th U.S. NAVY SYMPOSIUM ON MILITARY OCEANOGRAPHY THE PROCEEDINGS OF THE SYMPOSIUM Jo-1z fee. IGG7 VOLUMETL NAVAL RESEARCH LABORATORY WASHINGTON, D.C. Each transmittal of this document outside the agencies of the U.S. Government must have prior approval of The Oceanographer of the Navy, 732 N. Washington Street, Alexandria, Va. 22314. publication substantially be requested from the on Station, Alexandria, th U.S. NAVY SYMPOSIUM ON MILITARY OCEANOGRAPHY THE PROCEEDINGS OF THE SYMPOSIUM VOLUMETL NAVAL RESEARCH LABORATORY WASHINGTON, D.C. nme ; ee A a a : “t Be. SPONSOR Oceanographer of the Navy CO-SPONSORS Office of Chief of Naval Operations Office of Naval Research Office of Naval Material Naval Ships Systems Command Naval Electronics Systems Command Naval Air Systems Command Naval Ordnance Systems Command Naval Facilities Engineering Command Headquarters Naval Air Development Center Naval Civil Engineering Laboratory Naval Oceanographic Office Naval Ordnance Test Station Naval Research Laboratory Naval Underwater Weapons Research and Engineering Station Naval Underwater Sound Laboratory Naval Weather Service Navy Electronics Laboratory Navy Mine Defense Laboratory Naval Radiological Laboratory Naval Applied Science Laboratory Navy Marine Engineering Laboratory Naval Ordnance Laboratory Supervisor of Salvage STEERING COMMITTEE Captain R. A. Zettel Commander R. W. Haupt NGG5 IBS Ke (Goose Mr. Fred Knoop Dr. Victor Linnenbom Dr. Donald P. Martineau Mr. W. H. Hymes Mr. Murray Scheffer PROGRAM COMMITTEE Dr. V. Je Linnenbom, Naval Research Laboratory, Chairman Dr. N. S. Rakestraw, Naval Research Laboratory Mr. J. Schule, Naval Oceanographic Office Mr. M. Schefer, Naval Air Systems Command Commander R. W. Haupt, USN, CNO (OPO9B5) Dr. H. McLellan, Office of Naval Research ii TABLE OF CONTENTS Page OPENING ADDRESS Honorable Robert A. Frosch. - »- « - -= s+ «6 «© - ©» - » » os ef vii BANQUET ADDRESS Honorable Robert H. B. Baldwin. ........+..+..2... X1il SESSION A - GENERAL SESSION AN ANALYTICAL REVIEW OF LESSONS LEARNED FROM THE H-BOMB SEA SEARCH OFF SPAIN Bo AS And ties: sibs veh ite! foc fel i: ferba lita Gey hehe ree eters Sige) Nom ceiratl Meaiomnctemte nie rte ce 3 SUBMARINE TOPOGRAPHIC ECHOES FROM CHASE V J. Northrop 29 NEL MANNED SUBMERSIBLE OCEANOGRAPHIC SYSTEM, MODIFICATION I A. J SSHVOSSET | ie. Wisi fd Veh celpetaet ew tot ac aire mnrcMARS at hMre OO Sellen stuasrn) eit 42 SESSION B - THE OCEAN BOTTOM ENVIRONMENTAL LIMITATIONS TO DEEP SEA SEARCH F.N. Spiess, J. D. Mudie, and C. D. Lowenstein. ---+.-.-- 69 DIRECT MEASUREMENT OF BOTTOM SLOPE, SEDIMENT SOUND VELOCITY AND ATTENUATION, AND SEDIMENT SHEAR STRENGTH FROM DEEPSTAR 4000 E.C. Buffington, E. L Hamilton, and D.G. Moore-.....- 81 VARIABILITY IN DERIVED SEDIMENT SOUND VELOCITY AS A FUNCTION OF CORE ANALYSIS TECHNIQUES AND ITS EFFECT IN DETERMINING THEORETICAL BOTTOM REFLECTIVITY J J..(Gallaeher 0 2S ie so ah ait eee iG te Sal STRANGE HOT WATERS AND MINERALS AT THE BOTTOM OF THE RED SEA J. M. Hunt and D. A. Ross; - +--+ +++ + + + » «+ « « 2 es 102 MILITARY SIGNIFICANCE OF DEEPLY SUBMERGED SEA CLIFFS AND ROCKY TERRACES ON THE CONTINENTAL SLOPE 2 apa een DEH Nee ceearormonion-olor Oyo. Yokvoua -oeO. biG a.)'9 row Ono Oo oO 106 iv SESSION C - OCEANOGRAPHIC PREDICTION REAL-TIME OCEANOGRAPHIC DATA FOR OCEANOGRAPHIC PREDICTION NMR AN es B Vesa Hide Mtr Lele Rope w coated 21 With the addition of a few new facilities, we can meet the non-military needs of both government and private industry. The Naval Civil Engineering Laboratory, for example, now cooperates with industry regularly to carry out joint tests of new industrial materials in the Lab's pressure test facilities, to the great benefit of both. Let me mention the third field in which the Depart- ment of Defense can accept the national mission-ocean weather prediction. The Navy has been in this business for a long time and we can't operate without it. To provide this service, the Department of Defense operates a world-wide observational, collection, and communica- tions network. Navy relies on the Coast Guard for accurate, timely, oceanographic data and services from their ocean stations, ice patrol, ice breakers, and oceanographic platforms. DOD observations are supple- mented by those from merchant ships, island and coastal stations both foreign and domestic, as well as from the Departments of Interior and Commerce. The product consists of storm warnings, weather forecasts, sea state and thermal profile production, and optimum track ship routes distributed by unclassified messages on known frequencies for use of specific military units as well as any one else. We are making sure that marine and fishing indus- tries will have the full benefit of the Navy's increas- ing knowledge in this field. These are just a few of the areas outside the basic sciences where the Navy is contributing to the non- military exploitation of the ocean. Let me mention one sticky point right here. When- ever Navy cooperation with the non-military community is mentioned, the subject of classification comes up. We are accused of classifying every piece of information we lay our hands on including telephone books. There has certainly been overclassification at times in the mili- tary, but Admiral Waters tells me that an estimated 90% of all raw data gathered from all our oceanographic platforms is completely unclassified. XVili The Navy, because the ocean is its daily operating environment, is spending over half the FY 68 Federal oceanographic budget of 462 million dollars. It not only makes good economic sense, but seems essential, that the oceanographic information and technology for which the Navy spends your money should do double duty by being made available to all other agencies and industries. In any case there is no real separation anymore between national defense and national, and even inter- national, welfare. The United States cannot be indefi- nitely safe in a hungry world anymore than we can be safe without a credible strategic deterrence. Our efforts in space should not detract us from devoting increased efforts and resources to the seas. Not only our welfare but our very existence may depend upon it. The Renaissance was an age of geographical explora- tion and an age of science. It was also an age of hope in Europe after centuries of instability and violence. We are living in a new age of exploration--of both outer space and underseas--which is dependent upon new scientific discoveries, sophisticated technology and advanced engineering. We also live in a divided world-- politically, militarily and economically--which offers promise of great hope to the well-off but fears of hunger and despair to the poorer. The exploration and exploitation of the oceans are vital to our military security, and that of our allies, until a more stable world framework evolves. Perhaps more important in the long-range, is that oceanography in the broad sense offers avenues of satisfying certain basic human needs essential to stability in future decades. It offers both avenues of world-wide coopera- tion and a challenge to American leadership, governmental and private, to devote our scientific, technological and management skills in furthering joint efforts. xix Session A GENERAL SESSION AN ANALYTICAL REVIEW OF LESSONS LEARNED FROM THE H-BOMB SEA SEARCH OFF SPAIN Part I: Search, Classification, and Recovery F. A. Andrews Consultant, Ocean Systems, Inc. and Research Professor, Catholic University Washington, D.C. , INTRODUCTION On 17 January 1966, an unarmed nuclear weapon was lost in the Mediterranean Sea off Palomares, Spain, following the collision of two aircraft of the U. S. Air Force. The weapon was retrieved 80 days after the accident from a depth of 2,850 feet six miles off the Spanish coast by Task Force 65 of the U. S. Sixth Fleet. Inasmuch as the entire operation (named SALVOPS/MED) and the organization of Task Force 65 was "Ad hoc" and staffed by personnel ordered tem- porarily from other duties, it was impracticable for the Search Commander to produce the definitive type of operation report required. This task, therefore, devolved upon the so-called "Technical Advisory Group SALVOPS/MED". It was this group, sitting in Washington and meeting daily, which arranged logistic support, provided technical guidance and in general terms was "home base" for the Search Commander, Two reports and an executive summary were actually written. The first, an interim report! which presented a description of the operation and listed the problems encountered was published on 15 July, 1966, for immediate use by navy planners and technologists. The second, a final report and a more thorough analysis which listed conclusions and recommendations from both SALVOPS/MED and a postulated recurrance was published on Feb. 15, 1967. The latter report is 1200 pages in five volumes. The executive summary3, meant to present to executives and supervisors in both defense and industry the significant lessons and implica- tions for the U. S. Navy found in the final report, was also pub- lished on 8 April 1967. The two reports and the executive summary are available through the Defense Documentation Center, The actual task of drafting all reports was assigned to Ocean Systems, Inc., under contract to Captain W. F. Searle, Jr., Supervisor of Salvage, U. S. Navy, who was a key member of the Andrews AVOSI AO AVa uteds }SeayINOS JO SOD - T “*31q DjD9 aq adp> [ + ) SbdiguogiD>) f 7, DYDNIUDS) uos|!4 dwo> ge Nf? SasDWO|Dq SODHDI|IA a4is 4aysoW 0329g ~~ Se|!W Se PHPOW sojinby a41§ BADIS XI4 IH D299q [ Ss SH So]Dq aq adp> puaboji07) NGSX ee «SY Andrews Technical Advisory Group, as well as the officer responsible for the assignment of much of the technical equipment and personnel to Task Force 65. Ocean Systems, Inc., also acted as prime contractor for all commercial assistance to Task Force 65 in carrying out the H-bomb salvage operation. The general purpose of this paper and the paper which follows is to present in brief the significant information contained in the two reports cited before, and in the executive summary report. The author of the first paper was general editor of both the interim and final report and was with Task Force 65 for 10 days assisting in the establishment of an on-scene operations analysis group. The author of the second paper was the principal writer of the executive summary. The specific theme of this first paper will be the analysis of the Search, Identification, and Recovery phases of SALVOPS/MED. Significant lessons and recommendations derived from these phases will be listed. A BRIEF DESCRIPTION OF THE OPERATION'S SCOPE AND ENVIRONMENT The general area of interest is shown in Figure 1, The Naval Task Force which was collected in this area to conduct the salvage operations was ultimately composed of 25 navy ships, four research or commercial shins, four submersibles, and over 3000 men. Twenty civilian contractors were used to assist in the operation at a monetary value of effort of 2.1 million dollars. The search areas designated by Rear Admiral Guest, Commander Task Force 65, are shown in Figure 2. The first, named Alfa I, was a circle one mile in radius, located approximately 5 miles off- shore, and centered at a point where a local fisherman sighted a large parachute and object falling into the water. The second, named Alfa II, was approximately a semi-circle with radius 4500 yards, located immediately adjacent to the beach area where three of the four weapons had been found by Air Force search parties. The other areas, named Bravo and Charlie, were defined from Air Force calculations of possible bomb splash points given various combina- tions of ballistic fall (without parachute) and wind-affected free fall (with parachute deployed). AI and AII were considered to be the highest priority areas; Bravo and Charlie were considered to be lowest. The bathymetry of the area is given in Figure 3. One can note the relatively flat bottom terrain in AII, Bravo, and Charlie. The depth of water in AII ranges from 0 to approximately 300 feet. The depth in Bravo and Charlie is 400 feet slowly increasing to 3600 feet. The bottom terrain in AII was a hard, cemented gravel plat- form with numerous exposed black rocks. The terrain in Bravo and Charlie, on the other hand, was soft silt predicted to be as much Andrews \ Villaricos|y) | Camp Wilsonig, my ST PY fer et Garruchale =--=—-- ALFA I High Probab ility Are stablished 24 Jan. 1966 -=-— = =——— Initial Placement of Most Seaward Survivor Entry Point Mee, 4, Mar. 1966 Refinement of Most Seaward [~ Survivor Entry P. ‘oint Fig. 2 - Salvops search area 76 Andrews Villaricos » / ata ii ( a Zen ior S es ” ac Garrucha's i sv i 3710 B me aSac | | \ Fig. 3 - Bathymetric chart Andrews as 50 feet thick in spots. One can also note the worst possible place where a bomb might be lost, in the south part of Area AI. The northern part of AI is flat compacted silt ranging in depth from 990 to 1800 feet. From this latter depth on the terrain drops abruptly away into deep canyons. The southern half of AI is the eanyon slopes comprised of bed rock and a thin intermittent covering of OOZE with heavier deposits of OOZE in the canyon floor. Depths in South Al are 1800 feet to 3000 feet. The presence of a probable ancient river bottom directed southeast from the edge of AII and passing below area AI can also be noted. As a matter of interest, the nearby land terrain was equally rugged with mountain ranges rising from a mile or two behind the shore line. Surface currents in the SALVOPS area were generally to the southwest and reached a maximum of about 0.7 knots. Currents on the bottom in those two areas where observations were taken were also generally southwest and varied from 0.0 and 0.3 knots. In the canyon areas in the southern part of AI, the current was observed by submersible pilots to vary from southwest through south to southeast and was thought to depend somewhat on tide. The maximum bottom current observed was 1.0 knot. Surface visibility was usually 2 miles or better with water and air temperatures never severe enough to hamper operations. The most persistent surface weather factor was an afternoon off-shore breeze which on numerous occasions necessitated the suspension of small boat operations and at times curtained submersible operations. General storm conditions prevailed for no more than 10% of the time, during which time all operations were suspended. THE SEARCH PLAN The stated mission of Task Force 65 was to detect, identify, and recover material associated with the aircraft collision. To carry out this mission the Westinghouse ocean bottom scanning sonar (OBSS), the Navy mine-hunter sonar (UQS-1) and the Honeywell sea scanar sonar were used for acoustic search. Navy EOD/UDT divers, television lowered on a wire, hard hat divers, and Perry Cubmarine were used for visual search and identification in medium depths. The manned and free vehicles, ALVIN and ALUMINAUT, the manned and towed vehicle DEEP JEEP, and the unmanned and towed NRL vehicle operated from USNS MIZAR, were all available and used for visual search in medium and deep depths. ALVIN and ALUMINAUT also possessed an acoustic search capability (Straza sonar on ALVIN and Westinghouse side-looking sonar on ALUMINAUT). Outside of 80 feet the search and identification policy of the Commander Task Force 65 was; step one, search acoustically in all high probability areas reporting all acoustic contacts to the flag- ship, where a contact log and plot were maintained; step two, follow-up the reported acoustic contacts by sending visual Andrews e S 52 LPOG OO .¥, PS KKK XRG 4 oe et x 13 % 9 Me os SO RRS ees 2 ex ‘ ne RSSSOO Scale! GVG, > & PS CM e ek eS ©, Notes: (1) SEP is given by the large bold faced numbers. The small light faced numbers give the square designation. LEGEND .95 < SEP < 1.00 .00 d3S>00° 0S > d3S> Ol: Yj S6° > dS > 0s" Le 00°l > das >¢6° QN3947 *d5S 40} sanjpa OM} auy 40 4126p} ayy uodn pesog s! Bulppyus (€) "(2'ystwissed pup d14s!wi4ydo) uolysayap JPNSIA jNOgGD suoljduinssD judsa}41p OM WO) 4]NSe4 aupnbs YydDe Ul GAS 40 SAN|DA OM] auy (Z) ‘uolyouBIsap aipnbs ayy eaiB ssaquinu pap} 4yb1) ;jDws ayy 1l Andrews reported contact area had been made. An experienced member of the analysis team, (Lt. Cdr. George Martin, former pilot of TRIESTE I) determined which of the two values to choose. Figure 4 and 5 show the values of SEP in areas AII.and AI respectively the day before the weapon was found. Note in Figure 5 that search in AI was far from complete on the day the weapon was sighted in Grid C4. The effect of navigation degradation on the search is covered in detail in Part 2, Chapter V of the final report and shows amongst other factors that if one makes a random search in a given area (A), that (Pp) = 63% when W (sweep width) x L (transit distance) = A (area of nth grid) regardness of the navigation error. In fact, one can do quite well in detection by random search alone. The navigational inaccuracies will, however, impose a severe handicap if one wishes to revisit a contact which has previously been detected. The definition of SEP and its use in the method developed on SALVOPS/MED are considered valid and most useful for future deep ocean search operations. However, the particular numbers calculated as shown in Figures 4 and 5 are considered to have unknown inaccuracies because the inputs Pp and Pc were little more than educated guesses. Thus no pre-operational test data were available, nor were any valid on-scene controlled tests performed to check the accuracies of the assumed data. For interest, the values of sweep width (W) and navigational error (co) which were assumed are given in Fig. 6. The major recommendation resulting from this analysis phase of the Search and Identification operation were: « Establish an Operations Development Group which will develop search tactics, measure search parameters under realistic operational conditions, and improve further the method of operations analysis developed in SALVOPS/MED. « Provide all search forces with a trained analysis team and adequate computing facilities. » Improve facilities and procedures for obtaining and rapidly displaying environmental information so that the search Commander, search teams, and the analysis team can better assess day by day performance. Quantitative values of bottom currents, bottom sediment and strengths, and the charting of a micro-bathymetry survey are suggested outputs. The current measure- ments made by NAVOCEANO scientists were considered good, but did not cover the area sufficiently. Valid bottom strength and sediment measurements with adequate coverage were made by NCEL scientists, but they could not be processed soon enough nor displayed optimally. A bathymetry survey was made by the USNS DUTTON but showed the usual inaccuracies due to the wide directivity pattern of the UQN fathometer. 12 Andrews das sutyep~nojed ut pesn (0) 10170 uoTjestaeu pue (mM) UJptmdooms Jo sonTea poumnssy - 9 °3Ty (TenstA) LAVNINATV pue NIA'TV BIOWCD peMOL UVZIWN Ssao G T-son yorees wopuey GG G8 13 Andrews Camp Wilson Villaricos / ps Sasi fee ee 7 oat NZ Se ee SYS ak ES LEGEND 1 @® Aircraft Debris 4. * Contact 261 2h OQ Non-aircraft 5. All Contour Lines are in Fathoms Except Debris 80 foot Contour. 3. + Natural Objects 6. Odd Grid Numbers Omitted. Fig. 7 - Visual sightings 14 Andrews THE SECOND OPERATIONS ANALYSIS (POST-OPERATION) The second analysis, conducted after the operation, sought to determine any relation between acoustic and visual search. A plot of all visual contacts was made. This was then considered to represent the true picture of debris and rock distribution on the ocean floor off Palomares. Figure 7 shows this plot. An important point to note is the high debris density in AII and the great number of rocks also sighted in AII. The location of the nuclear weapon is shown. With this chart and with data available from the contact logs, each acoustic contact was reviewed and placed into one of the following categories. NF - no visual follow-up NS - no sighting in the visual follow-up OC = a sighting was made in the visual follow-up and its position, its depth, and for OBSS only, the approximate size of the contact correlated with the acoustic report OD = a sighting was made but only one of the criteria sited under OC seemed to correlate with that initially reported by acoustics. A correlation in position was considered to have occurred if the reported acoustic position was within 500 yards of the reported visual position. For depth the figure was 80 feet. For correlation in dimension, agreement within 1-2 feet for large objects and 50% for small objects was required for one dimension only. The other dimension had to agree within only an order of magnitude. The contacts in categories OD and OC were further divided by considering (a) the availability or nonavailability of DECCA to both detection and classifier units, (b) the location of the contact in an area of high or low density as determined by the chart of visual contacts. Figure 8 gives a summary of the analysed data for the UQS-1 and OBSS Sonars. The large number of OBSS contacts in Class NF was due to lack of time to follow-up all reported contacts and the fact that many of these contacts were outside of the priority areas AI and AII. The significant conclusions from this study were 1. The use of the OBSS sonar in SALVOPS/MED was relatively ineffective because of the tactics used, the high percentage of rocks in the operations area and the conduct of part of the OBSS search prior to the availability of DECCA. Specifically OBSS was ordered to make single passes through an area once with a small overlap. No attempt was made to regain iS Andrews $}9ejUOD OTysNode Fo stsATeue uy - g °3BIq pej10dey $]0BjUO0D [e1OL 16 Andrews $]DeJUOD TENSTA FO uOTIeINGey, - 6 “3tq UV ZI ‘HH pue AL daar daa ANTUVINNO NIATV LOVNIWOATY Lan/doa SLOATC “HH AL f Bre sTiqo Ajjuepuede dn=Mo]]O Sj09lqO qyerorry Taqod s3urysis I P pul TO jeanyeNn FeIOITY pests UL pezqsts LOTIIUSpT uUON [e1OL (q) (ez) yesi1ey, odAy, 17 Andrews or reinvestigate any contact. Instead, if the contact looked food on the OBSS trace readout, it was reported by radio to the control ship. Thus the navigation error of detection ship and classification ship were accumulative, possible communication errors could be intro- duced and no opportunity was given to the OBSS operators to learn whether the targets they were reporting were good or bad. As a result there are only 13 OBSS contacts in the OC class. Of this number 9 were rocks. Of the 66 CBSS contacts in the OD category, 39 were reported or followed up by units with no DECCA, and 32 were reported in areas later determined to be high density areas (i.e. > 5 visual sightings/ 1/4 mile sq.). 2. The UOS-1 was of limited use on independent broad area acoustic search and only in shallow water (i.e. < 300 feet). Specifically, the minesweeps with UQS-1 carried out certain mine-hunting tactics in which a contact was regained several times before it was considered a valid detection. Thus the percentage of UOS-1 contacts (16/34) in the OC category was an improvement over the OBSS, Still the inability to know whether the contacts reported were good or bad preempted any learning process by which the sonar operators could improve their skill. A second tabulation Figure 9, shows the performance of visual identifiers. The significant conclusions from this study were: 1. Cubmarine, and the EOD/UDT divers were the leaders in sightings of aircraft debris in AII. The relative immobility of hard-hat divers and "dipped" TV, and material problems with DEEP JEEP accounts for the poorer sighting performance of these three elements. 2. The team combination of a minesweep with UQS-1 and the EOD divers or Cubmarine was a most successful method for searching acoustically and simultaneously identifying a target. Mr. Barringer, Senior Pilot of Cubmarine, indicated that Cubmarine was vectored to 75% of the contacts listed in Figure 9. This is a significant fact since it correlates with a report by COMIN DIV 84 that the ability of his UQS-1 operators to distinguish rocks from debris increased signifi- cantly when they received back an immediate report from Cubmarine or an EOD diver concerning the nature (rock or debris) of the target which had just been obtained acoustically. A tabulation of sea scanar performance (not given here) showed that this sonar mounted on a small boat and teamed with divers was also relatively effective. 31 acoustic contacts were made of which 24 visual sightings were made in the immediate follow-up. However, over 40% of the contacts sighted were rocks. A major difficulty was 18 Andrews (THON JO YITUIS *yY 10;7e) Yous] uodeoM - OT “STA i J “AUND Bulso jiady 7 uo yurod siyy wos Aiadoray °¢ Ri “[udy Z UO NIA TW Aq p2t20ja1 uodoay, “yp | 7 “YIDW PZ UO (3)TYVHD UP{q) sdweayyp Asaaoda. Buiunp pajiod aul) a1ojaq wojjoq Woy 4J0 441] $o yUlod payowlysy “E “YD4DW SL UO NI ATV Aq uodbam jo BulyyBis ysur4y -Z | \ : *y201y suodDam jo Buruuibeg "l aay SS “TOTP IES. ce“ t ee ee ) 2 1 ~ “ yidep “4 OSsz_ 1, IAOTS oO€ ILVWIXOUddV a, Ne 25 ve -. yidap"yy OSIzZ “® \ LAL ee } ee BAG We VadO1S oOl = ee ate ACS aLVWIXOSddV SS NY é 19 Andrews in accurate navigation of the small boat, which had to be vectored into position by radar posits from the USS Boston. 3. Acoustic search in the south part of AI was totally ineffective because of the inability of sonar to receive anything but bottom reverberation in the highly discontinuous terrain. Visual search was effective even though detection ranges were rarely greater than 15 feet. The weapon was found by a modified random visual search in which a combination of fortunate circumstances counterbalanced the handi- cap of limited vision. ALVIN elected to search along terrain contours, and correctly deduced that the bomb might have slid some distance down one of the steep slopes, leaving a discernible track. Thus ALVIN's courses were designedly perpendicular to the track of the weapon sliding down a slope. This effectively enlarged the target to 1000 feet, the length of the track at the base of which the parachute enshrouded bomb lay. (Figure 10) The major recommendations of this post-operation study were: . Incorporate into the design of future search systems the target classification lesson learned in SALVOPS/ MED. This lesson indicated that in the rocky terrain off Palomares, acoustic search alone was generally ineffective because of inadequate sonar classification. Future vehicles which combine the integrated use of acoustic, magnetic anomaly, visual, and other detection means are indicated. In addition, emphasis on long range research and development in acoustic classifi- cation is indicated. ° Accelerate the development of mobile hull-based and bottom mounted transducer underwater navigation systems. Efficient bottom search, the revisit of contacts previously reported, and a study of the location of a contact relative to the bottom terrain are all impossible without precise sea-floor navigation. AN ANALYSIS OF THE RECOVERY The extreme depth at which the bomb was lost required that the attachment of the lifting device be done remotely either from the surface, or from within a small submersible. Lifting, once the attachment was made, appeared most practical from the surface. Certain schemes involving an unassisted lift by ALUMINAUT were considered but finally abandoned because of the possible risk to the vehicle, its personnel or to the weapon and parachute. 20 Andrews NOTE: Lizard had attached: 2 - 37ke Pingers 1 - 16ke Transponder ‘4 - 40# Concrete Clumps ALVIN was to snare chute/shrouds/ weapon with grapnel and hook. Lift by 3" Nylon. 250° 3" Nylon —m [I / aa ee os Brass “ 1-1/2" Dia. V Mild Stl. Fig. ll - Recovery plan ALFA 21 Andrews Three lift plans were actually attempted using various devices constructed on scene. The first two failed, the third resulted in lifting the weapon clear off the bottom. However, the lift line parted owing apparently to chafing and the weapon fell back to the ocean floor to be temporarily lost for "9 agonizing days." The NOTS (Pasadena) controlled underwater recovery vehicle (CURV) finally brought the bomb to the surface on 7 April 1966. The following description of the three recovery attempts (plan Alfa, Bravo, and Charlie) and the successful lift by CURV is taken largely from a memo by Lt. Com. M. MacKinnon, who participated in the operation as part of the staff of CTF65. The initial plan (Alfa) involved the use of ALVIN to place an anchor stake in the sea floor (compacted silt) to which a small polypropylene line was attached, leading to the surface (Figure 11). A traveling "lizard" was locally constructed to slide down the poly- propylene line carrying a lifting line and shorter attachment pendants. After the lizard reached bottom, ALVIN was to use her manipulator, pick up the attachment pendants and hook them into any combination of chute, shrouds, risers, and weapon. Once attached the lift was to be made from the surface using the USS Hoist (ARS-= 40). On 16 March, ALVIN succeeded in placing the stake in the bottom, but on 18 March when the catenary was being removed from the polypropylene line the strain imparted pulled the stake from the bottom. Only one of the two flukes on the stake had apparently taken hold. A second plan (Bravo) was employed using MIZAR's towed sled to place the attachment pendants adjacent to the weapon (see Figure 12). The final attachment then was to be made by ALVIN in a manner simi- lar to that of plan ALFA. A key aspect of the operation was the position keeping of MIZAR to enable close placement of the sled. This precise station keeping proved impossible and hence the plan was abandoned. The third attempt (Charlie) again used MIZAR, but this time in conjunction with placing an anchor adjacent to the target (Figure 13), The use of a 16KC transponder and MIZAR underwater tracking system enabled the placement within 100 feet. The lizard of plan ALFA was modified and secured to the lifting system. Three lifting pendants, one on the anchor, and two on the modified lizard (now called POODL) were to be attached to the weapon by ALVIN as in ALFA and Bravo. On 24 March ALVIN succeeded in attaching to the weapon a lift line from the anchor, by ensnaring at least 6 shroud risers. The other two lifting pendants were fouled and could nct be used. The decision was made to lift, but shortly after the strain was the heaviest the lifting pendant parted. 22 Andrews UTE Hydrophone (H/P) <= _ 3/8" Wire MIZAR Tow Vehicle Cargo Hook ve Grapnel Fig. 12 - Recovery plan BRAVO 23 Andrews a: een a 3" Nylon <—@ Swivel $ in the frequency range of interest (4-400 eps). The hydrophone had a sensitivity of -79 dB re 1 volt per microbar. The signals were passed through a frequency dividing network? and split into two bands, 3-20 cps and 20-200 cps. Higher frequencies were passed on to a third stage amplifier and rectified prior to recording. Signals in the low- and intermediate-frequency bands were passed through a 40 dB (variable) amplifier. Each amplifier had high- and low-gain outputs with 20 dB separation between channels. The signals were recorded on an oscillographic camera, a pen recorder, and magnetic tape. Analysis of Data Arrival times and signal levels were picked on the records and total path distances computed from the elapsed time between shot and receiver, assuming a speed of sound in water of LolAy km/sec. The events recorded after the direct hydroacoustic wave were correlated, where possible, with known submarine promi- nences by the use of a calibrated scale in conjunction with a 36 in. diameter globe. For each reflection travel time, one end of the scale was placed at the shot position, one at the receiver position, and an arc of possible reflectors traced. Two assump- tions were made in analysis: (1) that the hydroacoustic waves traveled along great circle routes between source, reflector, and receiver with no horizontal refraction, and (2) that the main packet of energy traveled via continuously refracted SOFAR (Sound Fixing and Ranging) paths in the deep sound channel. RESULTS The Precursor The earliest arrival on the FLIP records was a low- frequency (~5 cps) precursor that emerged from the background 23 sec ahead of the direct hydroacoustic wave (Figs. 3 and 4). The precursor probably corresponds to the ground wave of normal-made propagation and the main peak to the "rider" wave of Pekeris. 30 Northrop Similar precursors have been observed from large underwater explo- sions in the Pacific (C. T. Johnson, personal communication) and earthquake T phases.9 The Direct Wave The main hydroacoustic wave was recorded on FLIP at 06he5m175 GMT indicating a velocity of 1.47 km/sec, which is the speed of sound in the SOFAR channel. Although the shot depth was slightly below the depth of minimum velocity, energy was introduced into the SOFAR channel because of the sound velocity structure in the shot area (Fig. 5). The sound energy traveled outward from the source via a combination of Surface/Bottom reflected (SR/BR) paths and continuously refracted (RRR) paths. Over the long ranges utilized in this experiment, the SR/BR rays have reflection losses of 1 or 2 dB at each reflection and eventually become unimportant (the rays converge about every 35 miles, indicating at least 50 surface reflections and an equal number of bottom reflections over the path to Hawaii). Thus, only RRR sound rays need be considered in this analysis of the underwater acoustic signals received on FLIP. Sound energy was received at the off-axis FLIP hydrophones at a level of about 66 dB above background noise (Fig. 6). An oscillographic record of the direct arrival (Fig. 4) shows disper- sion in the low- and intermediate-frequency bands. Peaks in the high-frequency band correspond to individual sound ray arrivals before the main (overload) arrival. After the main arrival, the first and second bubble pulse arrivals are evident in the high- frequency (rectified) trace. Topographic Reflections In this section, the prominent reflections are identified from arrival times and signal levels of Figs. 3 and 6. The location of these reflectors is shown in Fig. l. The first reflection appears on the record before the complete decay of the reverberation from the direct wave and is difficult to identify. However, the observed travel time is approximately equal to that computed for a reflection from the Mendocino Escarp- ment. This reflection is followed 2 min later by a long-duration return from the continental slope of the North American continent (Reflector A of Fig. 1). The next series of reflections is from the Hawaiian Arch (Reflector B of Fig. 1). ‘The reverberation lasts for 25 min (Fig. 6) and includes reflections from Hawaii, Maui, Molokai, Oahu, Kauai, Nihoa, Necker, and Midway islands, as well as the many reefs in the vicinity of French Frigate Shoals and Gardner Pinacles. Individual, short duration reflectors from the neigh- borhood of Mellish Seamount (29°N, 171°W) are also present in the data. 31 Northrop Reflections from the Aleutian Islands (Reflector C of Fig. 1) last for 15 min. Travel time considerations indicate that the first of this group of reflections is from the insular shelf south of Kodiak, Alaska. Continued reverberations were received from the Aleutian Ridge as far west as the insular shelf south of Adak Island. A series of short reflections from the Emperor Seamount Chain (Reflector D of Fig. 1) account for the next set of reverberations. The first of this set of reflections is from a seamount on the southern end of the seamount chain near 32°N, 179°E. The other seamounts then cause a series of short (~10 sec) reflections on the record. Echoes were recorded from the southern seamounts before the northern ones. A reflection from the seamount near 4OON, 169°E is the last to be received because the more northern ones are in the shadow of the Aleutian Ridge. Identification of individual reflectors has been impossible because of the multiplic- ity of seamounts and relatively poor bathymetric charts in the area. After 8 min of low-level noise, a series of reflections was received which lasted for 12 min. Individual bursts within this time lasted from 10-40 sec, with an average of 21 sec. The only mapped submarine topographic feature that could account for these reflections is the large (about 600 miles(960 km)long) shallow rise known as Volshouki Ridge (Reflector E of Fig. 1). This set of reflections is therefore ascribed to individual peaks on the Ridge. Reflections from the Kuril Islands next appear on the records (Reflector F of Fig. 1). The first echo is from Long Island (46°N, 150°E), the more northern ones being in the shadow of the Aleutians for great circle paths from Cape Mendocino. The Kuril reflections all exhibit relatively low signal-to-noise ratio (Fig. 6) because of: (1) the long ranges involved (~8000 miles (1300 km)), (2) the interference of the intervening Emperor seamounts, and (3) the onset of propeller noise from an approach- ing merchant vessel in the vicinity of FLIP. The last reflection identified on the FLIP records corres- ponds to the submarine slope off the Island of Hokkaido, Japan (Reflector G of Fig. 1). This signal lasts for over half a minute before it is finally lost in the increasing ship noise which dominated the remainder of the record. Signal Level The signal level from a one kiloton charge at SOFAR depth can be calculated from the equation: TR=I, - (71+15 logig R) - 0.0035 R (1) 32 MPL-U-20/67 Northrop where Ip is the level in dB re il dyne /em? at a distance R in nauti- cal miles from a shot of intensity Iy» the effective source strength (intensity at a range of 1 yd). For spectral energy within a specific frequency band, an auxiliary formula may be used: Te @ to ey) = 1, + 10 log 2/n ae (/*" = tan (e,/s ) (2) where fy and f. are the upper and lower frequencies respectively, shatel an “aS Gehal fisast somes frequency at the range where the pressure is 100 psi. Using these formulas and assuming spherical spreading to a distance of 10 miles and semi-spherical spreading (4.5 dB per distance doubled) beyond (Equation (1)), a calculated level of 69 dB re l dyne /eme is computed in the frequency band 3-20 cps. It is difficult to speculate on the observed signal level for this direct arrival because of recording system overload. From the data available, it appears that levels near 66 dB above ambient noise were recorded for the low-frequency band. The level for intermedi- ate frequencies was about 20 dB lower because the greater part of the energy for large explosions is in the low frequencies and ereater attenuation is experienced at higher frequencies (0.005 instead of 0.003 dB per nautical mile). Discussion Recording of topographic reflections from CHASE V was selective at the FLIP site in that the Hawaiian Arch prevented reception of reflected signals from the South Pacific. Even with this reduction of reflectors, it was still difficult with the method used to identify all of the arrivals noted. The first few returns after the decay of the main arrival were fairly easy to identify, but after that the multiplicity of targets that satisfied the travel time considerations increased. For example, the round- trip travel time shot to Vancouver to FLIP is the same as that from shot to Hawaii Ridge to FLIP. Therefore, even though multiple reflection paths (shot to Hawaii to Aleutians to FLIP for example) were excluded because of travel time considerations, the multiplic- ity of arrivals was such that individual targets were difficult to identify. Further identification by arrival time differences between FLIP, TERITU, and YAQUINA data were attempted to resolve these reflectors. However, the YAQUINA recorder was secured because of poor signal-to-noise ratio shortly after the Hawaiian reflection was received, and the TERITU in the interval between the Hawaiian and Aleutian reflections. Therefore, groups of arrivals with nearly the same signal level were isolated and labeled A, B, C, D, E, F, and G on Fig. 1. These general groupings are believed to be correct, although as the total travel time increased, the number of possible reflectors multiplied rapidly and errors in the later identifications are more probable than in the early ones. 33 Northrop Recorded signal levels for reflected sound are all below that of the direct wave because of increased spreading loss over the greater travel paths involved and because of scattering loss on bottom reflection. The signal level for Hawaiian reflectors is greater than the California Slope reflectors, probably because of the steeper slopes of the former and the lack of sedimentary cover. More distant reflectors produce levels not inconsistent with spread- ing and attenuation losses observed by Sheehy and Halley. f A curve of the level vs range computed from Equations (1) and (2) is shown on Fig. 6 to indicate the relative loss for individual reflectors. CONCLUSIONS 1. A low-frequency seismic precursor was recorded 20 sec before the SOFAR arrival. 2. The main hydroacoustic wave was recorded at a level near 66 dB above ambient noise. 3. Topographic reflections were recorded for D1 /2D hours after the main arrival. 4. Reflections from the continental shelf off California were 35 dB below the calculated level for a non-reflected path, whereas the other reflectors were only 20-25 dB lower. 5. Reflections from many seamounts of the Emperor Seamount Chain were received, as well as from Volshouki Ridge. 6. Reflections in the 3-20 cps band were 20-25 dB higher than in the 20-400 eps band. ACKNOWLEDGMENTS This work was supported by the Office of Naval Researcn. R. W. Raitt provided some of the records and, with P. Rudnick, gave valuable suggestions in data analysis. Information on the Ocean Bottom Seismograph results was provided by H. Bradner; the TERITU records by A. Furumoto; and the YAQUINA records by R. P. Meyer of the University of Wisconsin, and A. C. Jones of the Marine Physical Laboratory. HE. W. Werner gave valuable assistance in data reduction. 34 Northrop REFERENCES 8. 10. Underwater Systems, Inc., CHASE V Source Data Tech. Prog. Rep. 11, 10 pages, June 1966. Raitt, R. W., Geophysical measurements, Natl. Acad. Sci., Natl. Res. Council Publ. 309, 70-84, 1952. Northrop, J., and R. H. Johnson, Seismic waves recorded in the North Pacific from FLIP, J. Geophys. Res., 70, 311-318, 1965. Ea Squier, HE. H., FLIP acoustical self-noise, Univ. of Calif., San Diego, Marine Phys. Lab., Scripps Inst. Oceanog., SIO Ref. 66-12, 8 pages, July 1966. NDRC Summary Tech. Rep., Div. 6, Principles of underwater sound, reprinted by Committee on Undersea Warfare, National Research Council, Washington, D. C., 1946, Vol. 7. Jones, A. C., Seismic equipment manual, Univ. of Calif., San Diego, Marine Phys. Lab., Scripps Inst. Oceanog., TM-143, 48 pages, January 1964. Sheehy, M. J., and R. Halley, Measurement of the attenuation of low-frequency underwater sound, J. Acoust. Soc. Am., 29, 464-469, 1957. Pekeris, C. L., Theory of propagation of explosive sound in shallow water in Propagation of Sound in the Ocean, Geol. Soc. Am., Memoir 27, October 1948. Johnson, R. H., Spectrum and dispersion of Pacific T phases, Hawaii Inst. of Geophys. Rep. 34, 12 pages, June 1963. Johnson, C. T., W. P. de la Houssaye, and T. McMillian, Hydroacoustic signals at long ranges from shot SWORDFISH, U. S. Navy Elect. Lab. Res. Rep. 1212, appendix 73-94, 10 March 1964. 35 Northrop LEGEND CHASE FLIP CONTINENTAL SLOPE HAWAII ALEUTIANS EMPEROR SEAMOUNTS VOLSHOUK! RIDGE ee mmo oD bP O * KURIL ISLANDS BATHYMETRY OF THE PACIFIC BASIN HOKKAIDO Fig. 1 - Location of area showing shot point, FLIP hydro- phone position, and location of prominent reflectors 36 Northrop dITa wo3zy ureyshs uotsuedsns suoydorphy - 7 “BT F’ 311108 N3SNVN SNOHdOYNGAH SUYIM DIHdDVYSONGAH 3Sv9 AY¥311LVE C_»\ CUT fy UpLot_y Ux“ SivO13d LNVAONS ATIVYLNIN »»))D, slvols COU Y3LINMOT3 Y3SLSNMO14 YSINGSNVYL SYNSS3Yd NOU LOUSIA NIVHO YOLSINYSHL 318Vv9 SNOHdDOYGAH LNAWdINO| SNIGYO93Y 37 Northrop wyodo'z; SGNV7S! TMH SE are None es rear nen taptennmconenacicesgeetesacn ins rere a cw i ee OSU0N oe ores cele amemnorSnveseor sca & sees ncaa nen iibhl ieeet mine een tee ee ~ 8do Si-p puog Asuenbely SLNNOWVSS YOuSdWA es XZ 3SVHO WOS SNOILOS1IS3Y DIHdVYDOdOL ane a ean ~ QPol | euoydoupAy -@ «euoydouphy The record strips are not 20 cps. Fig. 3 - Sample brush records from two hydrophones in the frequency band 4 from which the 10nSs . led. i continuous but show typical reflect 6 were comp data of Fig. 38 Northrop spueq Aouenborj-ysty pue ‘-o}eTpourtezUt ‘-MOT UT DACM oTysnooevoapAy UTeUT BY} JO p1ODET eTaUTeD OtydersdOTIIOSO - F “BTA asind asind ADM aqqng pug = 9ygqqng 45/ sei HWA my iT : )Py!S284 || 4-4 ‘UD fi NIV NL TIMAT TN | HOIH HIN THIN | | eu AT eae ST AGATA =O sd9Q000¢€—00S ~ Aouanbay ybiy sd900Z—Oz ~ Aouenbay aypipewuejul sddQ2—¢ ~ Aouanbea.) Mo} " = x Ty " Low JI — 39 DEPTH (ft. x10?) {o} N RANGE (Nautical miles) ite) I Northrop ‘Et ia SOUND VELOCITY (ft./sec.) 4850 4900 4950 5000 ae Fig. 5 - Ray diagram and vertical-velocity profile in the shot area 40 28, above ambient 60 40 JO 20 10 300040005000. 60007000 6000 9000. 10000 000 12000 wo SEISMIC PRECURSOR Northrop Wg LEGEND Observed Level — = 3-20cps so = 20-200cps Calculated Level +b Wy my) 5 a wn a < E 4 uJ = E z o 41 O — SEAQUAKES (?) Distance, km Fig. 6 - Sound pressure level vs range for CHASE V hydroacoustic waves. Pressure level is in dB above background noise. Slope of calculated curve is 4.5 dB per distance doubled (Reference (10)). 41 13000 NEL MANNED SUBMERSIBLE OCEANOGRAPHIC SYSTEM Modification 1 A. J. Schlosser U. S. Navy Electronics Laboratory San Diego, California 92152 Abstract: Manned oceanographic research within the Navy utilizing deep research vehicles began with the assignment of the bathyscaph TRIESTE I to the Navy Electronics Laboratory in Sep- tember 1958. In June 1966 the first complete manned submersible system was established by the leasing of the Westinghouse Diving Service (DEEPSTAR DS-4000) and integrated with the scientific and professional capability established at NEL for eight years. During this eight-year period the technical and professional experience was gained from the use of the bathyscaphs TRIESTE I, ARCHIMEDE, and TRIESTE II, and the Cousteau Diving Saucer. This achieve- ment provided an integrated and diversified system capable of conducting geological, biological, acoustical and physical oceanog- raphy investigations. Geological investigation systems involved use of coring devices, shear measurements of the sea floor, and photography of the various geological features prominent in the area. Biological investigations used listening devices, sampling methods of biological specimens and tracking of the scattering layer coordin- ated with echo soundings from a surface ship. Acoustical investiga - tions were made throughout the water column measuring sound speed, pressure and temperature. Shear velocity through the bottom was measured by an explosive source and geophones. Addi- tional investigations were accomplished by probes containing trans- ducers implanted into the sea floor. Physical oceanography results were obtained by water-sampling temperature probes at various in- crements above the sea floor. The above instrumentation systems will be discussed. Manned submersibles, unique as they may be, are ineffective tools unless integrated into an oceanographic system with specific scientific mission profiles. 42 Schlosser INTRODUCTION Oceanographic research received a new tool with the acquisi- tion of the bathyscaph TRIESTE in September 1958. Scientists at the Navy Electronics Laboratory began developing techniques that would allow maximum utilization of the bathyscaph for data collection. As technical competence increased the acquisition of more meaningful data was made possible by specially-developed instrumentation. Valuable knowledge was gained also from diving experience on the French bathyscaph, ARCHIMEDE, and on TRIESTE II. The minimal- payload Cousteau Diving Saucer was utilized to conduct shallow-water investigations. Due to the payload and attachment restrictions of this latter vehicle, new instrumentation techniques were necessitated. One of the most important lessons learned was that data collection applic- able to the oceanographic disciplines of acoustics, biology and geology was possible from any manned submersible if the instrumen- tation could be designed with a degree of flexibility to allow this interchange. Paramount to any instrumentation system was the requirement for adequate ship facilities to support the submersible and the scientific program. DISCUSSION The various scientific disciplines at the Navy Electronics Lab- oratory engaged in the military aspects of oceanography require a submersible with a complete-facility support concept. The support requirements include: (1) a surface support ship; (2) extended at-sea facilities with adequate berthing; and (3), the personnel to maintain and operate the ship and submersible. This service was best avail- able through lease of the Westinghouse DEEPSTAR DS-4000 and its support facilities (see Fig. 1), which provided the scientific community with a highly mobile manned-submersible system and freed the technical community to concentrate on the instrumentation required for each particular scientist within his discipline. By merging a leased diving service to a scientific program with firm mission profiles*, a manned-submersible oceanographic system was established. The leasing service was based on ''oper- ating days" throughout the period of contract® in order to satisfy the quantity of dives required. 43 (esnoysutjsem jo Asojyrnos hq yderso0j04d) AGIL HOUN| ‘dtys y10ddns oy} preog uo 000F-SC UVLISdAAC Psnoysurjsem - [ “S1q Schlosser 44 Schlosser SURFACE SUPPORT SHIP The surface support ship was equipped with a Gifft echo sound- er and recorder which was incorporated as standard instrumentation. This equipment could be trailed over the side for bottom profiles by each scientist at his discretion prior to diving at a given site. To further provide scientific support, a small shipboard van was utilized with the basic at-sea requirements for minor servicing and repair of electronic components and instruments. INSTRUMENTING THE SUBMERSIBLE Standard instrumentation aboard the submersible was estab- lished for the purpose of recording each mission profile. These instruments were included as part of the scientific payload, but were made removable where possible to provide payload flexibility. The movie camera was removed on many dives in order to add other scientific instruments. External Attachments Jettisonable instrumentation brow. The flexibility necessary for the "systems" approach called for an instrumentation-attachment device which would not only provide attachment points, but increase the payload to fulfill mission profiles. The NEL-designed, remov- able instrumentation brow’, utilizing syntactic foam, provided this capability and payload was increased by 220 pounds. Shipboard changes could now be made, providing mission versatility and more efficient utilization of the submersible. Tethering. The ability to hover at various increments of height from the sea floor necessitated development of a simple teth- ering device. Fig. 2 illustrates a typical tether arrangement. The procedure followed on this type of dive was for DEEPSTAR to descend full-tether distance from the bottom. By means of forward propulsion during final descent, the tether was allowed to string out aft and lay in that position as measurements were made on the sea floor. Small weights equal to the weight of the first tether were then dropped and DEEPSTAR would ascend to the desired height above the sea floor and provide a stable platform for further measurements in the desired increments of distance from the bottom. Lighting. DEEPSTAR was equipped with external lighting con- forming to specifications’, and sufficient for visual identification and viewing. 45 50 FT. SOFT. Schlosser DEEPSTAR ELECTRICAL CABLES FOR SEPARATION BOLTS BOLTS %, COTTON LINE ISLB. WEIGHTED WITH WEIGHT SOLDER WINDING IS LB. WEIGHT Figure 2 46 Schlosser Manipulator. A manipulator, or mechanical arm, capable of collecting specimens was provided as part of the submersible. The arm was incapable of use for or adaptation to extensive core sampling but met the minimal scientific needs. Instrumentation Power and Signal Requirements The instrumentation power and signal requirements were made independent of the control and power supply systems of DEEPSTAR. Silvercell batteries supplied the DC power (24-28 volts). AC power was provided by static inverters. The power package was located within the personnel sphere beneath the observer's couch, Three power packages were assembled, providing two spares to be carried aboard the support ship, one of which was on recharge during each dive. A separate terminal board within the sphere consisting of 50 connections was allocated, thus allowing instrumentation changes to be accomplished without interruption to or interference with the sub- mersible's system. Through-hull penetrations allocated to "science" from the terminal board were externally terminated in eight female connector junction boxes, located forward of the sphere and adjacent to the brow. Four junction boxes providing 32 connections were available for brow instrumentation hookups. A single eight-connector junction box was located aft of the personnel sphere. All junction boxes were Electro Oceanic-type, therefore all instrumentation was equipped with suitable male connectors. This standardization was made on all scientific equipments. Standard Sphere Instrumentation Standard instrumentation is that which is added on request of the user laboratory or which may be already a part of the submersible, as covered in negotiations and specified as a contract requirement. These items may vary, dependent upon mission profiles established as part of the scientific program. Fig. 3 illustrates the sphere layout. The starboard side is devoted to the scientific program. Viewports. An observation capability to enable the scientist to view the water column, sea floor, or the performance of the ex- ternal instrumentation, was a ''must''. To insure continuous view- ing a separate viewport from that of the pilot was specified. This "instrument" obviously must be a part of the submersible. Cameras. Mission profiles established the necessity of photo- graphic capability, in color, for both still and motion pictures. The 300-frame, 70-"h still camera was mounted externally of the sphere 47 Schlosser Fig. 3 - DEEPSTAR DS-4000 shpere instrumentation controls 48 Schlosser with strobe lights providing illumination. Motion pictures were ob- tained by an internally-mounted 16-7h camera with a 400-foot film capacity. Film cannisters could be changed during a dive, providing unlimited motion-picture coverage. A separate camera viewport was located between those of the pilot and observer. Both the still and motion-picture cameras had the same focal point and the frame control was indexed at the observer's window. The control panel was located at the scientific observer's station, providing him com- plete control of the photography. Voice tape recorder. A tape recorder was made part of the standard sphere instrumentation, providing the scientist-observer with the freedom to continue vocal observations of the sea floor or the experiment without interruption or loss of visual contact. From past experience, the recorder reliability was deemed paramount over all other basic requirements set forth for this instrument. The high humidity within the personnel sphere had, in the past, precluded any reliance on the availability of continuous and complete dive records. The recorder selected was a battery-powered Stancil- Hoffman, hermetically sealed by an ''O'' ring, capable of 6 hours of continuous operations. The speed used was 15/16 ips. Two recorders were used throughout the program, each containing its own rechargeable batteries. A recharge unit was carried in the equipment van aboard the support ship, which provided for one unit to be on charge at all times and insured availability of power during extended at-sea operations. Two-channel oceanographic instrumentation package. In order to measure the temperature, pressure and sound speed through the water column, a standard suite of instruments was carried during each dive. This package was called the '2-channel oceanographic instrumentation package!” and provided data recorded on tape by frequency division. This information is now being correlated and will provide information on anomalies which occur within the water column at the various dive sites. This instrumentation package was located externally and aft of the personnel sphere. The sound- speed and temperature sensors were located so as to provide maxi- mum exposure within the water column (see Fig. 4). INSTRUMENTATION TESTING Prior to use or installation on DEEPSTAR each instrument was tested to provide as high a reliability factor as possible. These tests, while not always sophisticated, did achieve the desired re- sults. All electronics components were thoroughly tested within 49 a RIA Schlosser Fig. 4 - External temperature and sound-speed sensors for the 2-channel instrumentation package 50 Schlosser the laboratory and those to be externally mounted on DEEPSTAR were then encapsulated within their individual pressure cases. All pressure cases were proofed beyond the minimums established by the Safety Certification requirements. Each component was then tested in the environment by means of a dockside evaluation. Those items performing mechanical functions were tested at pressure in a 20, 000-psi pressure vessel. As each individual component tested satisfactorily, it was integrated into its particular system anda full- system test made. This latter was done as a bench check; however, a dockside test was always conducted when feasible. Upon installa- tion of each component in DEEPSTAR, a complete functional recheck was accomplished. This thorough testing procedure provided a high confidence level in the instrumentation used. BIO-ACOUSTICS Bio-Acoustics aspects of the mission profiles were primarily concerned with the scattering layers”. Fig. 5 is typical of the instrumentation added to DEEPSTAR for this program. The "slurper", developed by the Mechanical Engineering Division at NEL, is a specimen-collecting device actuated by the observer within the sphere. The transducer was mounted on the brow and enabled DEEPSTAR to go below the scattering layer and obtain up- ward observations which could then be correlated with echo sound- ings made from the surface ship. All photographic equipment was utilized during scattering layer operations. For some Bio-Acoustics applications, another transducer replaced the still camera (see Fig. 6). A metal sphere suspended from the brow was used for calibration of signal strength 6 GEOLOGY The vehicle configuration necessary to accomplish the varied sea-floor studies” *” ** within the mission profiles was "bare-boat" (Fig. 7) or with the brow (Fig. 8). The brow was used to enhance lighting for motion-picture photography by the addition of two quartz-iodide lamps. The standard lighting was reconfigured for these specific dives. Core samples were taken by use of the mech- anical arm which would select the core from the fairing attachment and drop it into the hydraulically-actuated specimen basket. This method was not completely satisfactory due to the deficiencies in the mechanical arm design as noted previously. However, the arm did prove valuable in collecting rock and biological samples from the sea floor. 51 Schlosser Fig. 5 - Bio-Acoustic instrumentation for scattering-layer studies Bye Schlosser Fig. 6 - Bio-Acoustics instrumentation showing transducer replacing still camera (lower left) 5)S) Schlosser Fig. 7 - “Bare-boat” for sea-floor studies 54 Schlosser Fig. 8 - Additional quartz iodide lamps and standard lighting reconfiguration for sea-floor studies. Note fairing attach- ments (lower right) for sediment cores. 5 Schlosser UNDERWATER SOUND PROPAGATION AND ATTENUATION STUDIES One of the more challenging applications of instrumenting the DEEPSTAR vehicle was that of the Underwater Sound Propagation and Attenuation Studies » 4 .-.) he) scientific payload of 350 pounds was not approached, but the mass of the external instrumentation presented shipboard-handling problems and the internal sphere arrangement utilized all available space. Fig. 9 shows coring tubes and probes to measure compressional wave velocity and absorption. Fig. 10 shows equipment for generating shear waves by firing dyna- mite caps,and geophones for observing them. To provide significant data, DEEPSTAR made an early-morning dive at a specified site using instrumentation shown in Fig. 9. An afternoon dive was made at the same location following an at-sea instrumentation change as shown in Fig. 10. Sea-state conditions limited diving operations in some instances, causing handling problems when geophones and dyna- mite caps occasionally became entangled during launching operations. PHYSICAL OCEANOGRAPHY Physical Oceanography studies called for specialized equip- ments capable of reliable performance at the 4000-foot operating depth of DEEPSTAR. The need to obtain water samples at the sea floor and at approximately 1-foot increments dictated that the design of water samplers be compatible with DEEPSTAR's maneuverability. Deep-sea solenoids developed at NEL performed reliably throughout the program. Fig. 11 illustrates the external instrumentation con- figuration used for this specific mis sion’. The measurement of current flow at the sea floor in ''virgin water"! (without interference of the vehicle hull) was accomplished by means of depositing the large bale attached to the current meter on the sea floor and backing off a distance of 10 to 12 feet for observation and recording. The Signal was carried from the meter to DEEPSTAR by a retractible communication cord, similar to that used on telephones, which allowed minimal wire length and precluded entanglement during meter recovery. Temperature probes to be embedded in the sea floor were carried in a ''cocked" position (see Fig. 12) so that the probes could be actuated by the scientist without disturbing the water sampling. The underwater separation bolt” provided an extremely simple device for actuating the probes. A dye marker was also used for current data. Dye, intermixed with sand and encased in "baggies'} was released into the current upon rupture of the ''baggies'' by the point of the temperature probe. 56 Schlosser Fig. 9 - Temperature probes and core samplers Sif Schlosser DEEPSTAR BOOM IN STOWED POSITION DYNAMITE CAPS lL, COTTON tt HSI 8 LINE Figure 10 58 Schlosser *soT110q 197eM pue sptouU -oTos ere yoeqT *oAp Butureqyuoo ‘yystr T9jUS. .,9133eq,, 910K] “punosdsoroF UI 1949 JUSTIN YIIM uoTIeJUSUINI;sut Aydersouess0 Teotshyug - [TT °3taq Bg Schlosser Fig. 12 - Separation-bolt release mechanism for temperature probes 60 Schlosser ACOUSTICS The acoustic program fully utilized instrumentation payload and available space, both within the sphere and that additional pro- vided by the brow” Teh, Attachments were designed to be readily removable to allow maximum flexibility by interchanging at sea. Fig. 13 is typical of the type of instrumentation used. The water- sample bottles are of the same design and type as in Fig. 11, but are mounted vertically with reversing thermometers. A dive series of three days at sea and at different locations teflected external changes as shown in Fig. 14. The maximum payload was utilized on these dives and, in many instances, the photographic equipments were removed to allow the scientist greater versatility in the instru- mentation system. Data was obtained and recorded by means of a 7-channel tape recorder. Fig. 15 shows a typical internal sphere arrangement for the acoustic mission profiles. ACKNOWLEDGEMENTS The author wishes to express appreciation to Dr. G. H. Curl for his assistance in reviewing this paper, to Mr. W. H. Armstrong for drafting services, andto Mrs. B. L. Hurt for editing. 61 Schlosser Fig. 13 - External acoustics instrumentation. Suspended at left are water-sample bottles with reversing thermometers; above left, velocimeters; top center, two temperature probes and velocimeter; and suspended at right, a salinometer. 62 Schlosser *LOJOUIOUTTeS JO UOTJeOOTEI yysITS pue *(103u99) aqoad oanzerodutay o81eT JO UOTIIppe ‘srojyomIOWI TOY} YWIM stoydures 19}3eM Jo TeAOUINI epnypout ¢{[ “3tq ut uMoYsS yey} WOLF *seties oatp Aep-¢ r0}ze uOTIeJUaUINASUT SOT]SNODY - FT ‘°3TqT sesueyy 63 Schlosser | EET Sat ey ey er = ee = = ~ SMe eae Internal sphere instrumentation for acoustics program ~15 - Fig 64 Schlosser REFERENCES Ne LO: ll. UZ 13. 14. 5) 16. "Specifications and Requirements for Lease Services of a 4, 000' Submersible, '' NEL contract N123-(953)54016A, 13 May 1965, "Appendix A: Original NEL DEEPSTAR Diving Program as Based on Scientists' Requirements (reproduction), '' NEL Deep Submergence Log No. 1, August 1966. Schlosser, A. J., ''Submersible Diving Service Leasing for Scientific Measurements, '' paper presented at Workshop on Use of Deep Manned Submersibles, WHOL, Woods Hole, Mass., 21-23 November 1966. Schlosser, A. J., ‘Instrumentation Brow for DEEPSTAR,"! NEL Deep Submergence Log No. 2, p. 31-34, October 1966. Seeley, R. L.,''Two-Channel Oceanographic Instrumentation Package, '' NEL Tech Memo in preparation. Barham, E. G., and Davies, I. E.,''Bio-Acoustics, '' NEL Deep Submergence Log No. 1, p. 31-38, August 1966. Davies, I. E., Barham, E. G., and Pickwell, G. V., ''Bio- Acoustics, '’' NEL Deep Submergence Log No. 3, p. 47-52, February 1967. Adams, R. L.,''Bio-Acoustics Scattering Measurements, "' NEL Deep Submergence Log No. 3, p.53-56, February 1967. Beagles, J, An, “Sea Ilo Siwchies, ") INJDIL, IDSEje) SleranewecineS HoreNon wl p. 25-28, August 1960. Buffington, E. C., 'Submarine Slopes and Gullies; Topography and Structure, '' NEL Deep Submergence Log No. 2, p. 27-30, October 1966. Dill, R. F., and Heezen, B. C., "Submerged Sea Cliffs and Rocky Terraces on the Continental Slope, '' NEL Deep Submer- gence Log No. 3, p. 61-62, February 1967. Bucker, H. P., "Underwater Sound Propagation, '' NEL Deep Submergence Log No. 1, p.29-30, August 1966. Whitney, J. A., "Underwater Sound Propagation, '' NEL Deep Submergence Log No. 3, p.57, February 1967. Hamilton, E. L., 'In-situ Sediment Sound Velocity and Atten- uation, '' NEL Deep Submergence Log No. 3, p.59-60, Feb- ruamy LOOT, Schlosser, A. J., Explosive Effect on Manned Submersible Viewports, NEL Tech Memo 933, 17 May 1966. LaPFond, E. €., et al, "Oceanographic Research (Sea Floor Inter -face--Chemical and Physical Properties), '' NEL Deep Submergence Log No. 3, p.17-45, February 1967. 65 Ito 18. Schlosser Mackenzie, K. V.,''Deep Ocean ASW Acoustic Research, '"' NEL Deep Submergence Log No. 3, p-63-72, February 1967. Mackenzie, K. V., ''Precision in-situ Sound-Speed Measure- ments, '' NEL Deep Submergence Log No. 3, p.79-99, Feb- ruary 1967. 66 Session B THE OCEAN BOTTOM yolk ENVIRONMENTAL LIMITATIONS TO DEEP SEA SEARCH F. N. Spiess, J. D. Mudie, C. D. Lowenstein University of California, San Diego Marine Physical Laboratory of the Scripps Institution of Oceanography San Diego, California 92152 During the study conducted by the Deep Submergence Systems Re- view Group in the aftermath of the THRESHER search it became clear that acoustic, magnetic and optical search systems should be developed to en- hance our ability to find objects on the deep sea floor. It was equally clear, as we investigated hypothetical search systems, that we had no information on which to estimate, even crudely, the limitations which natural roughness of the deep sea floor would provide to the effectiveness of system operation. This limitation might be either in raising the effective ‘noise’’ level at the system output in a general way or might appear in terms of false targets. With support of ONR and DSSP we have brought into operation a towed instrumentation system capable of observing pertinent fine scale aspects of the deep sea floor and have begun an actual program of operation at Sea to determine these limitations. The system consists of an instru- mented towed body which can be operated from a research ship, using an appropriate towing wire to provide both electrical and mechanical connection between the two. The observational systems which are operational at this time are precision down-looking echo sounder, up-looking sounder, bottom penetration sounder, side-looking sonar, camera and lights, and proton magnetometer. Transducers for most of these can be seen in the picture of the towed body (Fig. 1). High quality local navigation near the sea floor is provided by an acoustic transponder system .24/ All outgoing signals for the acoustic systems as well as trigger signals for camera and magne- tometer are transmitted from the ship down the coax core of the towing wire and all acoustic and magnetometer signals are sent back up the same line. We have, in the past year, operated in several areas, four of which are shown on the index map (Fig. 2). In three of these (shown by circles) transponders were planted and an area of 80 to 120 square miles 69 Speiss, Mudie and Lowenstein was surveyed in detail. In the fourth area a long east-west traverse (105 miles) and a shorter north-south run were made. The present paper will be concerned primarily with acoustic system data from the circle areas and magnetometer data from the long straight tows (although both acoustic and magnetic data were obtained in all areas). Magnetic search techniques are limited by inherent noise in mag- netometers, noise associated with their motion or the motion of the conduct- ing sea water in the earth's field, by ionospheric currert effects and by naturally occurring magnetic effects of the rocks of th. sea floor. The first few of these items lead to uncertainties of a few tenths of a gamma at most in well-designed systems. (One gamma is 107° oersted thus typically the earth's field is about 40,000 gammas.) These imply, then, that search for iron objects up to 100 meters across must pass at least within a kilometer or so of the target and thus that such searches at sea must be conducted near the bottom. With this in mind magnetometers were towed near bottom at depths of about 2.5 kilometers in the THRESHER search. $s shown in Fig. 3, taken from the paper of Maxwell and Spiess in Science,—’ and as more extensively documented by Heirtzler,-’ the natural geological background in that area, on the continental slope, was smooth. A 100 gamma anomaly, presumably due to THRESHER, is clearly visible against the gradual 30 gamma change occurring along the 1.5 km track shown in the figure. It has long been known, however, that anomalies of several hundred gammas can be observed at the sea surface over the deep ocean. One of the best magnetic maps showing these strongly lineated anomalies is that produced by Mason and Raff——’ as a result of surveys which they con- ducted in the middle fifties with cooperation of the U. S. Coast and Geodetic Survey. These show ridges which have north-south continuity in sections for 500 km or more and east-west wavelengths of 20 to 50 km, with ampli- tudes of 50 to 300 gammas. The question as to how these anomalies look when observed near the sea floor has now been answered with the long profiles at the location shown in Fig. 2. The result (as seen from about 80 meters off bottom) is shown in Fig. 4. The long wavelength component which would be measured at the surface is barely discernible, as are the small peaks introduced artificially by adding the anomaly of Fig. 3 repeatedly every 20 km. Itis clear that the naturally occurring anomalies, associated with variations of magnetization of material lying just below the thin sediment cover (50 to 300 meters), have strong short wavelength components which would make detection of ship-sized iron objects quite difficult. Since the anomaly of Fig. 3 has significant short wavelength content we have applied a simple signal processing scheme in hope of 70 Speiss, Mudie and Lowenstein improving the signal-to-noise ratio. The record of Fig. 4 was high pass filtered with and without the THRESHER peaks using a cutoff wavelength of 400 meters. The result is shown in Fig. 5, in which the lower trace includes the inserted targets. In some parts of the record these show clearly but there are many similar peaks generated by the background. Even with this treatment detection would be far from easy. Having found this strong short wavelength background in one area one may ask whether it should be expected elsewhere. The answer is that surface observations show similar lineated structures in much of the deep ocean and that the strong similarities observed among these various occur- rences make it highly likely that deep observations, when they are made elsewhere, will show this same fine structure. Figure 6 shows areas which we estimate will have background of a sort which will render magnetic search difficult. Acoustic search for objects on the sea floor can be carried out in either of two basic modes—one utilizing essentially a down-looking system and the other using more or less horizontal paths. If the down-looking mode is used in the deep sea from near the surface a reasonable sweep width can be achieved. Determination of the horizontal coordinates of suspected targets is uncertain without very sophisticated (and expensive) sonar systems. The downward looking technique can give good resolution if near- bottom equipment (towed and in future on free-running vehicles) is used. Figure 7 shows a profile over some small natural irregularities on the sea floor, at about 2200 fathoms depth 500 miles west of San Diego (area B). The upper portion shows the record of the surface-operated echo sounder over the same portion of track as that shown by the lower record from the deep tow. The lower record displays both the down-looking sounder output which gives the height of the towed body off the bottom, and (in the middle of the record) the sum of the down plus up echo sounder travel times, that is, the actual depth of the water. The small hill is very clearly discernible. While the near-bottom use of the downward path provides good resolution, this good resolution implies a very narrow search path (typically only a few fathoms wide). In fact once one has decided to operate near the sea floor it is immediately obvious that a nearly horizontal path is much more appropriate. Under such circumstances one should naturally utilize a system in which beam width and pulse length are adjusted to provide proper resolution and employ such a beam either fixed in side-looking mode or sweeping in searchlight or scanning sonar fashion. The most natural question to ask as one proceeds with such a search system design is what one might expect in the way of false targets arising from the natural rough- ness of the sea floor. It is this question for which we now have the begin- nings of an answer from the deep sea floor in the northeast Pacific. 71 Speiss, Mudie and Lowenstein In making such background observations it seemed most appro- priate to use a side-looking sonar. This, operated from a deep-towed fish, gives high resolution without unduly complicating the equipment problem (no scanning motors or circuits, easy mounting of the necessarily large trans- ducers, simple telemetry and display). At the same time results of such ob- servations tell of the sizes of irregularities on the sea floor in a form which can be made relevant in terms of other types of sonars. We have, to date, utilized a pair of Westinghouse transducers (shown on the towed body in Fig. 1) operating in the non-focussed mode at about 230 kcps. Height off the bottom is variable depending on the circumstances but records useful for our purpose, with objects as small as 5 feet high and 10 feet across in mind as targets, can be obtained at heights off bottom of as much as 200 feet. Normally a half-second or one-second repetition period has been used and most records show useful returns to ranges of 150 to 200 fathoms (0. 15 to 0. 2 nautical mile). Speed of advance has typically been about 1.2 knots. In all three of the areas of Fig. 2 a major fraction of the sea floor is essentially featureless to the side looker and would be very easy to search systematically and effectively at a rate in excess of a half square mile per hour. In every one of the areas there were, however, regions of sharp topographic relief in which the associated side-looking sonar record is patchy in a manner which would give rise to possible false targets or un- certainty as to whether or not targets were present. Three examples can be shown of side-looking sonar records as- sociated with rough topography. In Fig. 8 is shown a pass across a small canyon in the Hawaiian area (area D). This ditch was crossed at several points and was shown to be about 30 fathoms deep, 0.1 mile wide and at least 2 miles long. Its walls were quite steep (45°). The crossing shown here was made at an angle of about 60° to the axial line, a trend which is evident in the sonar record. The second (Fig. 9) was made in area A as the fish was being towed down the side of a seamount at a depth of about 1900 fathoms. Most striking from the viewpoint of possible false targets was Fig. 10 which was made close to the small hill shown in Fig. 7, in area B. In this area, along over 200 miles of track on which the total relief of two abyssal ridges was about 200 fathoms, there was only this one region, less than half a mile across, which showed other than a smooth side-looking record. As observed with the air gun and penetration sounder there was Significant sediment cover (at least 50 meters) in all these areas, except in the immediate vicinity of steep slopes. The only major exception 72 Speiss, Mudie and Lowenstein occurred in some portions of the Hawaiian location and in that region there were areas showing very thin sediment cover which correlated with very patchy sonar returns as shown in Fig. 11. This record is of double useful- ness inasmuch as it shows the sort of roughness which the non-sediment covered portions of the sea floor can have while at the same time giving indication of the resolution of which the equipment is capable. This provides reassurance that the much more extensive smooth areas are in fact (for ob- jects of this size) essentially featureless. In this picture the nearly round shadow at about 1157 implies an object 14 feet high and 90 feet wide along the track at a distance of about 400 feet to one side of the track. A knowledge of general geological sea floor patterns makes it clear that one can predict that search for objects having characteristic di- mensions of 10 feet or more should be highly effective in about 80% of the northeast Pacific. In volcanically controlled areas, as off Hawaii, the per- centage will perhaps drop lower, with areas showing no sediment cover pro- viding particularly difficult situations. It thus appears that any search for large objects in the deep northeast Pacific can be conducted using simple towed instruments in conjunction with small submarines with an appropriate division of effort between the two. Areas of steep terrain or those with scant sediment cover could be assigned initially for detailed search using manned craft with the towed gadgetry covering the usually much larger sediment-covered area and locating within it any possible targets or small patches of rough terrain for later manned investigation. At the present time, with only bathyscaphes available for detailed observations, this might still throw too much load on the manned portion of the system. However, as craft having greater mobility become available the combined system should approach compatibility with the environmental situation. It should be emphasized that only a small fraction of the sea floor has been sampled in this manner thus far, although areas were chosen to allow the most significant inferences to be drawn. Further observations in areas of different geological types are still required to allow formation of a truly worldwide picture. We would like to acknowledge the participation of the engineering staff which designed and built the equipment which made these observations possible: Messrs. M. S. McGehee, C. S. Mundy, D. E. Boegeman and M. Benson. It is also a pleasure to express our thanks to Capt. T. Hansen and the crew of the THOMAS WASHINGTON (AGOR-10) for their very effective support of our seagoing work, which was supported by ONR and DSSP. 73 Speiss, Mudie and Lowenstein REFERENCES McGehee, M. S. and D. E. Boegeman, MPL Acoustic Transponder, Rev. Sci. Instr., 37, No. 11, 1450-1455 (1966). Spiess, F. N., M. S. Loughridge, M. S. McGehee and D. E. Boegeman, Navigation: J. Inst. of Navigation, 13, No. 2, 154-161 (Summer 1966). Spiess, F. N. and A. E. Maxwell, Search for the THRESHER, Science, 145, No. 3630, 349-355 (1964). Heirtzler, J. R., Magnetic Measurements near the Deep Ocean Floor, Deep Sea Res., 11, 891-898 (1964). Mason, R. G. and A. D. Raff, Magnetic Survey off the West Coast of North America, 32°N Latitude to 42°N Latitude, Bull. Geol. Soc. Am., 72, No. 8, 1259-1266 (1961). Raff, A. D. and R. G. Mason, Magnetic Survey off the West Coast of North America, 40°N Latitude to 52°N Latitude, Bull. Geol. Soc. Am. 72, Not 8, 1267-1270 (1961): 74 Speiss, Mudie and Lowenstein Fig. 2 - Index chart.of operating areas 02 Anomaly, gammas Speiss, Mudie and Lowenstein —t WW ____-__—_—__}— TIME 0430 0445 0500 {2 JUNE 1963 Fig. 3 - THRESHER search magnetometer trace Or 40 80 120 Distance, kilometers Fig. 4 - Magnetic profile at location C with trace of Fig. 3 added every 20 km 76 Anomaly, gammas Speiss, Mudie and Lowenstein TARGETS ADDED O 40 80 120 Distance, kilometers Fig. 5 - Magnetic profile at location C, high pass filtered with 400 meter cutoff wavelength. Upper trace without and lower trace with THRESHER areaanomaly added every 20 km i | 4 Pa as Yi YZ E Gy , Yo Yo Y. LEGEND KNOWN {tes WITH MAGNETIC LINEATIONS OR HIGH SUSPECTED | MAGNETIC ROUGHNESS Fig. 6 - Magnetic lineations, world map 77 Height,meters Depth, meters 3300 4000 4100 Speiss, Mudie and Lowenstein | SURFACE ECHO-SOUNDER RECORD FISH ECHO-SOUNDER RECORD eee “BOTTOM PROFILE _ MEASURED Fig. 7 - Echo sounder trace in area B. Upper record from hull mounted echo sounder, lower trace (middle of lower portion) is record from deep towed sounder. Fig. 8 - Side-looking sonar traverse across canyon in area D 78 Speiss, Mudie and Lowenstein SIDE-LOOKING SONAR NEL AREA MARCH 12, 1967 =? Gal eal aN R/V THOMAS WASHINGTON : iy COSS, TAL KY ian © ; } “s. uy, "i i & _ 3200 METERS] I Fig. 9 - Side-looking sonar record on mountainside in area A | . Fig. 10 - Side-looking sonar record in area Bacross location of sounding re- cords of Fig. 7 19 Speiss, Mudie and Lowenstein 100 150 200 250 300 METERS 60 80 100 ISO FATHOMS M w = = 5 a x= WwW & = z 400 200 100-200 ) Fig. 11 - Side-looking sonar record over low-relief outcrop in area D 80 DIRECT MEASUREMENT OF BOTTOM SLOPE SEDIMENT SOUND VELOCITY AND ATTENUATION AND SEDIMENT SHEAR STRENGTH FROM DEEPSTAR 4000 Edwin C. Buffington, Edwin L. Hamilton, David G. Moore U. S. Navy Electronics Laboratory San Diego, California GENERAL Among the capabilities of deep submersibles which are especially useful, but poorly known, are those permitting types of measurements which set up, or complement, remote data customarily taken from sur- face ships. So much emphasis in the employment of submersibles has been toward those things which the submersible can do uniquely that frequently overlooked or, at least underpublicized, are those contri- butions which upgrade, calibrate, or give added significance to those data taken by wire-lowered or acoustic sensors. It is the purpose of this paper to briefly describe and discuss the significance of three such types of measurements which have been made with a high degree of success from the DEEPSTAR 4000. DIRECT MEASUREMENT OF BOTTOM SLOPE The sources of error inherent in echo sounding have been identi- fied almost from the beginning, but their correction has not always been a simple matter. The two principal causes are the speed of sound in water, and the shape or character of the sound cone produced by the echo-sounder transducer. While the first cause applies at all times, and is particularly significant in deep water, whether over a flat or rugged bottom, the second applies only on slopes. And the steeper the slope and the shallower the water, the more complicated and sizable the error. Only the second cause will be discussed in this paper. The determination of a true slope may be approached by utiliz- ing narrow-beam echo sounders, by taking lead line soundings, or by attempting a correction of the errors inherent in a slope measured with a wide-beam echo sounder. To date, the determination of these errors has been attacked by assuming given, apical angles for the 81 Buffington, Hamilton, and Moore sound cones from any given echo-sounder transducer. These angles are usually taken from a standard transducer calibration curve. Simple geometry has been developed and graphic solutions found. Acknowledging that the solutions almost certainly improve the situa- tion, there is still no way to check the absolute error without empiri- cal data on the slopes being measured. This empirical data can be provided by lead-line sounding, but inherent errors in this method give a very low level of confidence. This leaves only "in situ'' mea- surements, made directly on the spot with submersibles, as the best, and perhaps the only, basis for absolute comparison or calibration. One might expect that the measurement of a bottom slope from a submersible to be a simple matter, and indeed it is if one is content with accuracies which vary 2 or 3 degrees. Simple gravity- dependent instruments, such as pendulum or bubble inclinometers can be instal- led within the pressure-proof sphere in which the pilot and the obser- vers ride. From these, inclinations can be measured to fractions of degrees. However these instruments still permit only an approxi- mate measurement of the sea-floor slope on which a submersible rests because the pilot can only visually estimate that the horizontal plane of symmetry of the submersible, to which the interior inclin- ometers are adjusted and referred, is parallel to the plane of the sea floor. In the past three years, slope measurements have been made in 16 dives with the Cousteau Diving Saucer and the Westinghouse DEE P- STAR 4000. All dives were made where the sea floor had rugged relief. The largest concentration of dives was made off the town of San Clemente, California, where specific features had been echo sounded under good navigation control, and a large number of slopes calculated from echograms. These calculated slopes ranged from 7.5° to 31. 0° on the simple basis of a tangent resolved from depth differential, ship's speed, and time. Inthe same area, the same slopes, directly measured, ranged from 7.5 to a maximum of 43° The accuracy of these latter measurements was circumscribed a the limitation mentioned previously; namely, the pilot's ability to estimate that the horizontal plane of the submersible was parallel to the sea floor. Even so, the difference was most striking, and under- lined immediately the need for precise measurements and carefully controlled experiments. A device for permitting a submersible to measure accurate slopes has been developed at the U. S. Navy Electronics Laboratory (NEL), and tested with DEEPSTAR 4000. It has proved successful enough so that a series of dives, designed specifically for direct and precise slope measurement and comparison with echo-sounding records under rigorously controlled conditions, is planned for the coming dive ser- ies which will get underway when DEEPSTAR is recertified. The device is basically optical, and permits the pilot of a sub- mersible to adjust its attitude around both a fore-and-aft and a verti- cal axis while heading upslope and watching two light spots projected 82 Buffington, Hamilton, and Moore on the sea floor from outside the sphere. When the two spots coin- cide at a maximum measured fore-and-aft inclination with the thwart- ships axis horizontal, the horizontal plane of symmetry of the sub- mersible is exactly parallel to the sea floor and an interior reading can be made which measures, exactly, the slope of the sea floor. The two projectors are small, simple, and fortunately, not subject to refraction problems. The details are the subject of a technical memorandum currently being published by NEL. Fig. 1 illustrates the basic geometry involved in the installation aboard DEEPSTAR 4000. The forward of the two light projectors is mounted at maximum distance from the sphere, on the fore-and-aft axis, and is adjusted so that it projects its spot vertically downward or normal to the basic plane of horizontal symmetry of the submer- sible. The second projector is also mounted on the fore-and-aft axis, but next to the sphere. It is adjusted so that its beam shines oblique- ly, and its spot coincides with the spot projected by the vertical pro- jector on a plane which is tangent to the bottom of the sphere and, at the same time, parallel to the horizontal plane of symmetry of the submersible. OBSERVERS SPHERE FORE & AFT AXIS THROUGH HORIZONTAL PLANE SEA FLOOR OR CALIBRATION PLANE VERTICAL AXIS NORMAL TO HORIZONTAL PLANE Fig. 1 - Lateral Aspect. Geometry for initial mounting and adjustment of inclinometers for measurement of a level sea floor. The horizontal plane of the submersible is parallel to the plane of the sea floor (or plane tangent to the sphere base) when the light spots coincide. This situation obtains regardless of the slope of the sea floor. The only change is angular rotation of the entire relationship with regard to ab- solute horizontality. 83 Buffington, Hamilton, and Moore This particular geometry obtains at all times when the submer- sible is parallel to the sea floor and an accurate slope is being mea- sured. If the submersible is down by the stern, the light will separ- ate as shown in Fig. 2. —| <— 7° ! I. a SEA FLOOR Fig. 2 - Lateral Aspect. The plane of the submersible is inclined to the sea floor stern down. Resting on its “haunches,” the maxi- mum absolute separation of the lights is about 7°. This situation obtains regardless of the sea floor slope. The only change is angular rotation of the entire relationship with regard to absolute horizontality. Correspondingly, if the bow is down, the lights again separate, (Fig. 3) reversing their proximity to the observer's window. Close study of the geometrical relations between the submersible and the sea floor in various situations will show that to guarantee par- allelism between the horizontal plane of the submersible and the hor- izontal plane of the sea floor, it is necessary to control the thwart- ships axis running through the center of the sphere in the horizontal plane. This control can be achieved independently of any external observation or reference by simply observing the pendulum or inclin- ometer inside the sphere which is adjusted to measure deviations from an axis running thwartships. In Fig. 4 we have an anomalous situation where the two spots coincide on the sea floor and the thwartship axis is horizontal. Yet 84 Buffington, Hamilton, and Moore the horizontal plane of the submersible and the plane of the sea floor are not even close to being parallel, and the angle being measured is meaningless. SEA FLOOR Fig. 3 - Lateral Aspect. The plane of the submersible is inclined to the sea floor, bow down. The situation obtains regardless of the absolute slope of the sea floor. The remedy in this case (Fig. 4) is procedural. A standard part of the slope measuring procedure must include the rotation of the sub- mersible around an axis normal to its basic horizontal plane of sym- metry, while holding the thwartships axis horizontal with the two spots coinciding on the sea floor. Thus, when the submersible is facing upslope, the measured slope will decrease in both a port and starboard direction away from the azimuth of the maximum. The maximum is the true slope. 85 Buffington, Hamilton, and Moore *SSO[SUTUROU ST JUDUTOINSeOUW OUT, * LOOT eos oY} JOU ST YOTYM jJo suo ‘sjzutod om} Aq poUTfep st yoTyM odofs & sUIZIZOJIeIeYO poin -seoul st ojsue juorzedde uy ‘(q7) eptoutos s3zdos 4ystT [UN Aouasuey Jo yuIOd punowze sutjejor ‘moq 94} Aq possoidep st aftqrszrourqng *(eF) WI0}}j0q ey} 9AOGe apToUTOD sqystIT {T@oTJLOA OY} WIOIZ pooeTdstp ,o¢ st Adusduey Jo jJUTOgG ‘“TeJUOZTIOY stxe sdiysyremyy YIM LIOOTJ CAS OY} UO S}S9I 9TqIstouIqng *joodsy Te19je7T pue (uo-peosy) Mog - F ‘31q “q a) Ld4dSV 1V¥ALVI Ld4adSV NO -GV4H YSLAWONIIONI NO daynsvaw ONIFd JIONV NOILVLOY JO SIXV GNV ADNIONVIL JO LNIOd adO1S Andel SE S1OdS ADNIONVLY | ye NaGIDNIOD 4O LO LSS 86 Buffington, Hamilton, and Moore MEASUREMENT OF SOUND VELOCITY AND ATTENUATION Knowledge of sound velocity and attenuation in sea-floor sedi- ments is now considered to be a basic requirement for experimental and theoretical studies in bottom-bounce, underwater acoustics. At the Navy Electronics Laboratory, since 1954, measurements have been made in the laboratory, andin situ by scuba divers, and from the bathyscaphe TRIESTE. The present program in the DEEPSTAR is a continuation of previous work. These in situ measurements are necessary to validate laboratory studies on sediments sampled from surface vessels, and to determine the corrections that must be made on laboratory measurements to obtain in situ values. In addition, it is necessary to measure sound velocity and attenuation in natural, undisturbed sediments in which the original sediment structural strength is present. Successful measurements of sound velocity and attenuation were made during five dives in the present DEEPSTAR program. These Stations, made from the submersible, were supplemented by 12 sta- tions made from a diving boat (using the same equipment) with scuba divers to emplace the equipment and take sediment samples. The equipment used for the measurements was a second model of the probes used during the TRIESTE program. The new probes allow measurements at 14, 7, and 3.5 kc/s. The equipment con- sists of three probes, 22 inches in diameter, fastened to a rigid frame in such a manner that when the frame is on the sediment sur- face, the probes are inserted to a variable, preset depth in the sedi- ment; this depth varied from 6 inches to about 2 feet during the DEEPSTAR tests. The sediments were cored by plastic tubes attached to each end of the probe frame (Fig. 5). Three barium titanate transducers, designed by NEL's Trans- ducer Division, were used as sound source and receivers. Velocity was determined by measuring travel time over a 1-meter path length between receivers 1 and 2. Attenuation was measured relative to that in the bottom water (assumed to be zero for 1-meter). Path length between receivers was calibrated by using a water velocimeter. The unique feature of these probes is the ability to measure sound velocity and attenuation at three frequencies (14, 7, and 3.5 ke/s) without removing the probes from the sediment. The large diameter of the probes (22 inches) was necessary to obtain measurements at lower frequencies. Insertion of the probes into the sediment to about 2 feet is necessary to obtain measurements at 3.5 ke/s, and about 1 foot for measurements at 7 kc/s. Ina sedi- ment containing appreciable amounts of sand, the DEE PSTAR could not insert the probes a sufficient depth to obtain measurements at any frequency, and even in high porosity clayey silt the probes could not be inserted a sufficient distance to measure at 3.5 kc/s. However, 87 Buffington, Hamilton, and Moore successful measurements were made from DEEPSTAR at 14 and 7 ke/s, and at all three frequencies from the diving boat in water depths to 100 feet. Fig. 5 - Velocity probes and core tubes mounted on the brow of DEEPSTAR 4000 Measurements of sound velocity were made in sediments hav- ing ratios of sound velocity in sediment/ velocity in bottom water from 0.979 to 1.212. Sediments having sound velocities less than that in the bottom water were measured at several stations. The techniques of relating laboratory measurements of sound velocity to in situ values, determined during the TRIESTE program, were veri- fied. There were no measurements indicating a velocity dispersion (dependence of velocity on frequency). The attenuation measurements will be completed when the DEEPSTAR returns to NEL, and will be reported in a future paper. 88 Buffington, Hamilton, and Moore DIRECT MEASUREMENT OF SEA-FLOOR SHEAR STRENGTH The response of sea-floor sediment to imposed loads and struc- tures is becoming increasingly important to the U. S. Navy and other organizations considering sea-floor installations. The measurement of shear strength is, therefore, of special concern. This can be done from remotely collected samples if they are undisturbed and if the validity of the measurement can be established. To this end a carefully controlled experiment, to compare in situ and laboratory vane-shear strengths of marine sediment, was carried out in October 1966 from DEEPSTAR 4000. The tests util- ized a NEL-developed vane-shear strength machine, designed for operation from deep diving vehicles. This machine was mounted as a unit including two adjacent core tubes (Fig. 6) for simultaneous collection of in situ strength data and sediment for later laboratory study. The experiment was accomplished off San Diego, California at a water depth of 365 meters in an olive gray, clayey silt. Fig. 6 - Vane-shear machine adapted for use with DEEP- STAR 4000. The white strip below the housing protecting the vane in the center is 12 inches long. 89 Buffington, Hamilton, and Moore The vane-shear machine, mounted rigidly on the brow of DEE PSTAR in a frame with the two thin-walled tube samplers, operates at ambient pressures with a signal and power lead through to the pressure hull. The machine features a logarithmic accumula- tor which supplies torque to the vane and allows good sensitivity at a wide range of strengths. The data read-out is in electrical resist- ance calibrated to torque. For this test, the machine was set up to measure strength in the interval 20 to 22.5 cm beneath the sediment surface with a 2.5 X 2.5 cm vane. The tube samplers were 30-cm long with an inside diameter of 6.67 cm and were fixed at a distance of 26 cm on both sides of the vane. Thus, as the vane was inserted into the sediment for the in situ strength test the sample tubes also penetrated the sea floor and were filled. The samples were carefully handled and stored under sea water in their collecting tubes until the laboratory measurements were made. The measurement in the sea floor gave a shear strength of 22.13 gm/cm? and the adjacent samples tested at exactly the same depth intervals within the sediment showed strengths of 21. 81 and 27.91 gm/ cm? when later measured with a calibrated laboratory vane-shear strength machine. It is apparent that, in this test, the variation in sediment strength between the two cores taken 56 cm apart is significantly greater than the variation between in situ and laboratory measurement of shear strength. With this limited data, the tentative conclusion is that laboratory measurements of vane- shear strength are valid on gravity type cores at least 20-cm long, assuming a well-designed core tube and proper handling between collection and testing. For this type of comparative test, it is imperative that labora- tory measurements of shear strength be made on cores collected as closely adjacent to the in situ measurement as is practical, and at exactly the depth interval in the core which corresponds to the in situ depth measurement. Without this precaution the comparisons may be meaningless as lateral variations and, particularly range in strength with depth may be great. The laboratory vane strengths of these sediments, for example, varied from 5. 79 em/ em4, at 3to 5.5 cm beneath the sediment surface, to the previously stated values at 20 to 22.5 em which are about four times as great. 90 VARIABILITY IN DERIVED SEDIMENT SOUND VELOCITY AS A FUNCTION OF CORE ANALYSIS AND ITS EFFECT IN DETERMINING THEORETICAL BOTTOM REFLECTIVITY James J. Gallagher U. S. Navy Underwater Sound Laboratory Fort Trumbull, New London, Connecticut INTRODUCTION Acoustic field tests have been conducted over limited portions of flat-bottomed ocean areas. Core samples were collected from these insonified bottom areas, where each area was usually about two square miles. A multilayered, absorbing, mathematical model is being used to transform the geological core data into an acoustic description of the unconsolidated ocean bottom.'> This description is supposed to provide an understanding of the influences that a nonuniform bottom exerts on an acoustic signal as it travels through that bottom. The results of the model investigations are being compared with corre- sponding 3.7 kHz field data to determine whether the mechanisms affecting acoustic behavior are, in truth, being explained by the model. Relationships between geologic and acoustic properties of marine sediments have been studied by other investigations.°'” Inasmuch as the results of these studies will not be reviewed here, it should be noted that the products of these investigations have direct application to the mathematical model studies. The salient features of the multilayered model as currently programmed are (1) it treats up to 10 liquid layers overlying a solid layer; (2) it considers acous- tic absorption; and (3) it requires a smooth, nonsloping surface. This model is being employed as a research tool. In its present capacity disagreement between observed and predicted reflection coefficients of about 0.05, which may correspond to a change in bottom loss of several db, is con- sidered significant. The sensitivity: of this model to changes in sound velocity and layering thickness is of concern and is being explored at the present time. This report discusses the results of a preliminary investigation that was designed on Gallagher to test the effects of changing sound velocity values on the values of acoustic reflection coefficients obtained over a broad range of incident angles. DISCUSSION Two dissimilar sediment depositional areas were considered. The sediments in the’slow depositional area, the Hatteras Abyssal Plain (HAP), are character- ized by fine grained particles, low carbonate content, and high porosity. Sediments in the rapid depositional area, the Tongue of the Ocean (TOTO), are characterized by medium sized particles, high carbonate content, and moderate porosity. Two cores from each area were investigated. The distance between the two cores in both the HAP and TOTO areas is 0.4 mile. Variations in sound velocity are provided by values computed from Sutton's regression equation and from direct laboratory measurements on the core samples. HATTERAS ABYSSAL PLAIN (HAP) These variations are strikingly evident in core no. l,as seen in Table 1. The measured velocities for core no. 1 should probably be viewed with caution since they are considerably higher than what the accepted sound velocity versus porosity curves indicate; the porosities lie in the 70 percent range. It is possible that a constant error existed in the laboratory measurements. However, the Table 1 MEASURED AND SUTTON DERIVED SOUND VELOCITIES AND CORRESPONDING ACOUSTIC IMPEDANCE DATA — CORE NO. 1 —HAP Deih Sound Velocity Direct pads Impedance Direct eee) ft/sec Minus Difference x10* Ib/ft?-sec Minus Direct Sutton Sutton Direct Sutton Sutton 0-1 5266 4530 +736 13 47.00 40.85 +6.15 1-2 5181 4359 +822 15 47.80 40.67 +7,13 2-3 5336 4492 +844 15 47.60 40.47 +6,59 3-4 5266 4896 +370 6 49.30 46.26 +3.04 4-5 5266 5826 -555 11 75.09 5-6 5336 4756 +580 | DOO 46ers +5,47 6-7 5356 4369 +987 17 46.90 38.66 +8, 24 7-8 5246 4456 +790 15 47.80 40.28 GH sO2 8-9 5236 4458 +878 16 47.80 40.34 +7,46 9-10 5356 4405 +95] 18 47.80 39.73 +8.07 92 Gallagher measured sound velocity values for core no. 2, and for core no. 3, which is not included in this report, were determined with the same equipment, and both were lower. It is interesting to note that the laboratory vane shear strength values for core no. | were an order of magnitude higher than those determined from cores 2 and 3. During the core sampling operation it was noted that the bottom was quite hard in this general area. Core no. 1, in DEPTH (FEET) n > oo 10 Z(x104) 42 44 46 48 50 52 56 38 40 42 44 46 48 SO 52 Hees om o aur ae Tae [oem oan] i 75.09 oo URI DIRECT SUTTON IMPEDANCE 104 LB/FT?-SEC HAP CORE NO. 1 Fig. 1 Vertical Acoustic Impedance Profiles — Core No. 1—HAP particular, seemed so compact that if appeared to be almost dry when it was removed from the barrel at sea. However, the porosities did lie in the 70 per- cent range. The 3.7 kHz acoustic field data substantiate the nonuniformity URI DIRECT JP Sass SUTTON PEAK AMPLITUDE REFLECTION COEFFICIENT | 20 30 40 50 60 ANGLE OF INCIDENCE (DEGREES) 0 10 Fig. 2. Reflection Coefficients versus Incident Angles — Core No. 1—HAP of this bottom. Figure 1 denotes the impedance profiles for the directly measured and Sutton derived data for core no. |. The number and thicknesses of the acoustic impedance layers are not altered by the sound velocity differ- ences, but differences in the magni- tudes of these layers are produced. The resulting differences in the peak amplitude reflection coefficients over increasing incident angles for core no.1 areshown in Fig. 2. Considering the requirements imposed on this model study, the differences in bottom loss are considered significant. This is generally true for the succeeding data to be discussed. The measured and Sutton derived sound velocities and impedance data are indicated in Table 2. Graphical differences in the impedance profiles are shown in Fig. 3. For core no. 2, the differences in the sound velocity values obtained from the two cases of interest produced differences in both the layering and the magnitudes of the impedance layers. The reflec- tion coefficients computed from the measured and Sutton derived data for 93 DEPTH (FEET) Z(x104) 42 44 46 48 50 52 54 38 40 42 44 46 48 50 Toa Talay (aaa @ oc fF YN SO cS ae Ss ara 65.4 L 75.55 URI DIRECT IMPEDANCE x10‘ LB/FT?-SEC HAP CORE NO. 2 =*== |NFERRED CURVE (NO DATA AVAILABLE) SUTTON Fig. 3 Vertical Acoustic Impedance Profiles — Core No. 2—HAP Gallagher Table 2 MEASURED AND SUTTON DERIVED SOUND VELOCITIES AND CORRESPONDING ACOUSTIC IMPEDANCE DATA — CORE NO, 2—HAP Benin Sound Velocity Direct ese ft/sec Minus Direct Sutton Sutton 0-1 5026 4597 +429 1-2 5041 4564 +477 2-3 5046 -- -- 3-4 5036 4564 +472 4-5 5031 4464 +567 == 4413 +618 5-6 5026 4403 +623 6-7 5016 4310 +706 7-8 5006 4900 +106 8-9 5041 4565 +476 9-10 4926 4454 +472 10-11 4996 4633 +363 11=12 4956 4494 +462 12-13 5041 4592 +449 13-14 4881 4454 +427 14-15 4876 4496 +380 15-16 4906 4535 +372 16-17 4876 4589 +287 17-18 5066 5857 -791 18-19 4846 4457 +389 19-20 5026 4494 +552 20-21 == == == ye ae Impedance Direct Difference «104 Ib/ft?-sec Minus Direct Sutton Sutton 9 Ae /Omm ales h69 10 44.30 40.16 +4,2 -- 47.30 -- -- 10 ANS) S10) S)) -(0)7/ +4,2 11 45.30 40.76 +4.6 12 -- 39.14 AHO)5 22 12 ANAL KO) GS), 227/ +5.6 14 44.10 37.84 +6.3 72 48.50 47.43 Pils I 9 46.60 42.22 +4,4 10 45.80 40.71 tro II 7/ AG.90) A345 areo5) 9 46.40 42.06 +4,3 g A520 SA, +4,0 9 AN 8O) SI 2 $330 7/ 8 42.90 38.44 +4,5 8 44.20 41.04 tro Z 6 42.90 40.38 58) 16 OSL0 75.55 -10.1 8 44.30 40.73 EOnO 11 51.00 48.46 P25) U { 1 1 aS ie) aS Oo 1 i core no. 2 are shown in Fig. 4. The bottom loss values appear to be greater in core no. 2 for the two cases than tempting to attribute the lower losses in core no. | to the higher sound ve- locities reported, it appears that the magnitude of the sound velocity value has little influence on the value of the reflection coefficients for any given angle. For cores 1 and 2 the lower Sutton derived sound velocity data appear to produce generally higher reflec- tion coefficients. However, this is not for those in core no. 1. Although it is URI DIRECT ----- SUTTON , LOE all os. Ties eel oe Nic oneal | i o S| T L PEAK AMPLITUDE REFLECTION COEFFICIENT | 1 10 20 30 40 50 60 70 80 90 ANGLE OF INCIDENCE (DEGREES) Fig. 4 Reflection Coefficients versus Incident Angles — Core No. 2—HAP we Gallagher Table 3 MAXIMUM DIFFERENCES IN REFLECTION COEFFICIENT VALUES CORES NO. 1 AND NO. 2—HAP Sait Incident Direct Sart Bottom Loss Difference oe Angle Measurement EA. Direct Minus Sutton Core No. 1 aoe 0.17 (15.0 db) 0.45 (7.0 db) +8 .0 db 48° 0.20 (14.0 db) 0.54 (5.4 db) +8.6 db 58° 0.14(17.0 db) 0.51 (5.7 db) +11.3 db Core No. 2 VA> 0.06 (24.5 db) 0.39 (8.1 db) +16.4 db ES 0.09 (21.0 db) 0.54 (5.4 db) +15.6 db consistently true over all angles. The impedance values of the layers, the thicknesses of the layers, and the angles and speeds at which the sound rays are propagating through the layers apparently combine to form constructive and destructive interferences and thereby regulate the amount of acoustic energy returned to the interface. The maximum differences in the values of these reflection coefficients, and of bottom loss, derived from the measured and Sutton data, are shown in Table 3. It is interesting to note that whereas the magnitude of the sound velocity differ- ences between the measured and Sutton data in core no. 2 is somewhat less than that in core no. 1, larger differences in bottom loss are seen in core no. 2. These large differences may occur aperiodically over various angles, depending upon the layering effects of the model. It should be noted that the differences are smaller at other angles and at some angles there is a zero difference. It should also be noted that since the bottom loss differences are logarithmic ratios, a change in one unit of reflection coefficient at the lower range will produce a different corresponding change in db than will a change in one unit of reflection coefficient at the upper range of the scale. Table 4 indicates the differences in bottom loss between the measured and Sutton derived data in cores 1 and 2 and the data obtained from field acoustic tests at a frequency of 3.7kHz. Considering the suspected high measured velocity values for core no. 1, it is interesting to note that the field data report generally lower bottom losses for most incident angles. The exceptions are at 42° and at angles approaching normal incidence. Ina general sense, it would be expected that even higher sound velocities would be required to pro- duce higher impedance values in order to obtain the lower bottom losses reported by the field test data. However, in noting the previous remarks, it can be seen that this will not necessarily follow. Generally, the 3.7 kHz field test data exhibit lower bottom losses than those for both cases for core no. 2. Je) Gallagher Table 4 VARIATIONS IN BOTTOM LOSS BETWEEN FIELD DATA AND DIRECT AND SUTTON DERIVED VALUES—CORES NO. 1 AND 2—HAP Core No. 1 Core No. 2 Incident Field Field Minus Field Minus Field Minus Field Minus Angle Values Direct Sutton Direct Sutton 42° 8.0 db +2.5 db +6.7 db 0 -7.8 db 50° 9.1 db -1.2 db +3.6 db =5.8 db -2.6 db SEY 5.0 db -0.3 db -0.2 db -8.6 db -3.0 db 57° 6.0 db -4.4 db +0 .3 db -10.5 db -5.0 db 60° 4.4 db -4.7 db -1.6 db -21.6 db -4.4 db 62° 1.8 db -3.9 db -4.4 db -14.0 db -7 .0 db 65° 3.5 db -2.6 db -2.0 db -16.5 db -8 .5 db 68° 1.8 db -9.9 db -3.6 db -17.3 db =6.2 db 70° 3.1 db -3.8 db -3.3 db -10.9 db -7.9 db 75° 0.8 db -4.4 db -6.7 db -14.1 db -13.2 db 80° 3.1 db +1.8 db -1.0 db =-21.3 db -2.5 db TONGUE OF THE OCEAN (TOTO) In contrast to the large differences noted between the Sutton derived and directly measured sound velocity values in the Hatteras Abyssal Plain cores, the differences between these values in TOTO core no. 5 are small. These data, which are shown in Table 5, present an bia interesting example of the caution that 46, #8 50 52 54 56 58 60 52 54 56 58 60 62 64 6 should be exercised in using this model 2 for predicting in situ bottom losses. Small differences in sound velocity be- tween in situ values and those deter- mined from direct core measurements could be produced by mechanical dis- turbances in the sampling and handling processes. These differences would be — produced by the alteration of the sedi- TOO CORNOS ment aggregate structure. It may be === INFERRED CURVE (NO DATA AVAILABLE) assumed that these differences would be af IL J DEPTH (FEET) SS Oo e fs © N 3 @ 1) SSS aaa r re ae 4 fel i H Hq H H N N NAVOCEANO DIRECT SUTTON ° . Fig. 5 of the same magnitude as the differences Vertical Acoustic Impedance Profiles — reported for core no. 5 in Table 5. Core No. 5— TOTO i The acoustic impedance profiles constructed from the data from these two methods are shown in Fig. 5. Changes in the magnitudes of the impedance layers 96 Gallagher Table 5 MEASURED AND SUTTON DERIVED SOUND VELOCITIES AND CORRESPONDING IMPEDANCE DATA — CORE NO, 5—TOTO Depth Sound Velocity Direct pause Impedance Direct ierceval ft/sec Minus Difference x 10° |b/ft?-sec Minus Direct Sutton Sutton Direct Sutton Sutton 0-1 495] 4970 -19 0.4 SHleZ iil 5 740) -0.2 1-2 5078 4920 +158 So I O22 8 ole 20 talsow, 2-3 5103 5030 +73 1.4 Soya lleh OB GO) +0.83 3-4 5131 5140 -9 -- 70.80 70.90 -0.10 4-5 S27, 5200 -73 1.4 B58) BY/740) -0.81 5-6 4997 5020 -23 0.46 ACEO a0 20, -0.23 6-7 -- 4840 -- -- 49.97 48.40 lov, 7-8 4991 4720 +271 5.4 50.91 48.10 +2,.81 8-9 -- 4750 -- -- 50.91 48.40 2a 9-10 5080 -- -- -- 54.35 -- -- 10-11 5110 5010 +100 JS S472 Ss _ 010) PlloIZ2 11-12 5104 4820 +284 S55) TO Ola) +3.0 12-13 5140 4970 +170 Soc) 559 S46 70) +1.85 13-14 5116 4900 +216 4.2 {8,723 E3570) +2.38 14-15 5169 4870 +299 Sais) 56.86 53.60 +3,26 15-16 5142 5070 +72 1.4 J335 i) B5680) +2.31 and in the number and thicknesses of the layers were effected by the small sound velocity differences. The peak amplitude reflection coefficients for both the measured and Sutton derived values over a wide range of incident angles are shown in Fig. 6. Differences in the measured and Sutton derived sound velocity values are again large in core no. 6, as shown in Table 6. The resulting differences in the respective impedance profiles are shown in Fig. 7. These moderate differences in the sound velocity data produced changes in the number and thicknesses of impedance layers and in the magnitudes of these layers. N iS o oo = PEAK AMPLITUDE REFLECTION COEFFICIENT fo} NAVOCEANO DIRECT 9 nem ----- SUTTON , | 1 it It it ! 1 ! J 0 10 20 30 40 50 60 70 80 90 ANGLE OF INCIDENCE (DEGREES) Fig. 6 Reflection Coefficients versus Incident Angles — Core No. 5— TOTO The reflection coefficients for both cases in core no. 6 were computed for various incident angles, and the results are shown in Fig. 8. The magnitude ih Gallagher Z(x104) por 48 50 52 54 56 58 60 20m 32 13 56 58 60 62 a) aaa NAVOCEANO DIRECT oe -- o---- SUTTON ” oa oo S&S © Te oT Ta DEPTH (FEET) Ss SS T T 44 —_ PEAK AMPLITUDE REFLECTION COEFFICIENT NAVOCEANO DIRECT ‘i SUTTON (0) i J | IMPEDANCE 104 LB/FT?-SEC 0 10 20 30 40 50 60 70 80 90 | TOTO CORE NO. 6 ANGLE OF INCIDENCE (DEGREES) | = === INFERRED CURVE (NO DATA AVAILABLE) Fig. 7 Fig. 8 Vertical Acoustic Impedance Profiles— Reflection Coefficients versus Incident Core No. 6— TOTO Angles — Core No. 6— TOTO Table 6 MEASURED AND SUTTON DERIVED SOUND VELOCITIES AND CORRESPONDING ACOUSTIC IMPEDANCE DATA — CORE NO. 6—TOTO Death a Velocity Direct pateees Impedance Direct nape t/sec Minus Difeenee 10° |b/ft*-sec Minus Direct Sutton Sutton Direct Sutton Sutton 0-1 4923 5150 -227 5 48.34 50.05 -1.71 1-2 4925 -- -- -- 50.62 -- -- 2-3 5006 5540 -534 11 DAO ORO = D473 3-4 5344 5580 -236 4 58.27 60.80 -2.55 4-5 5029 == -- -- 51.45 -- -- 5-6 4954 5380 -426 9 S20) S780 -8.59 6-7 5015 5410 -395 8 SIP 5,00 -4,09 7-8 4949 -- -- -- -- -- -- 8-9 -- -- -- -- -- -- -- 9-10 5123 5680 -557 11 S/AK3 Bi, 70) =One2, 10-11 5139 -- -- -- 56.78 -- -- 11-12 5085 5710 -625 12 56.09 63.10 -7.01 12-13 5082 5540 -458 9 S235 57/0 -4.75 13-14 5097 5340 -253 5 SMS A. 10 -2.9) 14-15 5114 5810 -696 14 -- 65.70 -- of the differences in the reflection coefficient curves between the two methods for cores 5 and 6 does not necessarily reflect the respective ranges of sound velocity differences in these cores. The differences between the reflection coef- ficients computed from the measured and Sutton derived data for core no. 5, where the sound velocity differences are very small, are similar to those differences in 98 Gallagher core no. 6, where the sound velocity differences are quite large. The maximum differences between the measured and Sutton derived bottom losses, occurring at various incident angles, for both cases are shown in Table 7. Table 8 indicates the reflection coefficient values obtained from field tests at a frequency of 3.7 kHz and denotes the variations in bottom loss obtained from the core data relative to the field values. The bottom loss values obtained from the field tests are consistently greater than those obtained from the use of either method on the core samples. Table 7 MAXIMUM DIFFERENCES IN REFLECTION COEFFICIENT VALUES CORES NO. 5 AND NO. 6—TOTO Suen Incident Direct Gan Bottom Loss Difference Angle Measurement Sue Direct Minus Sutton Core No. 5 Ie 0.22 (13.2 db) 0.47 ( 6.6 db) +6.6 db 42° 0.27 (11.4db) 0.46 ( 6.7 db) +4.7 db 54° 0.26 (11.7 db) 0.54 ( 5.4 db) +6.,3 db 61° 0.27 (11.4 db) 0.74 ( 2.6 db) +8 .8 db 70° 0.25 (12.0 db) 0.95 (0.5 db) +11.5 db Core No. 6 20° 0.365 ( 8.7 db) 0.23 (12.8 db) -4.1 db 46° 0.285 (10.9 db) 0.515 ( 5.7 db) +5.2 db 56° 0.04 (28.0 db) 0.29 (10.5 db) +18.2 db 62° On57 (43 cla) O32 (( e7 Gla) +3.1 db Ow 0.25 (12,0 cls) Oss8 ( 0.5 cls) +11.5 db Table 8 VARIATIONS IN BOTTOM LOSS BETWEEN FIELD DATA AND DIRECT AND SUTTON DERIVED VALUES—CORES NO. 5 AND 6—TOTO Core No. 5 Core No. 6 Incident Field Field Minus _—_ Field Minus Field Minus _—_ Field Minus Angle Values Direct Sutton Direct Sutton 50° 15.4 db +4.7 db +14.7 db +6.6 db +0.9 db 552 17.6 db +5 .9 db +11.1 db =5.4 db +6.2 db 60° 17.1 db +5 .9 db +14.0 db +98 db +12.9 db 65° 14.5 db +2.5 db +13.3 db +0.5 db +10.5 db Ze 15.9 db +4.9 db 115)5) cls) +12.3 db +14,6 db US? 14.0 db +3.9 db +13.8 db +7.1 db +13.2 db 78° 15.9 db +7.1 db +15.7 db +10.5 db +15.3 db 99 Gallagher SUMMARY AND CONCLUSIONS The purpose of this single experiment was to test the multilayered, absorbing, mathematical model for sensitivity to changes in sound velocity. Two closely spaced cores from each of two dissimilar depositional environments were investigated. The slow depositional environment is represented by the Hatteras Abyssal Plain cores, andthe rapid depositional area, by the Tongue of the Ocean cores. The sound velocity data used were values determined from direct meas- urements on the core samples and computed values obtained with Sutton's regression equation for the same cores. Based on this investigation, it is concluded that for given density values higher sound velocity values will produce greater values of impedance, but they do not necessarily produce lower bottom losses over all incident angles. In addi- tion, for given density values, the magnitude of the differences in reflection coefficients derived from two sources over various incident angles does not necessarily reflect the magnitude of the differences between the sound velocity values of these sources. The impedance values of the layers, the thicknesses of the layers, and the angles and speeds at which the sound rays are traveling through the various layers in a nonuniform bottom apparently combine to set up constructive and destructive interferences, thereby regulating the amount of acoustic energy returning to the interface. This model is currently being employed as a research tool. In its present capacity, bottom loss differences greater than 3 db between observed and pre- dicted values will continue to be considered significant. Differences in bottom losses of 3db or greater over short lateral distances are evidenced by single source data obtained from two closely spaced core samples. If indeed this close range variability is widespread within a physiographic province, and there is every reason to believe that it is for some provinces, then the accurate predic- tion of changes in bottom loss of several db over extended distances, based on a few sample points, will be difficult. Results of acoustic measurements made at these core locations substantiate that a complex nonuniform acoustic impedance layering structure exists over small lateral distances. The effectiveness of this model in accurately predicting in situ acoustic bottom losses over core sample lengths appears limited by the sensitivity to changes in sound velocity. The effects of small changes on layer thickness and sound velocity will continue to be sought. The problem has already been mani- fested by the fact that in some cores the sensitivity of the model to changes in sound velocity produced more impedance layers within a core length than the model program was capable of accommodating. Therefore, laboratory experi- mental model studies are being planned to simulate the mathematical model to compare measured and computed reflection coefficients for given data. Attempts will be made to alter the number of original layers, while the original mass Gallagher of sediment is retained, to note the effect of integrating the layers on the reflection coefficients. 10. REFERENCES . M. C. Karamargin, ''A Treatment of Acoustic Plane Wave Reflections from an Absorbing Multilayered Liquid and Solid Bottom,'' USL Technical Memorandum No. 913-91-62, 16 July 1962, . M. C. Karamargin and B. J. Klein, ''Some Theoretical Computations of Distortions at Reflection from an Absorbing Multilayered Liquid and Solid Bottom,'' USL Technical Memorandum No. 910-169-63, 27 August 1963. F. R. Menotti, R. D. Whittaker, and S$. R. Santaniello, ''Analysis P roce- dure for the Bottom Reflectivity Measurements Program,'' USL Technical Memorandum No. 913=4-65, 26 January 1965. F. R. Menotti, S. R. Santaniello, and W. R. Schumacher, Studies of Observed and Predicted Values of Bottom Reflectivity as a Function of Incident Angle, USL Report No. 657, 28 April 1965. F. R. Menotti and S. R. Santaniello, ''Observed Values of Bottom Reflec- tivity as a Function of Incident Angle (TOTO),'' USL Technical Memoran- dum No. 913-286-65, 20 December 1965. E, L. Hamilton, G. Shumway, H. W. Menard, and C. J. Shipek, "Acoustic and Other Physical Properties of Shallow-Water Sediments Off San Diego,'' Journal of the Acoustical Society of America, vol. 28, no. 1, January 1956, pp. 1-15. G. H. Sutton, H. Berckhemer, and J. E. Nafe, ''Physical Analysis of Deep-Sea Sediments,'' Geophysics, vol. 22, no. 4, October 1957, pp. 779-812. G. Shumway, "Sound Speed and Absorption Studies of Marine Sediments by a Resonance Method —Parts I and I1,'' Geophysics, vol. 25, no. 2, April 1960, pp. 451-567; vol. 25, no. 3, June 1960, pp. 659-682. E. L. Hamilton, ''Sediment Sound Velocity Measurements Made In Situ from Bathyscaph Trieste, '' Journal of Geophysical Research, vol. 68, no. 21, 1 November 1963, pp. 5991-5998. E, L. Hamilton, ''Sound Speed and Related Physical Properties of Sedi- ments from Experimental Mohole (Guadalupe Site),'' Geophysics, vol. 30, no. 2, April 1965, pp. 257-261. 101 Strange Hot Waters and Minerals at the Bottom of the Red Sea John M. Hunt and David A. Ross Woods Hole Oceanographic Institution Woods Hole, Mass. 02543 During the fall of 1966, the R/V CHAIN conducted an extensive survey of the hot brine area of the Red Sea (21°10'N to 21°30'N). This survey included: detailed bathymetry using radar reflecting buoys; measurements of temperature with conventional hydrocasts and a temperature telemetering pinger; continuous seismic profiling; gravity and magnetic measurements; and a detailed sampling of bottom sediments with free-fall, gravity, piston and, box-coring techniques. This paper discusses some of the preliminary results of this survey. The kathymetric survey showed three deeps (Fig. 1), which contain hot highly saline water. These deeps are located in the central portion of the Red Sea rift valley. The ATLANTIS Il Deep is about 12 km long and 5 km wide, and contains several small topographic highs. This deep is connected to the smaller DISCOVERY Deep by a narrow channel; which is apparently sufficiently high, at present, to prevent mixing of the hot saline waters. The CHAIN Deep, which was discovered on this cruise, is situated in a saddle on this channel. Temperature profiles of the brine areas were obtained with a temperat- ure telemetering pinger, developed especially for this cruise by Benthos Inc. (Ross and Tyndale, 1967). This pinger emits two pings within a one second interval. The first ping is emitted every second, the second ping at a time interval after the first, this time interval is a measure of temperature. The time interval is maintained by a thermistor that produces a varying resistance which controls the telemetering electronics. Both pings were received by the ships 12 ke echo sounder and recorded on a Precision Graphic Recorder. Temperature could then be read directly from the P.G.R. In the Red Sea, a thermistor with a range of 20°C to 70°C was used. The accuracy of the measurement was +1.25°C. Determinations of absolute temperature were made by conventional methods using high-range thermometers. The temperature telemetering pinger 102 Hunt and Ross was used to define the water structure, position other equipment in certain water levels and to test other areas of the Red Sea for high temperature waters. No other areas tested contained anomalously hot water. The highest water temperature measured was 56°C. This was from below a depth of 2040 meters in the ATLANTIS Il Deep. This water had a "salinity" of about 317%. The "salinity" was determined by a Schleicher- Bradshaw salinometer on samples diluted by volume with distilled water. Salinity values so determined assume that the brine water has the same relative proportion of salts as normal sea water. This assumption is not correct. A better estimate of salinity is total solids by evaporation to dryness at 200°C, which for the 56°C water is about 255%, (F.T. Manheim, personal communication). The 56°C water of the ATLANTIS Il Deep is overlain by a layer of 44°C water (total solids about 1317). This 44°C water grades into normal 22°C Red Sea bottom water. The 44°C and 56°C water generally have oxygen values of 0.1 ml/I or less. The DISCOVERY Deep has water with a temperature of 44.7°C below a depth of 2038 m, which is overlain in some instances by a layer of 36°C water. This 44.7° water has a salinity similar to the 56°C water of the ATLANTIS II Deep. The CHAIN Deep has a maximum temperature of 34°C anda "salinity” of about 74%,. Its deepest part was not sampled. Continuous seismic reflection profiles have not been fully analyzed at present. Field observations suggest considerable rifting and faulting in and adjacent to the brine areas. Using a recording bandwidth of 37.5 to 150 Hz some reflections were obtained from the hot brine water in the ATLANTIS II and DISCOVERY Deeps (no records were made in the CHAIN Deep area). These reflections apparently are due to density differences between the hot brine water and the overlying Red Sea water. All gravity observations, bathymetry and navigation data were processed while at sea, and the results computed and plotted on line by the shipboard IBM 1710 computer. A gravity anomaly of about + 120 milligals is observed in the hot brine area. However, this value is within the range normally observed in the rift valley. A magnetic anomaly of - 650 gammas was observed over the ATLANTIS Il Deep, and an anomaly of + 350 gammas was observed over the DISCOVERY Deep. The large magnetic gradient between these areas may be due to higher sub-surface temperatures in the ATLANTIS II Deep. Temperature gradients in the sediments, as determined by heat flow measurements, were 10 to 20 times the world average of 1°C per 16 meters. 103 Hunt and Ross Fig. 1 - Bathymetry of the hot brine region in the Red Sea. Depth contours in meters corrected for sound velocity according to Matthews Tables. No correction has been applied for the increase in sound velocity in the hot brine water. 104 Hunt and Ross Extreme variations in gradients occurred along the eastern flank of the ATLANTIS Il Deep, which suggests that this area may be a likely source of the hot brines. Sediment cores obtained from both the brine and adjacent areas contained brightly colored material. The sediments generally were comprised of alternating yellow, brown, red, orange and black layers that are mainly amorphous iron oxides. Cores obtained using a square box coring device were especially impressive. These cores (4m long x 15 cm x 15 cm) collected relatively undisturbed samples. Because of its large size, this sampler collects enough material so that many detailed chemical and geological analyses can be performed on an individual layer. These analyses are now in progress. The preliminary interpretation is that the hot saline water and its associated heavy metals are ejected, probably forcibly, periodically from the ATLANTIS II Deep. The hot brines in the other deeps are the result of spillover. It is possible that similar brine pools may be found on the ocean floor in other parts of the world rift system. This work was partly supported by the Office ot Naval Research and by the National Science Foundation. (Nonr-4029 and GA-584) References Cited Ross, D.A., and Tyndale, C. A temperature telemetering pinger, in press, Geo-Marine Technology. 105 MILITARY SIGNIFICANCE OF DEEPLY SUBMERGED SEA CLIFFS AND ROCKY TERRACES ON THE CONTINENTAL SLOPE Robert F. Dill U.S. Navy Electronics Laboratory ABSTRACT A series of narrow, step-like, rock terraces and low sea cliffs (between depths of 325 and 1170 feet) have been observed on three DEE PSTAR dive-traverses up the continental slope off San Diego, California. The terraces are cut in bedrock and covered with coarse shelly sand of shallow water origin, and large rounded boulders. The flat terraces are less than 100-feet wide and the maximum observed relief of the adjacent cliffs is less than 60 feet. The best-developed cliff and terrace is between 600 and 700 feet (water depth), and extends at least 10 miles along the slope off San Diego. Although large when viewed from the window of a submersible, these features are usually overlooked on most echo-sounder surveys because of the lack of definition within the wide angle of the sound cone and the high speeds of the survey traverses. Restudy of existing echo-sounder and acoustic reflection profiles, especially those made with narrow beams, may show that these features are important world-wide top- ographic expressions. If they are related to still-stands of lowered sea-level during the Pleistocene, these terraces should be found in many areas at nearly the same depth, permitting correlation over great distances. | Acoustic reflections from the rocky cliffs and associated coarse sand and cobbles should be greatly different than those from the adja- cent slopes which are usually covered with a thick mantle of fine- grained sediment. This difference must be considered in any type of ASW or lost-instrument search and recovery employing acoustic sound sources. Also, large numbers of fish were over these rock areas and may be important sources of false echos on search sonars. In addition, these terraces could afford a firm, fairly level founda- tion for vehicles, weapons, and equipment placed on the sea floor. 106 Dill INTRODUCTION The U. S. Navy, although capable of operating throughout all the world's oceans, conducts many of its operations over the relatively shallow waters of the continental shelves and slopes immediately adja- cent to strategic land areas. The increasing reliance on sophisticated acoustic equipment for search in both ASW and Mine Hunting opera- tions along with a requirement to retrieve lost ordnance and valuable equipment from the sea floor, makes a detailed knowledge of the bottom in these operational areas necessary. In most instances fleet personnel assigned to the foregoing tasks must rely on published hydrographic charts if they want to know the nature of the bottom over which they are working. The deeply submerged rock terraces and associated sea cliffs discussed in this paper, although capable of greatly affecting sonar search capability, do not appear on these charts. Even more critical is the fact that they do not show on most echo-sounder profiles across the continental slope when made with conventional echo-sounding equipment. (Fig. 1). The existence of rocky areas in regions presumably covered with fine-grained sediment have been long known and exploited by fishermen. Abrupt changes in the bottom slopes recorded on echo- sounder profiles have led marine geologists (Emery and Terry, 1956; Emery, 1960; Terry, 1965; Goreau and Burke, 1966) and many others to propose the existence of deep terraces throughout the world. How- ever, the extent that these terraces were correlatable, their size, and the nature of the sediments associated with them were unknown until world-wide surveys could be conducted with sub-bottom acous- tic profilers (Moore, 1957; 1960; Buffington and Moore, 1963; Moore and Curray, 1963; Curray and Moore, 1964; Garrison and McMasters, 1966) and direct observations made from deep submersibles (Busby, 1965, Shepard and Dill, 1966). The profilers permitted a rapid sur- vey of the relationship between the bedrock forming the continental margin and its cover of recent sediment. The submersibles permit- ted a visual and photographic record of bottom roughness, provided bottom samples, and allowed accurate measurements of slopes and micro-relief in the vicinity of the terraces and sea cliffs. For the first time the in situ factors controlling bottom reflectivity and prob- lems associated with operating acoustic equipment in these areas could be made, and the areas could be assessed as foundations for vehicles, weapons, and equipment. Terraces and Sea Cliffs off Southern California The direct observation of deeply submerged terraces and sea cliffs were made during three DEE PSTAR dive-traverses up the continental slope approximately 6 miles off San Diego, California (Fig. 2). The traverses were made between depths of 325 and 1200 feet. Within this interval, 7 cobble-covered terraces and 2 small sea cliffs were found. The terraces are cut in bedrock and covered with coarse, shelly sand of shallow-water origin. Large rounded 107 Dill Sos 20°F Mt Knots Lend ae warr en Mh Lome, C we Fig. 1 - Typical echo-sounder profile across continental slope off San Diego, California and locale of terraces and sea cliffs (arrows). 108 Dill aie 50' VW MISSION BEACH SAN DIEGO 300 DIVE 192 312i EM 2) 117° 20' 10° Fig. 2 - Location of DEEPSTAR dives to observe deeply-submerged terrain and sea cliffs. Shaded area continental shelf. Depth in fathoms. 109 Dill boulders up to 2 feet in diameter were also observed. Exposed bed- rock has been observed down to depths of 1170 feet and up to the break in slope at 325 feet (Fig. 3). The terraces are less than 100- feet across and the relief of the adjacent cliff is less than 60 feet. The best-developed cliff is between 600 and 700 feet (water depth) and is known from dredging and echo-sounding traverses to extend con- tinuously for at least 10 miles along the continental slope off San Diego, California. Although large when viewed from the window of a submersible, the narrow terraces and rock cliffs are only barely dis- cernible on echo-sounder traces because of the lack of definition within the wide angle of most sound cones and the relatively high speed at which most surveys are run across the continental slope. In many instances it would be difficult to differentiate between irreg- ularities on the echo-sounder trace caused by high-sea surface swell and those caused by bottom irregularities. The terraces and cliffs are narrow zones of acoustically differ- ent bottom that extend for great distances along the continental ter- race. The bottom between the rocky zones are blanketed with a thick (over 10 feet) deposit of fine-grained, silty clays of uniform compo- sition. Associated Shallow Water Fossils Dredge samples at the terrace level have been analyzed by Dr. Edwin Allison of San Diego State College who states they contain a mixture of fossil and modern species. Associations of the fossil species from a ''beachrock'' sample taken at a depth of 600 feet dur- ing Dive 175 (see Fig. 2) indicate water depths no deeper than 60 feet existed when they were living and that the water was colder than exists at 60 feet in the present latitude of San Diego County. The fol- lowing species are indicative of the above environment: Olivella baetica Marrat in Sowerby, 1871 Turritella cooperi Carpenter, 1864 Acila (truncacila) castrensis (Hinds, 1843). Associated mollusks and micro-fossils from other samples taken along the continental slope off San Diego support the shallow- water origin of the fossils associated with the sea cliffs and found on the terraces. Abundant Foraminifera in the sediment collected during the DEE PSTAR dives include the following species indicative of similar shallow depths: Casidulina tortuosa Cushman and Hughs, 1925 Cassiduling limbata Cushman and Hughs, 1925 Elphidium crispum (Linnaeus, 1758). Dill E | Ww fg) As ae a REE 0 DEEPSTARDIVEN75) ts SUBMERGED BEACH LEVELS ALIR Ae i i | } | { DEPTH, METERS | | iS) ro) DEPTH, FATHOMS 2000 FT. —200 Fig. 3 - Location of sea cliffs and terraces observed on DEEPSTAR dive number 175. Note that they are not visible features on this echogram made at 12.5 Ke with a standard EDO transducer (60° sound cone) running at 4.5 knots. The sea surface was almost flat. Dill It thus appears that the sea cliffs were cut during the Pleistocene lowering of sea level. That such is the case is important because it permits the speculation that similar sea cliffs exist throughout the world and at similar depths, in those areas known to be relatively stable and unaffected by tectonic movements of the continental margin. Nature of the Sediment-Covered Slope Most of the continental slope off San Diego is covered with a deposit of fine-grained, greenish, clayey-silt. The sediment is cohesive and has shear-strength values indicating that it has built up slowly, probably as a prograding slope deposit over a considerable period of time (Moore, 1961). The benthonic Foraminifera associated with the clay are deep-water forms, characteristic of the depths in which they are found. The slope below Dive 175 has been investigated by Hamilton (1963) in the bathyscaphe Trieste I and with cores. He reported that the bot- tom was composed of a sandy silt at a depth of 188 fathoms (344 m). He states that the bottom, when viewed from the window of the bathy- scaphe, is relatively smooth and slopes gently toward the west. A 10. 5-foot long piston core taken at a depth of 187 fathoms (342 m) cored 50 cm of clayey silt overlying 250 cm of sandy silt. This was underlain by 10 cm of silt with abundant broken shell debris. Bottom Currents on the Slope Bottom currents were measured during the DEE PSTAR dives that reached velocities up to 0.4 knot. In most instances they were flowing diagonally down slope. However, in one instance in the vicin- ity of a sea cliff, currents were observed to be flowing up slope. This might have been a counter current set up by the prevailing down- slope current spilling over the upper lip of the small sea cliff. Simi- lar counter currents have been observed in submarine canyons where there are abrupt dropoffs in the axial slope. Another important observation was that scour depressions are frequent around large pieces of man-made junk that had been dropped on the sea floor. In one instance a bathroom sink approximately 2 feet in diameter had developed a scour depression of over 3 feet in diameter. It had settled at least 1- foot into the sediment and rested on alag deposit of broken shells. The constrictions, of flow stream- lines created by this foreign object caused the erosion (or lack of accumulation) of the otherwise stable fine-grained sediments that are prograding the slope. This indicates that the sediment forming and building up the continental slope are at equilibrium with the prevail- ing bottom currents. Scour. can therefore be expected around arti- facts placed on the sediment-covered slopes; an important considera- tion when planning the placement of bottomed equipment if instru- ments must remain upright or stable. Dill Small objects, such as tin cans, are covered by a deposit of fine- grained sediment, showing that although larger objects cause scour, smaller ones do not. The sediment cover shows that, in general, the continental slope at the present time is an environment of deposition. Fine-grained sediment, brought to the shoreline by rivers, form clouds of dense, dirty water that slowly spread seaward along the bot- tom as a turbid layer (Moore, 1960; Vernon, 1965). This material is kept is suspension by bottom currents and swell-induced surge as it crosses the continental shelf. When these suspended particles encounter the relatively quiet waters over the gently sloping contin- ental slope (less than 10 degrees) they are no longer kept in suspen- sion by storm surges and bottom currents, and build up as a prograd- ing sedimentary deposit (Moore and Curray, 1963). The occurrence of relatively thick deposits of sediment have been verified by sub- bottom acoustic profiles across most of the continental slopes of the world. It is then very important to note that the terraces and sea cliffs discussed herein are not covered by such sediments. They must be older than the sediment covering most of the continental slope and yet young enough not to have been covered. In most instances the fine-grained sediments are capable of being deposited on relatively steep slopes. Submersible observations in areas of sub-bottom acoustic profiles have shown stable sedimentary deposits up to ten or even hundreds of feet thick develop on slopes of over 40 degrees, a value much greater than the average continental slope. The lack of a sedimentary cover on the terrace and sea cliffs therefore indi- cates there has not been a large enough supply or sufficient time to bury these features. Youth must be the other criteria for the exist- ence of the terraces; the late Pleistocene age of associated fossils Support this contention. Significance of Terraces and Sea Cliffs Geologically speaking, we are looking at a slope that has been formed in the last ‘few minutes'' of earth time. A world-wide cor- relation of the terraces at similar depths would indicate that tectonic (mountain building) forces have not had time to warp and modify the continental edge since their formation. The occurrence of shallow- water fossils, those that lived in depths of less than 60 feet, in water ten times that depth, must indicate a greatly lowered sea level. We know that the great ice sheets that covered much of the continents during the Late Pleistocene lowered sea level; however, the deepest proposed lowering until now has been about 480 feet (Curray, 1960, 1965; Donn, et al., 1962; Shepard, 1964; Garrison and McMaster, 1966). The wide-spread, 600-foot terrace indicates a greater lowering of sea level than heretofore suspected. Large numbers of bottom fish are associated with the rocky areas. They do not venture far from the protection of the ledges and, at high frequencies, would constitute a definite, false-echo problem because of their air bladders. Large schools of an unknown Dill type of fish were observed to school above the rocky regions. The schools appeared as large clouds of reflectors (on traverses over the terrace areas) on records of an echo sounder operating at a frequency of 12.5 Kc/sec. Too little is known to assess the problems of false echos in the regions of the terraces and sea cliffs but the possibility that they may exist must be investigated. Another aspect of the biological population in the vicinity of the rocky terrace is that they constitute an abundant source of sessil organisms which could foul instruments placed on the bottom in these areas. These include sponges, corals, bryozoans, brachiopods, giant anemones, and burrowing clams. What are the military implications of such a great change in sea level? What were its effect on the sea floor, and how can a knowledge of this occurrence benefit the Navy? Most of the new electronic acoustic equipment in use by the Navy is in some way environmentally limited. Especially if it must differentiate between a foreign object in an environment that contains similar, naturally-occurring bodies, or false echo-producing organisms. Size of the object being looked for is important because it dictates the target strength and frequency of search sonars. It is in this respect that the sea cliffs and their associated sediments become important. It is almost impossible to find a bottomed mine, or a lost atomic bomb, by acoustic means in a boulder bed containing individual boulders that are the Same size as the object being searched for (Swanson, 1967). Even more important, how is one to find a sub- marine nestled against or cruising, submerged, along a sea cliff 60-feet high. Could an enemy ASW group detect a submarine that is running slowly along the 600-foot contour a short distance (100 yards) away from a steep cliff? If our submarine is also equipped with a side-looking sonar using rapidly attenuated, very high-frequency sound, it could safely maneuver clear of the cliff and still be acous- tically invisible to surface-search sonars. The other aspect of this problem is, could we detect an enemy submarine doing the same thing? I have been asked whether or not a submarine captain would care to venture into these rock-cliff areas. My answer is that these features were discovered by a small submarine, and that with the proper equipment (well within the state of the art), it would be no problem for a larger submarine to move with ease in the same areas. Such navigation would be much easier than an under-the-ice Polar traverse. CONCLUSION The great CHANGES of sea level that have taken place in the relatively recent past have greatly affected the nature of the sea floor. These changes are not visible on the charts provided the fleet. Practically nothing has been done to exploit the occurrence of terraces and sea cliffs as far as being areas of stabile rock for 114 Dill bottom placement of instruments in acoustic test ranges or the tacti- cal use of these areas as acoustic screens for hiding submarines. I'm reasonably certain that little is being done to improve the capa- bility of acoustic equipment to differentiate natural occurring objects on the terraces and those that may be lost or placed in these areas. We have only begun to investigate the nature of the deeply-sub- merged terraces on the continental slope. Their origin is compli- cated because they were formed during the geologic past when environ- mental conditions were different from those existing today. Little is known about the time necessary for them to form or their distribution throughout the world. A preliminary review of the existing literature indicates, in addi- tion to Southern California, that deeply submerged terraces occur in the Carribean, along the Atlantic coast, off Baja California, Mexico, in the straits of Florida, and off the Oregon coast. A well-developed terrace is indicated in sub-bottom profiles off the relatively unstable insular slope of Japan. The acoustic reflections from the rocky cliffs, coarse shelly sands, and cobble areas associated with the terraces must be very different from those of the intervening soft muds separating the ter- races. This difference has to be considered in any type of ASW or lost instrument search employing acoustic sound sources. Consider- able research remains to be done before we can determine how these sea-floor features affect acoustic search and whether or not they could be utilized tactically by the fleet or as underwater construction and installation sites. ACKNOWLEDGEMENTS The writer wishes to thank John A. Beagles, Bruce C. Heezen, and R. F. Busby for their assistance and discussion during and before the field work. Edwin C. Buffington, Edwin L. Hamilton, and FE. C. LaFond critically read the manuscript and gave many helpful suggestions. BIBLIOGRAPHY Buffington, E. C. and Moore, D. G., 1963, Geophysical evidence on the origin of gullied submarine slopes, San Clemente, Califor- tas) Jour Geol, Ve W.(8), 356-370: Busby, R. F., 1962, Submarine geology of the Tongue of the Ocean, Bahamas. Tech. Report 108, U. S. Naval Oceanographic Office, Wash. D. C., 84 p. Curray, J. R., 1960, Sediments and history of Holocene Transgres- sion, Continental Shelf, Northwest Gulf of Mexico, in Recent Sediments, Northwest Gulf of Mexico. Am. Assoc. Petrol. Geol. P. 221-381. Dill Curray, J. R., 1965, Late Quaternary history, continental shelves of the United States. In Quaternary of the United States. Princeton Univ. Press, p. 723-735. Curray, J. R. and Moore, D. G., 1964, Pleistocene deltaic progra- dation of continental terrace, Costa de Nayarit, Mexico, in Marine Geology of the Gulf of California - a symposium. Am. Assoc. of Petroleum Geol. Mem. 3, p. 193-215. Donn, W. L., Farrand, W. R., and Ewing, M., 1962. Pleistocene ice volumes and sea level lowering. J. Geol. V. 70, p. 206-214. Emery, K. O., 1958, Shallow submerged terraces of southern Cal- ifornia. Bull. Geol. Soc. Am. 0.69, p. 39-60. Emery, K. O.,1960, The Sea off Southern California: A modern habitat of Petroleum. John Wiley & Sons, N.Y., 366 p. Emery, K. O. and Terry, R. D., 1956, A submarine slope off southern California. Jour. of Geol., V. 64, p. 271-280. Ewing, J., Luskin, B., Roberts, A., and Hirshman, J., 1960, Sub-bottom reflection measurements on the continental shelf, Bermuda Banks, West Indies Arc, and in the west Atlantic basins. J. Geophys. Res., V. 65, p. 2849-2860. Garrison, L.E. and McMaster, R. L., 1966, Sediments and geomor- phology of the continental shelf off southern New England. Mar- ine Geology, V. 4, p. 273-289. Goreau, T. and Burke, K., 1966. Pleistocene and Holocene Geology of the island shelf near Kingston, Jamaica. Mar. Geol., V. 4, p. 207-225. Hamilton, E. L., 1963, Sediment sound velocity measurements made in situ from bathyscaph TRIESTE. J. Geophys. Res., V. 68, N. 21, p- 5991-5997. Moore, D. G., 1957, Acoustic sounding of Quaternary marine sedi- ments off Point Loma, California. U.S. Navy Elect. Lab. San Diego, Rept. 815, 17 pp. Moore, D. G., 1960, Acoustic-reflection studies of the continental shelf and slope off southern Calif. Bull. Geol. Soc. Am. V. 71, De AT oVlsGs Moore, D. G., 1961, Submarine Slumps. Jour. Sed. Petrol. V. 31, N. 3, p. 343-357. Moore, D. G. and Curray, J. R., 1963, Structural framework of the continental terrace, Northwest Gulf of Mexico. Jour. Geophy. Res. V 68, 1725-47. 116 Dill Shepard, F. P., 1964, Sea level changes in the past 6000 years: possible archeological significance. Science, V. 143, p. 574-576. Shepard, F. P. and Dill, R. F., 1966, Submarine Canyons and other Sea Valleys. Rand McNally, Inc., Chicago, Ill., 381 p. Swanson, L. U., 1967, Aircraft Salvage Operations Mediterranean, Lessons and Implications for the Navy. F. A. Andrews, general editor. Executive Summary of final report by CNO Technical Advising Group. Dept. of the Navy, Washington, D. C., 39 p. Terry, R. D., 1965, Continental slopes of the world. Ph.D disserta- tion, Univ. So. Calif., Los Angeles, 648 p. Vernon, J. W., 1965, Shelf Sediment Transport System. Report No. USC-Geol. 65-2, Coastal Engineering Research Center, U. S. Army Corps of Engr. DA-49-055-Civ. Engr. 63-13, 135 p. 4 ; ia rer fi : Lt ie a MR Ml ie } bg sie ee OP ae YG ay eli Be BOGE he fe ue 4 ‘ “oO by me * ed Parra! ay i y ‘ i 0 r my j / y } ¥ P he f o I i) u i t ee : i i , ‘ ; : , Sy A teea, 4 vy : Mi “J i ; ; 1g we oy ‘& reyes 4 a Reaper hat tt Session C OCEANOGRAPHIC PREDICTION REAL-TIME CCEANOGRAPHIC DATA FOR OCEANOGRAPHIC PREDICTION Kennard M. Palfrey, Jr. U. S. Coast Guard Oceanographic Unit Washington, D. C. The need for oceanographic forecasting is manifold and cértainly well recognized. In the field of military oceanography, particularly, oceanographic prediction is of paramount importance. The ability to describe and predict ocean phenomena under any scheme depends in part upon time-series observations at selected control points. Ever since the Ocean Stations were established during World War II, their potential for the collection of synoptic oceano- graphic data together with synoptic meteorological data has been recognized. During its first session, in 1962, the Intergovernmental Oceanographic Commission emphasised the value of the Ocean Station vessels for monitoring oceanographic conditions and recommemded that fuller use be made of their potential. The Ocean Station vessels, which are normally required to maintain station within a ten-mile Square centered on each ocean station, afford opportune platforms for time-series observations. Consequently, when oceanography was added to the statutory functions of the U. S. Coast Guard, in 1962, immediate attention was given to the development of this oceano- graphic potential. The high endurance cutters which had been routinely occu- pying the six U. S. operated Ocean Stations had been making meteoro- logical and bathythermograph observations for years. Since passage of the 1962 law, thirty-one cutters, all those performing ocean station duty, have been equipped with oceanographic laboratories, oceano~ graphic winches and related equipment to provide a basic Nansen cast capability. Four classes of high endurance cutters are employed in the occupation of the four North Atlantic and two North Pacific Ocean Stations operated by the U. S. Coast Guard. These are the 327' 121 Palfrey Secretary Class cutters, the 255' Lake Class cutters, and the 311' converted AVP Class cutters. The CGC HAMILTON, which was recently commissioned and is the first of a new 378' Secretary Class cutter, was designed and built for oceanography and will be manning North Atlantic Ocean Stations in the near future. Since early 1963, when CGC CASCO demonstrated the feasibility of oceanographic observations by Coast Guard Ocean Station vessels, through successful completion of a pilot project at Ocean Stations DELTA and ECHO in the North Atlantic, the program has been greatly expanded. At present the Ocean Station program is functioning at forty percent of its goal of continuous observations at all six stations. The achievement of the ultimate goal has been limited solely by the availability of deep-sea reversing thermometers. A problem which is not unique to the Coast Guard. Oceanographic observations are made at all stations except ECHO on an alternate patrol basis; a patrol being twenty-one days duration. The ECHO project is currently being accomplished on a seasonal basis. The oceanographic observation program at each Ocean Station requires a daily oceanographic station to a depth of 1500m and at least once during the patrol a daily station is extended to as near the bottom as practicable. Accepted sampling techniques and proced- ures are adhered to rigidly. In 1964, the Coast Guard Oceanographic Unit was permanently established in Washington, D. C. with the mission to develop and support the total Coast Guard oceanographic program, including the time-series program at the Ocean Stations. This support includes data collection, processing and dissemination, instrument development and calibration; establishing oceanographic techniques and procedures; and liaison and cooperation with other agencies. It was recognized quite early that the large number of vessels of opportunity involved in the Ocean Station program would prohibit scientific manning. In order to control the quality of data collected and to afford guidance to a vessel when deployed a two-fold program was established. An eight-week school was established at the Coast Guard Training Center, Groton, Conn. to provide two trained enlisted technicians as a part of the permanent crew of each Ocean Station vessel. In addition, a program of real-time quality control of data by radio was initiated. All observed data is transmitted by radio, generally radio teletype, to the Coast Guard Oceanographic Unit where it is corrected, processed by a digital computer and checked by a trained oceanographer on a twenty-four hour a day basis. This permits rapid feed-back to the vessel so that the sampling frequency can be adjusted to better observe changing phenomena and poorly performing instruments replaced. Over 1100 oceanographic stations were handled in this manner during 1966 and a total of over 1500 is anticipated for 1967. An analysis 122 Palfrey of transmitted data revealed an accuracy of 99.6% in comparision to original records submitted post-patrol by each vessel, It must be noted here that the high success of transmitting data is not attribu- ted to any oceanographic scheme but rather to the tenets of security, reliability and speed which govern the Navy and Coast Guard communi- cations system which are used. As a result of the real-time quality control procedures of the Coast Guard, corrected and verified oceanographic data can be made available to operational and research uses within twelve hours of observation as a maximum, and generally within six hours. Since late 1965 the U. S. Fleet Numerical Weather Facility, Monterey, Cal. has been a consumer for this real-time oceanographic data and has used the information in the development of synoptic oceanographic analyses. The Coast Guard has also participated in the National Oceanographic Data Center's project "HOTLINE" using the same real- time data scheme. The real-time data flow scheme employed by the Coast Guard requires that all data from the observing vessels pass through the Oceanographic Unit for correction and verification before reaching the ultimate data consumers. The CGC EVERGREEN and CGC ROCKAWAY, the Coast Guard's two primary research vessels, have on board data processing facilities and submit completed data direct to any consumer. In order to more fully realize the capabilities of the Ocean Station vessels, a system of standard oceanographic sections was established in 1966. These sections were designed so as to include the areas of maximum oceanographic interest and to take advantage of the normal routes of the Ocean Station vessels. At present, most of the sections are occupied at least quarterly, mainly by an Ocean Station vessel enroute to or from station or by a primary research vessel. Pacific Standard Section One is occupied semi- annually by a station VICTOR vessel, between patrols, as a part of the Cooperative Study of the Kuroshio which began in 1965. In the Atlantic Standard Sections 2, 3 and 4 are occupied several times a month during the winter and spring to provide ocean current infor- mation for the Commander International Ice Patrol. It should be mentioned here that the Coast Guard's interest in the real-time transmission of quality data stems from many years of experience in this field in the course of providing the International Ice Patrol Service. All observations along these sections are treated as for Ocean Station data and are available in the same time frame. An automatic salinity—temperature-depth measuring system is currently installed on EVERGREEN and ROCKAWAY, and will shortly be provided to all Ocean Station vessels by the Navy as an adjunct to ASWEPS. When the system is installed on the OSV's six-hourly obser- vations will be made, with the data entered into the present trans-— mission and quality control system. These observations will be made enroute as well as on station. The problem of long lead time in the 123 Palfrey procurement of thermometers will be overcome and salinity data will be available faster. In conclusion it is emphasised that this is only a cursory look at one system for real-time transmission of quality controlled oceanographic data. The mechanics of the Coast Guard's system are open for detailed inspection and I am sure it will holdup under the closest scientific scrutiny. 124 DATA REQUIREMENTS FOR SYNOPTIC SEA SURFACE TEMPERATURE ANALYSES R. W. James U. S. Naval Oceanographic Office Washington, D. C. 20390 I. BACKGROUND A common complaint voiced by analysts attempting to prepare syn-= optic oceanographic charts is that the present data input is insuf- ficient to produce reliable analyses. This is reflected in the use of composite groupings of observations taken over several days and by the increased demand for ship=-of-opportunity data. The constant requests for more observations, however, leads those in a planning capacity to conclude that the solution to the problem lies in satur- ating the oceans with observations. While it is obvious that more synoptic data are required, it is not so obvious where the obser- vations are to come from, how they should be taken, to what accuracy, and how many are required. Funds availseble for increasing the quantity of synoptic oceanographic data are limited and should be carefully apportioned between possible platforms to give maximum improvement in analysis accuracies. This can be accomplished only by first ascertaining answers to the questions posed above. This report describes the data requirements for sea surface temperature charts as deduced from a series of tests which compare accuracies of analyses based on various data inputs. These investi- gations are part of a continuing research effort concerning data requirements for the Navy's Antisubmarine Warfare Environmental Prediction Service (ASWEPS). II. OBJECTIVE The objective of this report is to establish guidelines as to the quantity and quality of synoptic observations required to ensure reliable sea surface temperature analyses. Although the study was conducted in the western North Atlantic ocean, and with ASWEPS in mind, most of the conclusions should be valid for other ocean areas. T25 James A secondary objective is to find the point of diminishing return where the cost and efforts expended to increase the data input are not balanced by the improvement found in analysis reliability. At this point further expenditures for more data are not justified unless it can be demonstrated that an increase in operational effec- tiveness compensates for the additional expense. III. PROCEDURES A series of analysis tests were made for two areas of about 250 NM on @ side. One srea has a complex thermal field typical of the region just north of the Gulf Stream while the second area is repre- sentative of the rather smooth thermal gradients found in the Sargasso Sea and other areas lacking major currents. The two areas will be designated for the remainder of the report as Area A, which is complex in nature, and Area B, which has a smooth thermal field. Basically the tests consist of starting with an assumed realis- tic isotherm pattern and a data field of observations over a closely spaced grid that are matched to the isotherms to produce perfect observations. Various percentages of the original data input are randomly selected and analysis made of these reduced data. Each analysis is compared to the original data field to compute the mean absolute error, and to the original analysis to ascertain the good- ness of fit. Analyses were prepared both manually and by computer techniques. Error functions are introduced to vary both the quantity and quality of data utilized in a test. In all cases the data field used for analysis is considered a 24 hour data input. Results of the tests reveal many relationships between the quantity and quality of data and the reliability of the analyses. The tests were divided into three major categories as follows: Area A =- manual analysis Area A = computer analysis Area B = manual analysis For each of these major categories there were 8 to 16 individual tests. In addition, there were two special tests to determine the degrees of freedom inherent in the use of random data distributions and to compute the relative advantage of selected data points over randomly spaced data. IV. RESULTS Results of the tests will be described in the following sections under the three major categories and the two special tests described above. 126 James Ae Area A = Manual Analysis of Perfect Data Area A was selected to be representative of the complex isotherm patterns found north of the Gulf Stream. Figure 1 shows the original data plot and analysis for Area A. Sea surface temperatures are considered perfect values, since the data were fitted to the iso- therm pattern. The grid spacing of 12.5 NM covers an area equivalent to a five degree square but a geographic fix is not indicated since the isotherm pattern is considered typical of thermal conditions over @ wide area north of the Gulf Stream. The major features of Area A are two warm tongues separated by a cold intrusion of water from the northwest. Colder water is also evident in the northeast portion of the area. The first test was to randomly select ten percent of the grid data and to analyze this new temperature field. This procedure was followed to see if the essence of the isotherm pattern could be maintained with only ten percent of the original observations. Tables of random values were used to select the 40 observations that would be utilized in the analysis. Figure 2 shows the resulting data plot of 10 percent data by a trained analyst with no knowledge of the original isotherm pattern. Comparison of Figures 1 and 2 indicates that although some distortion occurs in the pattern the major fea- tures are well described and the more prominent tongue to the east was still drawn as the main feature. A mean absolute error was com- puted by overlaying the new analysis over the original data. An error of l. 30° F was obtained, which is small compared to values nor- mally found for analyses in this area. This is not surprising in that there was an adequate input of observations and the data were of perfect accuracy. Three more tests of this nature were made using 5, 3.5 and 2 percent of the perfect data shown in Figure 1. That is, the data plot consisted of 20, 14 and 8 observations respectively. In each case the observations utilized were randomly selected and the analyst preparing the analysis did not have a pre=-conception of the pattern. Different analysts were used to avoid continuity of analysis. Figures 3, 4, and 5 show the resulting analyses. With 5 percent of data the isotherm pattern is still well defined although the source of the cold water intrusion is misplaced. Even with 3.5 percent of data the two warm tongues are properly located, although they are considerably smoothed. For optimum sonar routing, however, the pattern shown would still be of value. Using the 8 observations available with 2 percent of data the analysis does not describe any of the original pattern. The pattern consists of one smoothed warm tongue which is not too accurately centered. Unfortunately, as it will be showm later, present sea surface temperature analyses are based on data input comparable to the 2 percent chart but of mich lower quality than represented by the observations utilized in Fig- ure 5. Only through composite groupings of observations for 3 to 5 127 James A B C 55° D E 60° FE G H 65° J 60° x L M (OF N (0) 65° Pp Q R TO” s 75° T Fig. 1 - Original data and analysis 128 James Os SEP Cor SSP 5° 60° Sy 65° 60° tO? 65° 75° 70° Fig. 2 - Analysis of 10 percent of data 129 James SS” SO" Gs" SS” GO? 55° 55e 60° 55g Boe 65° 65° 70° 0OP 752 Ue Fig. 3 - Analysis of 5 percent of data 130 SOs 5)” Gee 70" iGo ao io CS Fig. 4 - Analysis of 3-1/2 percent of data 131 James 55° 60° 54 66 68 55: 65° 68 60° 69 70° 76 65° 73 73 70° 75s (e Fig. 5 - Analysis of 2 percent of data James days are sufficient observations obtained to prepare useful analyses. Table I gives the mean absolute error for each of the anal- yses discussed above plus a value for 30 percent of data. At 100 percent of data the error is obviously zero since the fit is assumed perfect. TABLE I. ACCURACY OF MANUAL ANALYSES FORAREAA 0 Percent Data Mean Absolute Error (OF) 100 0.00 30 0.43 10 1.30 3) 1.70 35 1.89 2 2.95 As expected the mean absolute error decreases with the in- creasing availability of data. Thus, by doubling the data input from 2 percent to 4 percent the analysis error is reduced by 39 per- cent. Tripling the 2 percent input reduces the error by 50 percent but above this point the increase of data does not produce noteable improvements in analysis reliability. For instance, although the error is reduced by half if the data is tripled it would take 12 times as much data to reduce the error by half again. Table 1 pro- vides information concerning the quantity of data required to ensure specified analysis accuracies, assuming perfect data. Since the present data used for synoptic charts are far from perfect the mean absolute errors must be recalculated using data typical of that pres- ently available. This was done in the second series of tests. (1) Error Functions For the second series of tests for Area A the perfect data were modified by introducing errors similar to those found in present ship injection temperatures. Two error functions were assumed; one for accuracy of temperature and the other for accuracy of ship position. The errors of the input temperatures were assumed to average 2°F, with a range of +T°F, end follow somewhat & normal probability distribution. With this error distribution applied to 10 percent of data, for example, there would be 6 perfect observations, 10 with +1.0°F errors, 8 with +2.0°F error, 6 with +3. O°F errors, 4 with +2. O°F errors, and 2 each of +5; +6, and +7 OF error. 133 James These errors are considered realistic based on past studies of water injection temperature reports. Saur (1963) investi- gated the difference between injection temperatures recorded in the log and specially observed temperatures for 6826 pairs of observations from 12 ships on 92 different trips. He found a generally normal probability distribution of errors centered around a small positive bias. Errors of 6°F were shown with a standard deviation of differ- ences of 1.6°F. Gibson (1960) compared sea surface temperatures as reported synoptically to those that had been recorded as a log entry and found 13 percent of the observations were in error due to coding discrepancies. These errors were small, however, generally not ex- ceeding 2°F, Other investigations of injection temperatures by Francesehini (1955), Kuhn and Farland (1963) and Beetham (1966) show the assumed error function to be realistic. In addition to the error in the temperature observation there is also a navigation error in present ship reports, which con- tributes to analysis errors. This was applied by assuming that one third of the reporting ships were within the quadrangle represented by @ position reported to the nearest tenth of a degree. That is, no position error within the capabilities of the code. A third of the reports were assumed to be +6 miles from the true position and another third with a +12 NM error. The magnitude of the movement errors were applied randomly while the direction of the deviation was uniformly distributed to the four cardinal directions. (2) Description of Error Tests Three cases of errors were treated using various per-= cents of the original data. In Case I the full range of accuracy errors and movement errors were applied as described above. The resulting data plots of "bad data" are considered equivalent to the present sea surface temperature data used in preparing analyses. The Case II tests were based on data in which the error was limited to 41°F although the navigation error was the same as in Case I. This would be equivalent to using only temperatures reported from ships with reliable instrumentation, such as the Near Surface Refer- ence Temperature (NSRT) System described by Beetham (1966), or per- haps airborne radiation thermometer observations when carefully taken. A third case was run where it was assumed reliable instruments were available so that the error would be +)°F and the navigation error sufficiently accurate so that the ship is actually in the quadrangle of area reported. An example of the three cases are shown in Figure 6 along with typical analyses. ‘The original data are altered in value and position according to the three cases previously discussed. The product of misplaced data and inaccurate readings obviously misleads the analyst into the wrong orientation of the isotherms in some cases. 134 James c a *69 °67 ORIGINAL DATA NE abe: sai 5 ll snes 27 67 73 MOY 0 ° Ti 70 69 CASE I ACCURACY AND MOVEMENT ERROR Mo CASE IL LIMITED ACCURACY AND MOVEMENT ERROR CASE It LIMITED ACCURACY ERROR ae Fig. 6 - Examples of modified data field 135 James Figure 6 is simply a model to illustrate the three error cases applied and was not part of the tests. er ere er ‘Figure 7 shows the analysis of 10 percent of original data which have been randomly modified by the Case I error function. This figure illustrates what happens when 40 observations of the quality presently utilized in synoptic analyses are analyzed instead of 40 perfect observations as in Figure 2. One of the two tongues is located relatively well but the cold water and the western tongue ere misaligned. Gradients are also over-emphasized in some areas. Analyses of 5, 3.5 and 2 percent modified Case I data are not shown but revealed a poorer fit to the true isotherm pattern than did analyses of the same quantity of perfect data. As would be expected analyses of the same quantities of data for the Case ITI and Case III errors produce isotherm patterns worse than those from perfect data and better than the analyses of Case I data. All anal- yses show that at 2 percent of data the isotherm pattern is not reliable but this qualitative information was not the goal of this study. Mean absolute errors were computed for all tests and it is these values that provide quantitative answers to some of the questions posed in the objective. The mean absolute errors were plotted versus data input with a family of curves resulting as show by Figure 8. For purposes of comparison assume that 2 percent data is our present data input; this will be justified in a later section. The following conclusions can be made from Figure 8: (a) Im all cases an increase in data produces a de- crease in the mean absolute error or an increase in analysis reli- ability. The reliability of the analyses can be increased consider- ably by simply doubling the present input data. Indications are that it would be desirable to triple or perhaps quadruple the pres- ent data input. (bo) The same improvement in SST analyses can be ob- tained by quadrupling the data input using present type SST reports, or by only doubling the input but using more reliable data. This could be accomplished, for instance, by using only ART or NSRT data instead of the present predominantly ship of opportunity injection reports. 2 Increases in reliability beyond 8 percent (quad- rupling the data) may not be desirable owing to the small improve- ment per data increase. 136 James 6571602 8 6 60° 62 65° 70° 70° LG 77 73 74 Ae 04 75° 80° 80° 0S Fig. 7 - Analysis of 10 percent of modified data (case 1) 137 MEAN ABSOLUTE ERROR (°F) James DATA INPUT (%) Fig. 8 - SST accuracy, manual analysis, area A 138 James (4) Absolute Data Requirements - Area A So far this report has dealt with quantities of data relative to the original base of 400 perfect observations. Tt was found that analyses of as little as 3.5 percent of these data re- vealed useful. information concerning the isotherm pattern. With 2 percent data input it was not possible to show the detail known to be present in the isotherm pattern. For several reasons the 2 percent data input is felt to be equivalent to the present data availability utilized in ASWEPS. The mean absolute error for the analysis of 2 percent of present data is 3.46 F. This value is close to the mean value of the mean absol- ute errors found in nine evaluations of present manus] sea surface temperature analyses. These evaluations included 1786 individual verifying temperature observations and_the mean absolute error aver= aged for all nine evaluations was 3.61°F. Reports by James (1966), Tuttell (1963), Shank (1966), Carman (1965) and James (1965), de- seribe these evaluations. The second reason for assuming the 2 percent data in- put is typical of vresent data availability is that 2 percent of the original data is eight observations and the average number of obser= vations found per day in an equivalent area during 1966 was 7.8 observations. On this basis 2 percent of the data used in the tests is considered equivalent to the present data input of ASWEPS. Thus from Figure 8, to reduce the analysis error to below 2°F would re- quire 8 percent of the present type data or four times as mich as presently available. This would be 32 observations per five degree square. The same analysis reliability could be obtained, of course, by doubling the data input but utilizing only accurate data as used in Case III. This would require only 16 observations. Of course the increase of data will be mostly ship of opportunity injection reports for a time and gradually the better instrumented reports will predominate. Thus, the true curve representing analysis accuracy with increasing data will slice across the three curves in Figure 8, approaching the Case III curve in time. (5) Non-Random Data Figure 8 shows the errors that result from analyses of various quantities and qualities of random distributions of data. Tf ship of opportunity data were rejected and only ART or buoy data utilized then the distribution of data could be pre-selected. To ascertain the value of specifying the data distribution instead of accepting a random field several ART tracks and buoy arrays were specified so as to give varying quantities of perfect data. IL Be) James The results were that the mean absolute error was re- duced by 20 to 40 percent, with the larger reduction occurring with the higher number of observations. The results of these tests are shown by Table II. TABLE II COMPARISON OF SEA SURFACE TEMPERATURE ACCURACIES FOR RANDOM AND SPECIFIED DATA DISTRIBUTION Mean Absolute Error (°F) Percent Data Random Specified 10 1.30 -69 5 1.70 1.08 365 1.89 1.39 2 2.95 2.81 The conclusion from these tests is that if more use is made of specified, reliable temperature observations instead of ran- dom ship of opportunity data the accuracy of the analyses will ap- proach the perfect data curve of Figure 8. Routine ART flights for instance would provide data of sufficient quantity and quality, and of a specified nature, to permit analyses of less than 1°F error. _ Be Area A = Computer Tests Tests similar to those described above using manuall ana- lyzed temperature charts were also made using computer analyses. This was done for two reasons: (1) To verify that the family of curves show in Fig- ure 8 are not biased by analysts skills. (2) To ascertain whether objective analyses introduce any special requirements or lead to difference emphasis on data requirements. The original data input to the computer analyses was the same as used in the manual analyses and shown in Figure 1. Similar randomly distributed data inputs of 10, 5, 3.5 and 2 percent were analyzed and the mean absolute error computed. Im order to maintain the same degree of pre=-knowledge as in the manual analyses a flat field was used. This means that every grid point was considered zero except those at which a true observation existed. In the case of the test of 2 percent data this means there were 392 grid points of zero and eight observed values. The computer program makes mitiple passes over the grid relaxing the grid values at non-observed points until the grid field shows certain minimm differences between grid values. A descrip- tion of the computer program from which this test program was drawn is given by Thompson (1966). In this case no smoothing was applied, 140 James since the area was so small. Also no forcing function was used in order to avoid injecting prior knowledge of the isotherm pattern. (1) Results of Computer Tests - Area A Figure 9 shows the computer analysis of 10 percent data. Although this analysis lacks the resolution of the manual analysis shown in Figure 2 the major warm tongue is shown as is the cold water intrusion from the northwest. An interesting aspect of computer analysis is shown by Figures 10 and 11. Both figures represent anal- yses of 2 percent of data but one of perfect data (Figure 10) and the other of poorer quality, Case I data (Figure 11). The interesting point is that there is little difference between the analyses in spite of the great difference in the quality of observations. The same similarity was found in comparison of computer analyses of other percents of good and poor data. This leads to the conclusion that the computer program han- dles imperfect data better than the manual, subjective analysis, since in the latter tests more differences are apparent. This char- acteristic of the computer analyses may be due to the relaxation pro- cedures which tends to modify the temperature field around an obser-= vation without creating extreme gradients or producing distortions in the field. An analyst on the other hand tends to extend the iso- therms to include high or low values far from the main warm or cold tongue. The mean absolute errors for the computer analyses of per= fect and Case I data are shown in Figure 12. For reference the curves for the same manually prepared analyses are repeated from Figure 8. Although the curves differ slightly the computer curves verify those derived from the manual analyses. No particular differ- ences in data requirements appear to result from these tests except that computer analyses can make better use of a minimm input of poor data than can manual analyses. As expected from the similarity of their patterns, the differences in accuracy between the computer analyses of 2 percent "good" and “poor" data is small. Although the manual analyses show slightly lower errors than the computer products this has no significance. The computer program used was not as sophisticated as that used operationally, since a simplified program was adequate to establish the curves. (2) Effect of The Degree of Randommess on Results In conjunction with the computer tests a test was also run to ascertain the effect of the degree of randomness in the distri- bution of data on the results. 141 James ISOTHERMS - °F Fig. 9 - Computer analysis of 10 percent of perfect data ISOTHERMS~°F Fig. 10 - Computer analysis of 2 percent of perfect data 142 James Co ISOTHERMS - °F Fig. 11 - Computer analysis of 2 percent of data (case 1) Fig. 12 - SST accuracy, computer analysis,area A MEAN ABSOLUTE ERROR (°F) DATA INPUT (%) 143 James It is obvious that the accuracy of an analysis of a given random data. plot is a function of not only the number of observetions but also their distribution. Ore analysis of eight random points does not clearly establish an analysis accuracy since 411 the data may, in that particular case, be located in one portion of the chart. To investigate this problem various random distributions of equal numbers of observations were analyzed on the computer. The results showed that for larger number of observations, 5 and 10 percent of data, the randomess of plot makes no significant differences. That is, with a sufficient number of observations any random distribution provides some values in all quadrants of the chart. With 2 percent of data, however, the randomess does make a difference. Eight random data plots of eight observations each were anal- yzed and the mean absolute errors computed. The distribution of ob- servations varied from cases when the data was evenly distributed over the charts to one case where all eight observations were located in the eastern half of the chart. The mean absolute errors varied from 2.6°F to 4.0°F with a mean value of 3.3°F. This mean value conm- pares well with the 3.2°F error shown in Table I for 2 percent data in the original tests. Tims, the family of curves shown in Figure 8 are considered representative of analysis accuracy. In eddition the slight smoothing necessary to establish a familv of curves tends to bring the value of an individual score closer to the true mean of all random samples. C. Area B - Manual Analyses The tests described so far have established some facts about data requirements for complex trend areas. In order to study the re- quirements for data in non=-comolex areas a similar set of tests were repeated for a relatively smooth isotherm field. This temperature field and analyses is shown in Figure 13, and represents generally the area of a five degree quadrangle similar in size to Area A. This area, will hereafter be referred to as Area B. Where Area A had 2 temperature eradient of 25 degrees across the area, Area B shows only a 6 degree temperature range. This is typical of areas such as the Sergasso Sea, the eastern North Pacific, eastern North Atlantic and other ocean areas away from major current systems. With such a flat thermal field the variation in temperature between grid points is less than 1°F. A wide belt of temperatures result for a given value rather than individual whole values as in the complex field. Thus the 64°F isotherm in Figure 13 is drawn through the center of a number of 64°F temperatures. This results since the values on one side of the 64°F isotherm may represent a true value of 63.7 while a 64°F on the opposite side may represent a 64.4°R but rounding to whole degrees for coding purposes gives the same values. 144 James ISOTHERMS — °F 64 Fig. 14 - Analysis of 2 percent of data 145 James It is obvious from this discussion that highly accurate analysis in smooth areas requires observations reported to tenths of - That is, to define a particular isotherm to the same exactness as is in a complex area requires more accurate data. This results simply from the fact that in Area Aa 1°F temperature change occurs in 12 IM while in Area Ba 1°F change may be spread over a distance of 100 NM. (1) Results of Area B Tests Inspection of the various analyses made for Area B re- veal that excellent accuracies can be realized in relatively smooth areas with a minimum of data if the data is of high quality. Note for instance the analysis of 2 percent of perfect data as shown in Figure 14. Although there were only eight observations available the analyst was able to portray a very reliable picture of the iso- therm pattern. The mean absolute error was only 0.4°F for this analysis compared to 2.95°F for eight observations in the complex Area A. On the other hand, poor data, in any quantity, leads to very erroneous results in analysee of Area B. Applying the same error functions as previously discribed in Section IV A (1) analyses were made of present tyne data. Figure 15 illustrates the analysis of 10 percent of this Case I data. Numerous features are shown that do not belong there; a total of six tongues as compared to the smooth isotherm pattern actually present. Of course operationally continu- ity from chart to chart and grouping of data tends to reduce this problem but nevertheless poor data has a very deleterious effect on analyses of smooth areas. The effect of poor data in deteoriating the reliability of analysis is obviously mich more evident in Area B than Area A. The reason for this is that an observation of 5°F in error and 12 1M out of position may create a entirely new warm tongue in Area B while in Area A the already high horizontal gradients mask the effect of one bad observation. Figure 16 shows the family of curves resulting from analyses of the four types of data described previously as perfect, Case I, II and III. As discussed above and show by Figure 14 very few observations are required for reliable analyses if the obser= vations are perfect. The perfect curve shows little improvement after 4 percent of data. Case II and Case III curves show mean ab- solute errors for analyses of temperature observations of 1°F accu- racy but with (Case II) and without navigation errors (Case III). The closeness of the two curves indicates that the navigation error is relatively unimportant in smooth areas if the data are good. This is expected since the thermal field is so flat. The difference between reliability of analyses of a given quantity of "good" and "pad" data is very evident from Figure 16. 146 James ISOTHERMS~°F Gas SS 56 60 Fig. 15 - Analysis of 10 percent of data, case 1, area B Fig. 16 - SST accuracy, manual analysis, area B MEAN ABSOLUTE ERROR (°F) See ‘ pig PERFECT DATA INPUT (%) 147 James In the analysis of 10 percent of present=-tyne data (Case I) the mean absolute error was actually lerger than for 5 per- cent data (the solid curve shows the expected curve). This was un- doubtedly due to the fact that greater quantities of poor data is more misleading than smaller quantities. The conclusions to be drawn from the Area B tests are as follows: (a) Very little more data is required for smooth areas than is now available but the data must be of a quality com- parable to that provided by NSRT or buoys. (b) It is not feasible to increase the deta input in smooth areas with low quality data, since this leads to little im- provements in analysis accuracies. (c) The use of ART is not required in smooth areas Since the need is for a small mumber of good observations, not a high density of data as provided by aerial survey. (a) Buoy observations would be ideal for smooth areas since they provide both a non-random data plot and high accuracy tem- perature data. (2) Absolute Data Requirements for Area B From Figure 8 it appears that present type data should be replaced as soon as possible by more accurate data and no great effort expended on increasing ship-of-opportunity data unless it is from a NSRT system. On the basis of the discussion in Section IV A(5) con- cerning selected versus random data distribution it is safe to con- clude that a fixed array of buoys would give accuracies comparable to the perfect data curve show in Figure 16. Thus, a 2 percent data input is completely adequate, which would mean eight observations or if accuracies of 1 F would be tolerated, perhaps five or six reports. Combining buoy observations and random ship reports from NSRT systems one could obtain reliable analyses from one buoy plus six to eight ship reports. Owing to the lack of advection in smooth thermal fields it is likely that continuity would contribute highly to analysis accuracy. An occasional ART survey over the area to obtain a high data density of good accuracy would provide a highly reliable anal- ysis which would be useful for continuity for 10 to 15 days. 148 James D. Estimated An is Accuracies In order to obtain some measure of analysis accuracy ex- pressed as a percent score the following formla was applied to the results of the tests described in this report: T- £5 A= rv Where A = percent accuracy total range of temperature in area, OF mean absolute error, ag E This formula obviously can not be used for all cases and areas but does give a means of comparing analyses in complex and smooth areas. It considers both the degree of variation in the area to be analyzed and the mean absolute error of the analysis. A mean absolute error of 2 F may be considered good in a highly complex area but not so good in a smooth thermal field. For Area A the temperature range was 25°F, from 53°F to 78°F. Thus the present 3.4 F error would be equivalent to an analysis accuracy of 86 percent. If the analysis error can be reduced in time to 1.4°F the accuracy score would be 95 percent. For Area B the temperature range was 6°F, from 61°F to 67°F. Present accuracies of 2.3°F show analysis accuracies of only 62 per- eent. Through suitable data sources and little new data this accur- acy would be improved to 97 percent. A question that comes to mine here is what accuracy is desired? This, of course, depends upon the use to be made of the analyses but generally the charts mst be reliable enough to satisfy the most stringent operator requirements. This would require accur= acies of 95 percent. V. SUMMARY AND CONCLUSIONS This report described a series of tests designed to show what the data requirements are for synoptic sea surface temperature anal= yses. It was shown that data requirements are not unlimited but that a definite quantitative value can be placed on how much data are re- quired. Data requirements for smooth and complex thermal areas were shown to be significantly different and should be so treated in plan- ning of synoptic nets. The number of observations required for reliable sea surface temperature analysis depends upon the area, the type of data and the reliability desired. Table III shows the data requirements estimated for various types of data input and analysis accuracies. 149 James TABLE ITT NUMBER OF DAILY OBSERVATIONS PER FIVE DEGREE SQUARE Complex Area Smooth Area Chart Reliability 85 90 95% 65 75 85 5% Data Type Present Ship 8 16 ho ney (De) tt: Injection Data Random NSRT Reports Te ee 35 5 © 8. € Plus Some ART, Buoy Mostly ART, Buoy Data 7 8 30 53 5 Gu 8 Plus Some NSRT *Not attainable with vresent type data The results of these tests are considered valid on the basis of the number of tests conducted including computer and manual, tests for randommess, for specified data distribution and for error func- tions. One procedure that was not treated in these tests was the use of continuity analyses. These charts obviously help in the prepar- ation of a synoptic chart when correct. If wrong they simple mislead the analyst. On the assumption that their popular use by all anal- ysis groups shows a degree of corrections it is concluded that all scores would be improved slightly by their use in the tests. The slopes of the curves, however, would not be materially altered. The major conclusions are as follows: (1) Planning for synoptic network should emphasize high concentrations of data with average accuracies to +1.0°R for corplex areas. In smooth areas the emphasis should ke on a few highly ac- curate (less than +1 F) fixed observations. (2) Data should be increased about four times in com- plex areas and improved in accuracy. For smooth areas the quantity of data need only be improved. (3) ART surveys should be confined generally to the highly complex areas with an occasional flight for continuity in the smooth areas. Buoy arrays should be utilized in smooth areas where their high accuracy and fixed position contribute greatly to anal- yses reliability. 150 James REFERENCES Beetham, C. V. "Test and Evaluation of the Near Surface Reference Temperature System", IM No. 66-10, Aug 1966, 14 pp. Carman, D. Re. "Comparison of U. S. Naval Oceanographic Office Sea Surface Temperature Charts with USNS GILLISS Survey Data, 2-5 March 196k", IMR 0.65-64, gan 1965, 6 pp. Gibson, Be. A. "Note on the Reliability of Transmitted Sea Surface Temperatures", IMR No. 11-62, Jan 1960, 5 pp, NAVOCEANO. James, R. W. "Accuracy of Sea Surface Temperature Analyses", Part ‘III, IM No. 66-23, Sep 1966, 14 pp. James, Re We "A Quantitative Evaluation of ASWEPS Sea Surface Temperature and Layer Depth Charts", IMR No. 0-39-65, Sep 1965, 37 pp. Kuhn, J. A. and Farland, R. "Sea Surface Temperature Measurement System (SURTEMS), IMR No. I-2-63, May 1963. Saur, J. F. T., "A Study of the Quality of Sea Water Temveratures Reported in Logs of Ships' Weather Observations", Journal Applied Meteorology, Volume 2, N3, June 1963, pps 417-425. Shank, M. K. Jr, Unpublished, "An Investigation of Grid Spacing and Data Requirements in the Preparation of an Objective Sea Surface Temperature Chart", Feb 1964, 10 pp. Shank, M. K. Jr., "Comparisons of Analyzed Sea Surface Temperatures With Observed Data (Jan - Feb 1966) IM No. 66-22, Oct 1966, 13 pp. Thompson, B. J., "IMwmerical Techniques", SP 109, ASWEPS Manual Series, Volume 9, 1966, 26 pp. Tuttell, J. T., "Comparison of NAVOCEANO SST and LD Charts With USNS DAVIS Survey Data, IMR No. 0-54-63, June 1963, 30 pp. 151 LARGE-SCALE ANOMALOUS SEA SURFACE CONDITIONS IN THE NORTH PACIFIC John D. Isaacs Scripps Institution of Oceanography La Jolla, California Aperiodic departures from the normal sea surface tempera- tures commonly occur throughout the world's oceans. Such anomalous Eluctuations often exceed the mean seasonal or annual fluctuations of temperature in many regions, and thus constitute major features of the oceans. These changes are associated with changes in weather, currents, the distribution of marine organisms, the success of fish- eries, and, undoubtedly, with the propagation of underwater sound, the background noise, and the frequency and distribution of natural tar- gets. It is thus important to inquire into the interactions, under- lying nature and causes of these changes. It has long been recognized that ocean conditions are changeable on a time scale much more abbreviated than the broad cli- matic changes, such as the ice ages. Perhaps the most striking example of these abbreviated changes in the last few centuries were experienced in the North Atlantic in the years 1812 and 1813, when conditions were so frigid that some of the now existing deep Atlantic water may have been generated, and snow lay on the fields of Northern Europe all summer. We are gaining some insight into the year-to-year range of ocean conditions within the last thousand years or so from the study of certain highly stratified basin sediments. The historical meteorological records also lend insight into the nature of these changes. The air temperatures at marine stations have been shown to follow the sea surface temperatures very closely. These records, hence allow us to extend the marine temperature record back in time for a few hundred years. In Figure 1, for example, is shown the air temperatures at three Pacific Coast stations, extending to 1925. This demonstrates a number of features, anomalously high and low seasonal temperatures following one another in a complex pattern. One feature of these records, besides their variations, is the peculiar lack of fluctuation in the decade 1946 to 1956. 52 Isaacs Such changes were long thought to be essentially local phenomena, generated by local variations in currents, winds and up- welling. Several workers, however, noted that extreme changes often occurred in the same year in areas as remote as Japan, San Francisco and Peru. This has been called teleconnection. The work of the Bureau of Commercial Fisheries in analyzing the sea surface temperature records from surface ships has shown that the anomalies are commonly of very large scale and long persistence. Figures 2 and 3 are examples of such large-scale anomalies. A pattern similar to Figure 2 persisted in the eastern North Pacific for a period of more than eight months in 1958 and that of Figure 3 during much of 1957. Hatched areas are colder than normal and clear areas warmer. We now believe that teleconnection is an expression of the very large scale of these features, certainly not some mysterious communication across the reaches of the sea between isolated events. The sudden change in the conditions of the North Pacific in 1956 to 1959 and which terminated the period of persistence, demonstrated afresh the importance and magnitude of the effects of these variations. One of the deepest and most prolonged meteoro- logical lows on record developed off Washington, remaining for three months. A strong narrow countercurrent developed along the west coast of the United States, carrying subtropical organisms as far north as Oregon, southern fish visited Alaska, the monsoon delayed its onset a full six weeks beyond its appointed time, desert isles of the Central. Pacific became clothed with green, the heaviest rains in a decade dampened California, everywhere was a stirring engendered by these events. Figure 4 exemplifies a subtropical marine organism carried far north by the narrow countercurrent, and Figure 5, from the records of the California Fish and Game, shows the invasion of tropical fishes into California waters in these times. The ocean-air interactions related to these changes are partly understandable in a qualitative way. For example, the onset of the 1956-1959 change is most marked in the region of the Aleutians where warm Central Pacific water was carried far north of its usual position by an unusual wind pattern in the winter of 1956-57. Namias has shown that the development of the major features of the North Pacific weather can be better hindcast when the persistence from ocean-air interaction is considered. The importance of attaining a better understanding of these events is now clear, and an approach to this understanding can be designed. 153 Isaacs The remainder of this discussion will outline the plans at Scripps Institution for a study of these large-scale anomalies in the North Pacific. One of the principal defects in our information about these events is the paucity of our knowledge of what is transpiring beneath the sea surface and how deeply the anomalous temperatures are dis- tributed. We do not know whether these anomalous regions result from an unusual transport of water or from some change in cloud cover, evaporation, mixing, heat exchange or other alteration in thermal flux. Some insight on the probable conditions has emerged from a pilot study that we have carried out. This study has brought to light a number of curious features of these changes, all of which guide us in designing a study of their nature. I will show only a few examples of these results, which, none-the-less will demonstrate the strong indications that they provide. Figure 6 presents data for a number of North Pacific stations and shows the log ratio of the monthly temperature change (temporal gradient) at the station and the long-term average monthly temperature change for that station, plotted against the sea surface anomaly for that station. One might properly expect that the temperature change at a station at the time of an intense anomaly might depart greatly from the normal, and that it might be near normal during normal conditions. Figure 6 argues that this is by no means the case, and, if there is any trend, the regions of intense anomaly follow the normal seasonal temperature cycles more closely than do the normal regions! Similar results emerge from the analogous relationship of the monthly horizontal temperature gradient and anomalies. These findings argue that these large-scale anomalies result from similarly large-scale homogenous effects, with variations in the effects occurring principally at the edges. The results also allow the possibility that the "normal" conditions are unstable and that two relatively stable conditions exist, one in which tempera- tures are above normal and one with temperatures below normal. Figure 7 is a plot of the relationship between the anomaly at a station for a given month and the anomaly of the monthly temperature change preceding the month. At first thought this appears to be a naive approach, and as would be expected, most warm anomalies are preceded by anomalous heating and most cold anomalies are preceded by anomalous cooling. Beyond this point, however, is a strangeness to the relationship. Warm anomalies often survive anomalous cooling, but cold anomalies very seldom survive anomalous heating. This is shown by the abundance of point in the fourth quadrant of the graph and the paucity of occasions in the second 154 Isaacs quadrant. Keeping in mind the fact that these data are monthly and surface temperature only, the best explanation for this behavior is that there exists an anisotropy between the heating and cooling pro- cesses. Undoubtedly this results from the production of a thin stable surface layer in anomalous heating, which represents a relatively low total thermal change. Anomalous cooling, on the other hand, must involve instability and convection, requiring a much larger thermal change. Thus anomalous heating is a rapid process and anomalous cooling a slow process. In our study on monthly changes we often catch the cooling in mid-step but the anomalous heating takes place so rapidly that it is most often complete within the monthly time scale. These two examples will serve to present the type of con- straints and insights that our pilot study is providing. I will now discuss the field program from which we hope to derive a much fuller understanding of these large-scale temperature fluctuations. The principal tool that will be employed is the small deep- moored instrument station that has been under development for many years at Scripps. These deep-moored stations were first attempted in the Pacific Proving Grounds at the IVY event in 1951, and later deployed in larger numbers in subsequent tests for the recording of fallout and other weapons effects. Figure 8 shows a typical deploy- ment in that period. These moorings were very successful. Depths of mooring were from 700 fathoms to greater than 3000 fathoms. Later developments allowed a greater penetration of sensors, and our recent models record to depths of 3000 feet and are showing a very satisfactory life. Catamarans as shown in Figure 9, have all remained moored and operating in the open North Pacific for six months or more, and one survived for over two years. Two moorings placed in the equatorial Pacific two and a half months ago have just been reported to be operating and in good shape. We thus have a number of the long deep-sea records of temperature versus depth. Figure 10 shows a spectral analysis of temperature depth fluctuations of a 100 day record taken about 600 miles off the California coast. In this record an incoherent lunar semi-diurnal fluctuation is the greatest. Curiously, the lunar and solar frequency are not significant but there is a strong unexplained coherent semi-solar periodicity. 1 show this analysis to point out that the moored stations are capable of yielding data that is amenable to spectral analysis and the determination of the periodi- cities involved in temperature-depth fluctuations. Another essential test of these data is to determine whether they are comparable to the ships' data from which the Bureau of Commercial Fisheries has derived the delineation of the anomalous 155 Isaacs sea-surface temperatures. Figure 11 shows the correlation between the sea-surface temperature as derived from monthly averages of hourly readings from deep-moored stations and the sea-surface temperatures from the published contoured charts from the Bureau of Commercial Fisheries for the same station and month. The relationships, unexpectedly, is excellent, with deviations between the two measure- ments less than 1/2° F. The prima-facie evidence is thus that the ships' data is very good and the instrument stations will delineate the same type of temperature anomaly (as well as, of course, obtain- ing comparable data beneath the surface). We therefore plan that within the next two years we will establish a series of clusters of deep-moored stations, spread across the North Pacific, in some such an array as shown in Figures 12 and 13. The cluster configuration in contrast to a regular grid is required in order to ascertain the direction of any discontinuity or wave-like motion passing through the area. The four peripheral moorings will record meteorological data, and sea-water temperatures well into the thermocline. In addition, insolation will be recorded. The central station is to be a completely submerged recorder, measuring current in the mixed layer. Some of these clusters may also include a large telemeter- ing station such as the Convair buoy. In preparing for the instrumentation of this program, we are reviewing the traditional meteorological measurements to see if other measurements may more directly relate to the interaction of atmosphere and ocean. For example, measurements of humidity are related to the evaporation or condensation of fresh water. A measurement of evaporation or condensation related to the local sea water would be more direct. This, in effect, would be a wet and dry bulb measurement using sea water for the wetting agent. Of course, such a simple approach as this would not be feasible because of the changing concentrations of the wetting agent as evaporation proceeded. Wind velocity is ordinarily recorded and converted into the Square of the mean velocity for evaluating its interaction with the sea surface. Clearly the mean velocity squared of a turbulent wind may be quite different from the mean square velocity, whereas in a steady wind they will be much the same. It is perhaps more direct and meaningful therefore to measure and record V2 rather than V. These are some of the instrumentation problems that we hope to resolve in the near future. We plan then to launch a significant attack on the unknown nature and causes of the large- scale fluctuations in sea conditions that have been instrumental in generating much uncertainty in man's meteorological forecasts and in his fisheries, agriculture, marine transport, underwater sound 156 Isaacs conditions, etc. Our study of the varved sediments, to which I alluded earlier, will provide us with perspectives on the range of conditions that may be within probably future experience. 157 Isaacs * BTULOJITED ‘oseatq ues (9) SeturofITeD “OOstoUeIY UeS (a) ‘uojsuTyseM ‘pues, ysooj}ey, (e) *(Tenuue ‘Tpey ‘Sutzds) soanjzerodurey rare ueew - [ “Sta so Bb o o oS fr) so z B es g & S 5 Vinee Use Tan | pa T c22Ur as a NG mA NVaW| TVANNY a 1 Sno 4-—>x + ae w Ave: Zs 3 NvaW TWv3 3 | I os = 'S6I 2561 40561 8 eG! 9b6l bb6l 2061 Opsi Uae oat DoH AEX VA a vee aS ve ae . -/~ T a =\ | I i L aI a ae al =i) ee re ee ee ee ee ai ala T Ts eal Se 1 =a | | | : | Pp —= 4 T T ' Aare Ne ONS _ ma ei es = g | a Eee . “ = Tie SWONNS Sap IL | SS é iN Sa 1S eA need 158 Isaacs "8G61 ‘OZ-I1I 1sNSnV °G7zz °O'H jo 42eYO UeSUT reok-Q¢ uto0rs (4¥,.) Panjeredurs} sdezans vas jo ATeUIOUY - 7 ‘314 V3uv Twnd3 1Y38WVT OIS 940 002 | “mary ys w a | 8S- 6-01 ‘MIMONOH | | SNOUWOILS3ANI ANBHSI4 SINV3OO 91410ud “LOaROwd 3YOOVETV AGSNN3Y-TVLSNOLTS 3HL 40 Luvd | SVVIOIONIWM 3 8 ANYON MT AB GauVdaud JOVY3AV NVH ¥30109 SVauy . | G3HOLWH S2z OH 4O LYVHO NV3W YV3A-O€ WOUd on San Soc (de) 3UNWHIdW3L 3OV4YNS WIS JO ATVWONY ‘ 8S6l O2-I1 LSNONV i == == —— = 159 Isaacs — AGO “Wea jt Aesesaeiey PSV? °GPISi WO daepe) weeewe: reoh-0¢ Wor (q,.) P1njeredutey soeyans vos jo ATeutouy - ¢ ‘314q 00d! o0f | Ot! 00S! eas 009 = a ede ae I pe 44 eee Bee ll al Vo CLZ2 2+ \ 8S-si-I MM ONOHSNOILVOILSSANI AYSHSI4 OINVS90 914100 “VTOIONSW M3 ONY AVMVT1V9 f'Y AB Gauvd aud ‘NobS NVHL Y31V3H9 SR0NLILV7 LV GaYyNOLNOD LON S3INYWONV S9VY3AV NVHL Y30709 SVauy Q3HOLVH S22 0H JO SLYVHO NVAW YV3A-OF WOU {4e) SUNLVYIDWAL 3OVINNS WES 4O ATVWONY Z2S6I O2-11 ANVANUL NAMA AAT 7 Y RZ en et Se f $1.02 160 Isaacs E. COLUMBIA R. number /1000m® = 1-49 TM] 50 - 499 500 - 4999 Fig. 4 - Nyctiphanes simplex, April, 1958 161 Isaacs LG6I JO tequieydeg ysno1y4 PFE Wor ysTZ Jo 1oqumu ut sotoeds [TeraAas Jo yo3ed yeoq Ajyred Tenuue [eJOT, - g “3tq q4f0de1 ATCUTUT [etd » gloqueydas ysnomyy = Coyg intr’s) Gon L9G ‘9QT 64g “QLT GLO ‘06h LG6T £90 ‘€2G og eT gl HOt TO 961 ‘62 €09 ‘19 9G6T 990 °96t @) OT il 60h ‘2d Q9t “9E 296 “GT GG6T O6T “ZG All 0G 0) gL0‘oL 219 6 Ot 2GG ‘2ge GET OnT £ZOG @) 612 ) Tee “9 AO) Paw ee 0GG ‘OLT €C6T 969 ‘Z9G g ge +E 619 62. £92 °6G 0GG‘°9EE ZG6T 616 £9GG 0) FASE 9G GL ‘HT Tel ‘2 GiG ‘692 TG6T +920 “HG T TE 9 6G€ “2 TL6‘9 LOE £9G2 OG6T G16 69h @) 6 TI GEG OWL LT Cat “99E 6h 6T LGL‘LOn 0 09 QT 616 HT 920‘ ET 90 “HQE Qn6T ft S6GE Gi 969 LET 964 “9E 81669 6+ £119 L+6T skeq USTZ yoefdtys veuny zeTsuy —- -uTYyd oq UT JMOTTOR OVTUODG TLTeIMOTTeX epnoer1eg I@OT 162 Isaacs 1 60 50 40 30 20 x0) NEGATIVE ANOMALIES = ( ie) e e h OOM : 4 + e 1956-57 REGULAR STATION POINTS + EXTREME ANOMALIES 10 20 30 40 50 60 Fig. 6 - Graph of ratio of 12-year mean monthly temporal temperature gradient to monthly temporal temperature gradient versus anomaly-deviation from 12-year mean monthly temperature 163 Isaacs yjuout snotaead wutoay AjTeuroue ut esueyo snsr9A - ainjyeroduisy ATyyuowi wesw reok-7] Woay uOT}eTASp-ATewoue Jo ydesryH - ) “S14 yquow snotaeid woiy ATeWOUe UT B8UeYD SATIESON So Q+ Gt v+ € Gy L+ le &é € v- S&S O=e Zs SSIIVWONYVY SWS3YLXS + Rie SLNIOd NOILVLS YVINOAY VSL ° yjuoul snotasid worz ATewoue ut sdueyd sATITSOg 164 Isaacs VMGAL 10} uotyeururequOS ynoT Te} pue uoTINgT44STp JFINS - g “Sta Of ited ote foes en ee 2 aaras iS eae eat : Z| Paaaocay suoojS | . Z| Payoarjoy suoyoss | (eT ay/ UG) pas "— fyDIS 40 W9aQ UO sy ssw XNOIS yO HOD] Asancd—ay -- - j ; uoijoubisag UolyDIS - xX i f Or d \ Fi % Ob /Obb Ob! /o7eZ i ieee t= : / ye XX \, Ff /' mee Ss) 61/98 Go | » os Ort AC nae "0/0 1 a By oe aie 002/008 en HSN | SS dd\ i ya 08/002! | { | Ne \ AWW a FES : SS, \ ays \ | aN : ‘ | ~ NS | Jo o@! \ nS 7 0/10: =a NZ! bl/Obz ———+ 02 /000£| — O2I/ 08S a fa) (0) se () (s) 44 ap a = HH \ 99 Leper ' (a ee Ne a ,01 | | A os \ O1/ v9 01/02! 002/0002, og¢e/oops ——- ants (a) — + — — — oO o—+>+>— — 0 | | 33 : aa aa | I) | | | | \ a | =x | 5! 3 | 5 : 5 a a : =| 165 Isaacs Ue IeUIEIED - 6 °3I4q 166 Isaacs setouenbery peyoetes ye (yyUOUT/aToAO [ yIptIm) pueq rod ASzouq - OT “SIA L-0 €:0 €-0 €0 €0 €-0 €:0 aoua1ayorD $10:0 940-0 110-0 080-0 $80:0 SEI-:0 160-0 (2D.) W OST 18 sINjeIOdWID €>S I-ZEE 0-1 10 $-602 $0 L-S8 — (,Wd) sapyy wntqimby OOc ~=EGT 8-1 It OO-T L6:0 €6:0 (Aep-Jad-sa]dA9) salauanbasy ze ZA punoisysegq sty W 'O 167 BCF CHARTS Isaacs Fig. 11 - Monthly mean temperatures (°F) from SIO moorings and Bureau of Commercial Fisheries sea surface temperature charts. © - BCF sea surface temperature charts, []- weather station November. 168 Isaacs *sdtys royzeem - g ‘AN “SuUo0m]e3s Fo Sieqsnyo = - = @] “star “SUOTIeIS PeLOOUL TOF SuUOT}eDOT[ aTqIssog 169 Isaacs Fig. 13 - Possible locations for moored stations 170 Energy Spectra of the Sea from Photographs Denzil Stilwell Naval Research Laboratory Washington, D.C. 20390 BACKGROUND The present work was undertaken to provide an independent measure of the surface conditions of the ocean in order to quantitatively correlate radar cross-sections with sea state. Previous attempts to define the sea state have been largely qualitative definitions, dependent not on direct surface measure- ments but on wind velocity and an observer's eye. Quantitative measurements have been made in such efforts as the SWOP program and glitter analysis, but these are not readily available techniques for the routine measure of sea clutter. Stereophotographic measurements have been attempted to supplement the SWOP data but the difficulty in its application at the very high range of the ocean spectrum (one meter wavelengths to millimeter waves) pre- eludes it as a convenient tool. Normal oceanographic devices offer little promise of obtaining the huge amounts of information re- quired to specify accurately the energy spectrum of a real sea, especially in the high frequency region, because of the tremendous difficulties involved in designing accurate measurements with the resolution required. Most techniques designed to obtain spectral information have one common feature, recognition of photographic recording as the only presently practical technique capable of providing a sufficiently large sample of information adequate to determine a spectrum with precision. The technique of obtaining Fourier transforms by optical techniques? is particularly applicable to the present problem since the basic data are the optical intensities which expose the photo- graph and the analysis is performed on the light amplitudes (which are simply related to the intensities). Optical Fourier trans- forms arise because the basic transfer function for lens has the form of a Fourier transform and when used with monochromatic, collimated light generates a light distribution which is simply related to the desired spectrum®’. Figure 1 illustrates the 171 Stilwell t simplest outline of an optical computer. The use of a laser is not absolutely necessary but the light intensities available from other sources are too low to be convenient. The laser beam diameter is enlarged after traversing the telescopic lens system to a size suitable for the data format, i.e., the negative size. The colli- mated light impinges on the transparency placed in the front focal plane (input plane) and is transformed by the lens into the output plane. The light amplitude in the output plane is proportional to the Fourier transform of the light amplitude in the input plane. The basic problem then is to determine the relationship of the light amplitude after passing through the optical system in terms of the information on the sea as recorded on the scene photo- graph. To accomplish this, it is necessary to establish the correspondence between a sea parameter and the optical density on the film, determine the exact form of the transform, and to relate the recorded optical density of the transformed information to the value of the sea spectrum. In what follows the first order theory for the Fourier components analyzed optically will be presented and the conditions under which the technique is valid will be indicated. SCENE PHOTOGRAPH REQUIREMENTS The initial problem in the photographic analysis is to determine the relationship between the optical density at a point on the negative with some parameter of the surface point it represents. Figure 2 is a plot of the characteristic curve of photographic emulsions for which the linear range has an equation of the form D = 97 lee K-a > (aL) where D is the optical density of the developed negative, y is the slope of the straight line part of the curve, K is a constant relating to the sensitivity of the film, Tt is the exposure time, and u is the power density incident on the film. Optical density is defined by the equation A -D u = Uy LO (2) where u is the light power density transmitted through the film with optical density D and Uy is the incident intensity. From these equations it is obvious that a knowledge of the light in- tensity leaving a point on the surface in a direction toward the camera is sufficient to determine the resulting optical density. The camera illumination (observer direction in Figure 3) is due largely to the light reflected from the surface since light leaving the water is normally of much lower intensity. Figure } is a plot of the reflectivity of water with angle. Visualizing 172 Stilwell a sinusoidal modulation of the normal angle to the wave surface, this Figure indicates that the resultant reflectivity variation would be a minimum for an observer angle of 90°, or vertical incidence. Grazing incidence would give the greatest sensitivity but the scene would be highly distorted and much of the wave surface would be obscured by the peaks of the waves. The intermediate re- gion, that near Brewster's angle, can be used. For this region a polarizing filter is required since the vertical component would not yield a unique reflectivity with wave angle and would therefore introduce extraneous spectral components. Reflectivity does not completely define the light directed toward the camera since the sky itself may have variations of brightness. With sky luminance entering the analysis the problem loses definiteness but certain requirements can be deduced. That part of the sky from which light is reflected into the camera must have monotonic variation of luminance if there is to be a one-to- One mapping of the surface normal angle onto the film. ‘Thus, the sky must be either clear or uniformly overcast for the technique to work. Denoting the product of luminance and reflectivity by g one can expand this function in terms of normal angle variations in the form g(x,y) = eg [1 + e/a, @ (xy) +-- -] (3) Noting that the power density in equation (1) is proportional to g, one can obtain aD gt Oe i ey = mM dep (do) YG (4) The B term in equation (4) will relate to the average density of the scene negative and can be estimated by a direct measurement of this density. The scene photograph under these’ conditions would consist of an increasing density in the direction away from the observer which by knowing the camera field of view would allow a measurement of density variation to yield an average of the angular gradient of density from D ee (5) To '= Ko This quantity expresses the sensitivity of the photograph density variations to the angular disturbances of the surface. It should be pointed out that waves traveling in a direction other than toward or away from the observer will not be transferred onto the Stilwell film with a density variation as large as the angular excursion would indicate using the value of D_ obtained above. This results because only the projection of the normal angle excursions in the vertical plane directly away from the observer is effective in causing reflectivity variation of light into the camera. The sensitivity must be reduced by some function of the azimuth angle of the waves » to be called p(y). The 2(y) function can be calculated by considering the projection of arc length for a skewed wave onto a plane parallel to camera direction. The approximate equation is p?(y) = Sin® 6 + Cos? 6 Cos? 4 where 6 is the camera depression angle which for § = 45° is p(y) = E (3 + Cos 2 4). OPTICAL ANALYSIS The scene transparency with the properties previously described is inserted into the input plane of the optical computer and illuminated with a collimated beam of laser light. Modifying equation (2) to a form involving light amplitude gives -D/2 (6) Substituting in equation (1) with the power density being the g function of equation (4), one obtains BB £& @€ AO fo) a =a, | fag) 27] (2+ 8 nly)e+.. 3% (7) The quantity ay (the incident laser light amplitude) can be written as proportional to the square root of the laser power density = 6 /G_)o fe) The term in brackets can be written as i oma (8) where the bar denotes an average over the area of the photo illu- minated by the laser beam. The subscript 1 refers to the scene photograph to distinguish similar terms referring to different photographs occurring later in the analysis. If a small angle or density gradient assumption is made the term in parenthesis can be approximated by 174 Stilwell 1-2 E ply) = 1-3 Dd, ly) ¢ (9) O The light amplitude can be seen to consist of a constant term plus a term proportional to the normal angle of the ocean wave. The constant amplitude term will transform into a finite aperture equivalent of a delta function and will contribute to the transform light intensity only near the region of zero spatial frequency and can be excluded from further consideration. The Fourier transform is then performed on the normal angle as is desired. The higher order terms ignored in writing equation (9) are less than a tenth the first order term for usual values of D with wave angles up to about 40 degrees. The resultant error in the final spectrum will then be less than 5%. Taking the light amplitude in the output plane as propor- tional to the Fourier transform one can write (10) where the constant q relates to the proportionality of the trans- form (the light amplitude distribution being proportional to the Fourier transform) and the constant ¢ relates to the proportionality of power density and light amplitude. Un IS} (Glale) abiahe(sualsjalioyy jyplaul@lal will expose the film placed in the output plane to record the spectra. q can be evaluated by knowing a transform pair for the Optical system. A convenient pair is just the gaussian shape of the laser amplitude with off-axis position which transforms into a@ gaussian shape. Then invoking the requirement that the power flow in the input and output planes must be identical, one obtains Cnt © 2m 2 2 2 2 a ] 4 #0 exp {- Bay! rdrd@ = | | gq? (2x) <= ot fo) 62 oO S exp {- y®e*} edede (11) where 2 = (i) e (12) and then a = (oy) = 175 Stilwell Thus q is identical to the scale factor in equation (12) relating a sinusoidal spatial wave of spacing & in the scene plane to its focal point position in the transform plane as derivable from Huygens Principle. The power density of equation (10) exposes the film placed in the transform plane for which the equation (using an alternate form of equation (1) is ich? = i #¥(a) Tar (1) where A is the ratio of D to y and the subscript 2 refers to the transform photograph (the second photo involved in the analysis). Thus equations (7), (8), (9), and (14) yiela _ 4 a 10ke aay eG a! Oe where the dc term has been suppressed. (The cross product terms in the squaring operation is weighed so heavily to the delta function that they can be ignored.) This equation can be simplified by noting the term K,upTz is equivalent to a term like 108 in which A, is the optical density arising by exposing film in the scene plane to the laser intensity uj. This film must be developed along with the transform in order that the constants K be the same. It is not always convenient to expose the third film for the same time used for the transform photo but the identity aL 1082 = 10 Yols (16) Ug T3 allows a different exposure time and laser intensity to be used. Thus the expression for the square of the Fourier transform is just ao = 4(LA)? Ug 73 10 -Ls3 +As S 1 Dy p? UgTs ( 7) SPECTRUM The photographic spectrum is obtained from equation (27) by an integration over an elemental area in the transform space as is indicated by 176 Stilwell pe oe q~ J | @ axay (18) Since the finite aperture utilized in the scene plane limits the spatial frequency resolution, an elemental area in the transform plane (reciprocally related to the gaussian beam cross section in the scene plane) has constant g®. Denoting that area by Tk)e Fe ae ata (19) The integral is trivial and one obtains the photographie spectrum c- Sg (20) The constant & expresses that factor of the gaussian half width o which is effective in the transform process. Further analysis is required to obtain the actual value of = but it is of the order of /2 which is the equivalent square pulsewidth of a gaussian function. The expression thus far derived has not involved the size of the sea photographed, only the photographic density variation. It is possible to envision a situation in which a sinusoidal wave of some fixed angular excursion propagating on the surface could be photographed and result in a photo exactly identical to a wave of different wavelength with the same angular amplitude but photographed from a different height. Using (Et) = (2x) | 1i,t) NG Het) oak x (21) from Kinsman® one can determine that the spectrum is proportional to the area analyzed. Thus the ratio of two spectra is propor- tional to the ratio of the respective areas. Extending this analogy to the photographic spectrum one can write that the sea spectrum is proportional to the ratio of the sea area to photo- graph area multiplied by the photographic spectrum. Or then @ = ape ft (22) where 8 is linear reduction factor from the sea to photographic 177 Stilwell length. The constant @ can be determined by computing the spectrum for a known situation (not necessarily realizable) by equation (21) (modified to have dimensions of wave angle rather than wave height) and also by the photographic technique. The ratio of the two solutions yields -2 o = (2x iia 10) (23) The solution for the spectrum is manifest in either equation (2) oon 5) AntD, -A3 2 oof iC, ~ £2) ee (2h) a “lia, © Dep (y) Ug Uy Ge = As 2 { ts (25) A, is the area of the sea analyzed and the other terms are previously defined. EXPERIMENTAL RESULTS Figure 5 is a composite of the scene photograph, the transform photograph, and an isodensitometer trace of the transform density contours. The photograph was taken from the South Capitol Street Bridge crossing the Anacostia River in D.C. from a height of about 30' in a wind of about 4-5 knots. Since for a given analysis equation (24) is simply a constant times the base ten power of the optical density (letting gamma be one for discussion purposes) the density contours correspond to spectral contours. The radial distance from the de term is proportional to the wave number of the water wavelength. The spectral information is what is expected for the low wind velocity involved when the analyzed spectral region spans the minimum in the wave velocity-wavelength curve for water waves. The high frequency spectral peak is about 1.5 cm wavelength and the corresponding "gravity wave" is about 4 em long, as indicated by the peak in the spectrum. It may be noted that the curves of velocity vs wavelength would not give the above values for paired frequencies. The surface tension on the polluted river is probably significantly larger than that for pure water which would shift the null in the velocity curve to longer wavelengths so the measurement is not unreasonable. Figure 6 is a similar composite of data taken from Wilson Bridge across the Potomac River from a height of 50'. The wind was about 12-16 mph, blowing almost directly into the camera from the north. The bench parameters for this analysis resulted in a large magnification of the region about the de term in order to 178 Stilwell analyze wavelengths approaching the size of the camera field of view. The high frequency peak was beyond the camera resolution and does not show. The low frequency spectral peaking is quite evident in the region around 20 em water wavelength. CONCLUSIONS The small wave theory developed here should allow a useful estimation of the spectrum of the ocean under the conditions of clear or overcast sky, camera depression angle about he} use of a polarizing filter, multiple photographs to reveal directional spectra and to obtain spectral smoothing, etc. The technique will have to be compared with the spectra obtained by present installa- tions but should allow an extension of measurements into the capillary wave region. It should be useful in studying the build- up of wave systems (the transient spectral changes) and allow the study of the turbulent structure and energy transfer properties of the air-sea boundary layer. Although the technique is not applicable in all circumstances for which sea clutter data is re- quired, it should allow a tabulation of high frequency spectral characteristics under a wide range of local wind conditions which would enable non-spectral measurements to yield a good estimate of the actual sea conditions. REFERENCES 1. C.A. Taylor and H. Lipson, "Optical Transforms," Cornell University Press, Ithaca, N.Y., 1964. ae L.J. Cutrona, E.N. Leith, C.J. Palermo, and L.J. Porcello, "Optical Processing and Filtering Systems," IRE Trans. on Info Theory, June 1960. 3. B. Kinsman, "Wind Waves: Their Generation and Propagation on the Ocean Surface," Prentice-Hall, Englewood Cliffs, N.J., 1965. 179 Stilwell Figure 1 Log E Figure 2 180 Stilwell VERTICAL HORIZONTAL aoe SNE PG Figure 3 | I HA, 05 = Ca Oe Figure 4 181 Stilwell Figure 5 182 Stilwell 25 WAVE LENGTH [em] rs loo Figure 6 183 Numerical Prediction of Geostrophic Flow Derived from Sea-Surface Temperature James G. Welsh The Travelers Research Center, Inc. Hartford, Connecticut INTRODUCTION The author has previously described the application of the equivalent barotropic model to an ocean of variable depth. This paper extends the model to allow sea-surface temperature prediction and to incorporate the irrotational flow associated with horizontal divergence. A simple approach to deriving a geostrophic flow field from sea-surface temperature fields is described; the inclusion of irrotational flow into the model and the nature of the derived velocity potential for the Gulf Stream are discussed, and experi- mental forecasts of sea-surface temperature for the western North Atlantic Ocean are shown. GEOSTROPHIC FLOW EXPRESSED IN TERMS OF SEA-SURFACE TEMPERATURE Geostrophic flow in the ocean can be computed when both the vertical distribution of horizontal density gradient and the actual flow at some reference level are known. For numerical prediction, these must be known for the area in which prediction is carried out. Unfortunately, it is only for sea-surface temperature that there may be adequate observations to permit a meaningful depiction of the horizontal distribution. This leaves salinity, the flow at a refer- ence level, and the vertical variation of temperature gradient to be described. It will be assumed that salinity is constant, and the flow vanishes at the bottom. The vertical variation of the horizontal temperature gradient VI will be prescribed by n Wiz) «= e,(a - 2) (aly) 184 Welsh where d is the depth of the ocean, T, is the surface temperature, and n is a parameter of the model. Note that this prescribes a temperature gradient which has the same direction at all depths and, together with its n- 1 derivatives, vanishes at the ocean floor, for n,> 0. The equation for the vertical shear of the geostrophic flow V can be written — Viz) = ak x VI(z) (2) where f is the coriolis parameter and a is the linear coeffi- cient of thermal expansion for sea water. Now Eq. (1) can be sub- stituted into Eq. (2) and the result integrated in the vertical to yield oy ” agd 5 n+l “y V(z) = Ga) F - :) k * VT, (3) where the assumption of zero bottom flow has been used. From Eq. (3), derive the further parameters d reat eS lan ee Ce ie i | Hees “EGan) Gm) ten (4) =e n+1 a@y os S22 @eala-Z) (5) 7 d 5 2 eae aS 2 5 hoe 2e ee 2 | tac dz ——— (6) 0 and note that Ay = A(z = 0) = mdb D (7) Furthermore, it follows from Eq. (4) that the stream function v of the vertically averaged flow WV ie given by = agh, u £ , (n+1) (nt2) ts ion Z (8) where agh, is £ , (nt1) (nt2) (9) is the conversion factor, and h) and f, are suitable mean values for d and f , respectively. 185 Welsh Figure 1 shows the profile A(z) for a selection of values of the parameter n , and Table 1 lists values for A) , AZ sandy whey, as Jee ae etig 3 and hy , have been assigned the values 1.8 x TO ant Chama 980 cm sec~* , 9 x 10°° sec! and 4000 m , respectively. Table 1 8 9 2 3 4 5 6 7 8 9 10 11 6g} Ibo Bo2D 2678 Bo2I Sal ho2 Gol D620. Sod 5-0 1250 6.00 3560 2640 2570 1530 1500 0580 O05 * k x 10>2@m2 sees! °C) To select an appropriate value for n , we shall be guided by previous results [1]. These showed values of the vertically inte- grated stream function at the inflow boundary ranging from zero to 3 x 10* m2 sec-! . For comparison, the sea—surface temperature for March, 1961 at the inflow boundary ranges from 5 to 20 °C which would require a value of 2000 for the conversion factor k , corre- sponding to n in the range 4 to 5. After some trial and error tests, a value of 1000 for k was applied to the 31-day mean sea- surface temperature for March, 1960 to give a volume transport com- parable with published values of the maximum transport, which is around 80 million m*? sec-! . Warren [2] and others have ques- tioned whether the actual maximum transport is not somewhat higher than this figure. THE INCORPORATION OF IRROTATIONAL FLOW The generation of relative vorticity by horizontal divergence is significant for the numerical ocean-prediction model. In the previous report [1], it was shown that the irrotational flow associ- ated with horizontal divergence can be derived from a velocity potential A given by vV2A = -— (x Wt VA)*Vh 2 -~ 10, in)“ (10) 0 0 where J is the Jacobian derivative. The prediction equation can then be written Af.) a7 a EAT fh [ve - 2] SY (Ex ow + al ev[areey +e - | = 05 Gal) 0 in which the irrotational flow VA is also used to advect the 186 Welsh <7 n7® ares ave Fig. 1 - Isopleths of A(z) for n = 0 ton = 10 187 Welsh "potential vorticity," [a2v2y + f - f)h/h)] . To implement the irro- tational flow, the velocity potential 2 is computed at each time step by solving (10). Then (11) is solved for the tendency d/dt . In order to solve for i , it is necessary to describe i on the peripheral boundary of the prediction grid. Setting A to zero on the boundary is not acceptable at the inflow to the prediction area, so an experiment had to be carried out to determine reasonable values. An area (see Fig. 2) was selected with the prediction area inflow well surrounded. The 3l-day mean analysis for March, 1960 was used for W , and the velocity potential \ was computed from (10) using zero values on the boundary. The velocity potential is shown in Fig. 3; the units have been converted to degrees Celsius. Note that the rotational flow is directed along the surface iso- therms, and that the irrotational flow is directed normal to the A isotherms. Consider the velocity potential of Fig. 3. There appears to be a boundary-value distortion on the lower boundary, and perhaps a slight amount on the upper boundary, but the existence of a maximum near 36N, 74W is definite. To the southwest, beyond the computa- tional area, the velocity potential probably is shaped like a trough which contains the Gulf Stream. Although the magnitude of the irrotational flow is an order-of-magnitude less than the rota- tional flow, its direction, more-or-less normal to the Gulf Stream near Hatteras, allows its effect to be significant for surpressing meander growth and stabilizing the Gulf Stream position. SEA-SURFACE TEMPERATURE PREDICTION* This section describes experimental sea-surface temperature (SST) predictions made with the model. The first two experiments were designed to test the flow from the SST conversion scheme. The third experiment is a series of predictions from 5-day SST analysis. The initial field for the first two experiments was computed from 31 days of ship reports for March, 1960. This long-term analysis is smoothed some but still shows relatively small-scale features, which may or may not be real. The first experiment uses a conversion factor of 1000 m* sec7! °C , which gives a reasonable mass transport for the Gulf Stream. The second experiment uses 3000 m2 sec~! °c » which corresponds to a mass transport much higher than generally suggested. The values for A and A, were taken directly from Table 1; they are shown together with the flow con- version factors in Table 2. For both experiments, predictions have been carried out to 30 days after the initial map. * The quantity predicted is = kT, where T, is the sea-surface temperature. Because k is taken to be a constant, it is con- venient to label wt in °C and to refer to it as sea-surface temperature. 188 Welsh S --7 ma-- Ku000! - ~~ Fig. 2 - SST (°C) map from which velocity potential is computed 189 Experiment 1 Experiment 2 To appreciate the results of these experiments, I have plotted the initial and successive 5-day predicted positions of the 16°C isotherm for each experiment in Figs. 4 (Experiment 1) and 5(Experi- ment 2). A comparison of these two figures shows Experiment 2, with its greater transports, to produce greater development... Although our primary purpose here is only to show the general envelope of predictions, continuity between 5-day positions can be discerned in the vicinity of 65°W, where Experiment 2 clearly predicts faster movements and greater amplitude development. The indicated wave speeds in this area are about 7 cm sec-! for Experiment 1 (Fig. 4) and 18 cm sec’! for Experiment 2 (Fig. 5). Warren has suggested a wave speed of 5 cm sec-! » with which Experiment 1 more closely agrees; as already noted, Experiment 1 gives volume transport in close agreement with accepted values. For the third experiment, six successive, non-overlapping, 5-day SST analysis were computed for the period March 1-30, 1960. The analysis should be valid for the middle of each time period, 16@65 2655 Wod5 Wod5 tWod5 22555 Gel Z27/5D Gays sinEO ene imeimesa From each analysis, a prediction was made out to day 32.5 (April 1, 1960), with prediction outputs every 5 days after the analysis time. For this experiment, the conversion factor 1500 m see-> °@-! AZ = 3.5 » and A, = 7.5 were used. Unfortunately, the six analysis, are themselves disappointing, for they show poor map-to-map continuity. Moreover, it is the ex- perience of the author that successive, non-overlapping periods of ship reports cannot generally be analyzed with good continuity be- tween maps. It is not surprising, then, that the predictions in this experiment did not bear much resemblence to the verifying analysis. The analyses were simply inadequate. For all three experiments, a computation grid with about 60 km between gridpoints and time steps of 6 hours were used. The value for f£,)/h) on the right-hand side of Eq. (11) is 2.5 x 107° m} sec , which corresponds to f, = 1 x 105° seeg! for) h- = 4000) m ~ This represents a reduction of the effect of the bottom consider- ably greater than the reduction used in our previous report. To of the 16°C isotherm eager Gs! itions sive 5-day pos for Experiment 1, k = 1000 m Fig. 4 - Succes 193 °C isotherm oc-l of the 16 2 sec-l Fig. 5 - Successive 5-day positions for Experiment 2, k = 1000 m Welsh Early experiments with SST predictions, not implementing irro- tational flow, showed severe sensitivity of the inflow pattern to bottom weight. That SST should be more sensitive than the hypothe- tical flow field used in our previous work is probably related to the confinement of the SST gradient to a more narrow band which is also located further up the continental rise. The incorporation of convergent flow reduced the sensitivity to bottom weight and allowed more realistic development near the inflow, but the full bottom weight, f)/h, = 2.25 x 10-8 m=! see , could still not be permitted. The value 2.5 x 107? m7! sec for fy /ho seems to give relatively undisturbed flow near the inflow boundary. SUMMARY Plausible predictions of sea-surface temperature can now be made using the equivalent barotropic model--plausible in that pre- dicted wave speeds, meanders, and mass transport are close to ob- served values. The lack of meaningful synoptic SST analysis, how- ever, does not allow a definitive testing of the model. Future work must be directed toward improving the analysis of sea-surface temperature. REFERENCES larnason, G. and J.G. Welsh, 1966: Numerical flow prediction with the equivalent—barotropic model. Part IV of "Studies of techniques for the analysis and prediction of temperature in the ocean." Interim Report, under Contract N62306-1675 with the U. S. Naval Oceanographic Office. *Warren, B.A., 1963: Topographical influences on the path of the Gulf Stream. Tellus, Vol. 15, pp. 167-183. 194 THE FNWF SOUND MAP PROGRAM Captain Paul M. Wolff, U. S. Navy LCDR Peter R. Tatro, U. S. Navy LCDR Louis D, Megehee, Jr., U. S. Navy Fleet Numerical Weather Facility Monterey, Califormia ABSTRACT A high precision computer ray trace program developed at the Fleet Numerical Weather Facility in Monterey is used as a research tool to investigate the effects of temperature and salinity variations on the path of sound through the sea. In one case, the temperature profile as a function of depth is described in terms of its basic parameters: sea surface temperature, mixed layer depth, thermocline gradient, and 400 meter temperature. These are systematically varied, individually and in combinations, to determine their effect on convergence zone formation. In a second case, a highly detailed cross-section of the Gulf Stream taken by XBT's is used to demonstrate the effect of water mass and current boundaries on convergence zone propagation. INTRODUCTION A ray tracing program developed at the Fleet Numerical Weather Facility for use on the CDC 1604, 3200 and 6400 computers has been used as a research tool to investigate the effects of oceanic variability on the paths of sound through the sea. The program is a highly sophisticated one incorporating a correction for the curvature of the earth and utilizing a time step of 1/128th of a second or less. It Operates in two dimensional x - z space, and is capable of accepting any arbitrary specification of the sound velocity in x and Zz. This represents a significant improvement over earlier numerical ray trace programs, many Of which were capable of handling only linear gradients of sound velocity or sound velocity profiles which could be LOS Wolff, Tatro, and Megehee described by a particular mathematical function. THE RAY TRACE PROGRAM In practice, the inputs to the program are temperature and salinity as a function of depth for as many positions in x space as are desired. The program utilizes Wilson's equation to convert these variables to sound velocity. The only other variables which must be specified initially are the source depth and the bottom depth. The program is capable of handling reflections at the air-sea and sea-bottom interfaces and of accepting an irregular bottom profile. It incorporates several options, such as terminating a ray after the first convergence zone, or after a specified number of surface or bottom reflections. Three types of output are available from this program. Figure 1 shows the printer output showing the values for all the computed parameters at each time step. Figure 2 shows a condensed printer output which shows only the initial point, the intermediate tuming points, and the end point of a ray. In this example, the terminate at end of first convergence zone option has been utilized. Figure 3 shows a plotted output. This is plotted off-line on a CALCOMP plotter. The upper box in this figure extends in depth from the surface to 500 feet, and in range from zero to 7200 yards. It displays the direct path sound field. The lower box covers the same depth range, but the horizontal range is from 50,000 to 80,000 yards. The sound which was all strongly refracted downward in the upper picture is turned at depth and often returns to the surface at a considerable range to form a convergence zone. The :lower box displays the convergence zone. From any of these three forms of output it is possible to determine the range and width of the convergence zone, and from the printer output the depth of water required for convergence zone formation. The ray trace program has been described in detail by Ayres et al (1966). EFFECT OF SOUND VELOCITY MICROSTRUCTURE Careful in-situ direct measurements of the sound velocity as a function of depth by means of a velocimeter lowering indicate that the sound velocity profile is not a smooth curve as it is often depicted, but rather that it has a sort of perturbation superimposed on it with an average wave length of approximately 10 meters and an 196 TaTQg 23 INITIAL ANGLi = *02,0000 TIME (SE5ONIS) ANGLE(RADIANS) 0030000,0000)000 #000,0349065 0030000,007591250 #000,0347725 0090000,0154$2500 2000.0346384 0090000,02343750 0000000,03125000 0030000,03905250 0090000,04637500 0090000,03453750 0090000,0525000 0090000,07031250 0030000.07812500 0090000,03593750 0090000,09375000 0090000,10155250 0090000,10937500 0000000,11713750 0090000,1250)000 0090000,13231250 0090000,14052500 0090000,14943750 0000000,13525000 0090000.15405250 0090000,17137500 0030000,17953750 0090000,1875)000 0090000,19531250 0030000,2031 2500 0030000,21093750 0000000,21375000 0090n00,22555250 0090000,23437500 0030000,24213750 0090000,25003000 0090000,25731250 0090000,25362500 0000000,27343750 0030000,29%25000 0030000,29905250 0030000,29437500 0030000,30453750 0090000,31259000 0030000,32031250 0090000,32812500 0090000, 33593750 0090000, 34375000 0030000,33155e50 0090000,35937500 0090000,35713750 0030000.3750000 0090000.33231250 0030000, 39852500 0090000,39843750 0090000,40625000 0030000, 41405250 0030000,42197500 0090000,42953750 0030000,4375)000 0030000,44531250 0090000,45312590 Wolff, Tatro, and Megehee #000.0345040 #000.0343694 0000,0342350 2000.0341006 +000.0339664 #000.033835235 #000.0336984 #000,0335646 2000.0334310 €000,0332975 #£000.0331642 2000.0350511 #000,0328981 2000,0327653 ©000.0326327 #000.0325001 +000,0323678 *#000.0322355 #000.0321034 #000.0319715 £000.0318397 #000,0317081 +000,0315766 €000,0314453 #000.0313141 #000,0311831 #000,0310522 ©000,0309214 #000,0307908 #000.0306601 #000.0305294 €000.0303987 #000,0307680 #000.0301574 €000,0300067 #000,0298759 #000.0297453 0000,0296146 #000,0294840 #000.0293532 #000.0292226 #000,.0290919 #000.0289613 0000,02883505 #000.0286998 #000,0285692 #080,.0284384 4080,0283077 #000.0281771 #000,0280464 0000,0279157 #000.0277850 ©000,.0276544 #000.0275236 ©000.0273929 2000.0272623 INITIAL DEPTH = RANGE ( 00510,721 3 ENJO-RAY 9 BSOTTI4 REFLECTIONS 13967 INITIAL DEPTH = 14 FEET RANGE ( (S) j=) DEPTH (METERS) an oS i) o>) (=) oO Wolff, Tatro, and Megehee TEMPERATURE 6 8 10 12 14 Fig. 10 - Variation of thermocline gradient 208 16 18 Wolff, Tatro, and Megehee 12 10 (o>) THERMOCLINE GRADIENT DEG F PER 100 FEET ) aS Fig. 11 - Temperature correction for thermocline gradient 209 DEPTH (METERS) 100 200 300 400 j 500 600 700 800 Wolff, Tatro, and Megehee TEM PERATURE 6 8 10 12 14 Fig. 12 - Variation of 400 meter temperature 210 sInjerodwis, 1919 OOP 0} enp UoOTJIeTIIOO sinzverodwuisy, - ¢][ °stq (Dye) it O.+ 8+ O+ Wolff, Tatro, and Megehee S)IL 8T @ Mk th Wolff, Tatro, and Megehee completely, however, in the Mediterranean or any other area where the water conditions approach isothermal and a half-channel sound velo- city profile results. Possibly the addition of a deeper temperature parameter will allow the extension of this system to cover these cases. EFFECTS OF CURRENT BOUNDARIES Near current and water mass boundaries the horizontal gradients of temperature and salinity, and thus sound velocity, become quite large. To investigate the effects of these large local gradients, a highly detailed cross-section of the Gulf Stream taken with expendable BI's by Sippican and the Woods Hole Oceanographic Institution was digitized into 71 profiles. Figure 14 shows the cross section used, and the locations of the sources which were initially considered. Sources 1 and 2 were in a cold secondary front which extended up to about 150 meters. Sources 3 and 4 were taken on the warm side of the Main Gulf Stream wall and run through the wall. Source 4 was also run away from the wall into an area of fairly homo- geneous water. Sources 5 and 6 are on the cold side of the main wall and were also run through the wall. In all cases, a convergence zone was formed; however, they were generally broader and more diffuse than the well defined zones found in an area of horizontal homogeniety, and in many cases did not reach the surface. Figure 15 shows the convergence zone formed from source 4 running through the Gulf Stream wall. On this plot the depth scale is from 0 to 2000 feet rather than 0 to 500 as before. The convergence zone formed is a subsurface one at approximately 25 nautical miles. Figure 16 shows a run from the same source in the opposite direction. In this casea surface convergence zone is formed at 33.7 nautical miles. CONCLUSIONS Based on the Ray Trace runs which have been described here and others of a similar nature, the following conclusions have been drawn: 1. The micro-structure in the sound velocity profile is essentially self-cancelling. It does not affect the convergence zone parameters. 2. In areas of reasonable horizontal homogeniety it is possible to predict convergence zone range by empirical methods if the temperature structure is known. 3. In areas of high local horizontal gradients, convergence 2aZ Wolff, Tatro, and Megehee zone formation is highly variable and may often be a function of the direction from the source. In these areas the only reliable method of prediction is to use a computer program. BIBLIOGRAPHY PIIP, Ants T. Precision Sound Velocity Profiles in the Ocean, Volume I, Technical Report No. 3, Columbia University Geophysical Field Station, 1963. Ayres, E,, P. M. Wolff, L. Carstensen and H. C, Ayres. A Ray Tracing Program for a Digital Computer, Fleet Numerical Weather Facility Informal Manuscript, 1966. 213 Wolff, Tatro, and Megehee NSO. MQ. MSL) Hida KYW (IOHM pue weortdditg Asoz1n095) Azepunoqg juerTiNno roperqey wreedyg FIND ey} ysnory} uotjDesS LaqX UV - FI “SIT S3IWN Of _ 02 OF Oo = as YOLVYOOT 1981S ~ OO} esos ces : cL €2 1401 29 Ofh Bib bia £6 638 18 08 214 and Megehee Wolff, Tatro i ) XN ils WO = 1 N \ \ C9 ee) NN Mf \\ NY NI c} a a a st w “ul rai IL “al is % fs UN XS / & W/W) I 215 and Megehee Wolff, Tatro, ol uw JM C -E <¥ ; SS Sh & f 74 4 E-"4 Ohe 74 U3"4 e~) 33 a5 Li) Zz8 eee ns 216 Session E OCEAN ENGINEERING AND TECHNOLOGY et CONCRETE HULLS FOR UNDERSEA HABITATS JERRY D. STACHIW U. S. NAVAL CIVIL ENGINEERING LABORATORY PORT HUENEME, CALIFORNIA ABSTRACT Exploratory experiments have shown that concrete is an accept- able construction material for hulls enclosing undersea habitats at atmospheric pressure. Models of spherical concrete hulls with and without penetrations have been built and tested to destruction in simulated hydrospace. Test results indicate that positively buoyant concrete hulls of spherical shape are feasible for location to 3500 feet depth, while negatively buoyant hulls of the same shape may be able to be placed at depths to 10,000 feet. Both economical and military considerations seem to favor concrete hulls for permanent ocean bottom installations. Ai) Stachiw INTRODUCTION The conquest of hydrospace ‘requires both mobile and fixed under- water structures capable of housing instruments and men for extended periods of time. There is a long history of research on the proper- ties of materials and the design of hulls suitable for submarines; however, the research into materials and designs for static under- water hull structures is just beginning. Although many materials developed for submarine or torpedo hulls are also applicable to fixed, ocean-bottom installations, there are materials which have not received careful study because of their manifest inapplicability to high-speed, deep submergence submarines or torpedoes. One such material is concrete. The purpose of this paper is to describe several brief explo- ratory investigations into the applicability of concrete to the fabrication of structural hulls for deep submergence structures.* The scope of this series of experiments was limited to models of buoyant spherical hulls of 16 inch external diameter for 3500 feet depth cast from the same concrete mix. Variables were introduced into the study by varying the method of hydrostatic testing, as well as by incorporating into the hull different kinds of penetra- tions and inserts. BACKGROUND Concrete has been used in harbor installations for many de- cades, but it has not been used for the construction of underwater habitats. There are several reasons for this. Since concrete is not as desirable for submarine hull construction as other materials, no research was done on its properties under seawater hydrostatic pressure prior to the recent interest in fixed, ocean-floor instal- lations. Furthermore, the impetus of research has been directed towards the discovery of new materials that would give buoyancy to a deep submergence hull even at greatest depths in the ocean. The most potent argument used against concrete in the past was that buoyant concrete hulls are limited by concrete's compressive strength to depths less than 5,000 feet and therefore cannot satisfy depth requirements that may arise in the future. Thus the philosophy appears to have been that since buoyant concrete hulls were defin- itely depth limited, there was no need to conduct research on structural characteristics of concrete as a stop-gap solution to the problem of finding suitable materials for deep submergence structures. *In this paper, "deep submergence" is used to refer to depths greater than 600 feet. 220 Stachiw Recently, materials like glass and ceramics have been dis- coveredl,2 to possess such high compressive strength, modulus of elasticity, and resistance to corrosion that buoyant deep submer- gence hulls for fixed or mobile installations can be built for any depth. Unfortunately, although glass and ceramics have the potential of providing man with hulls of ultimate depth capability, the engi- neering problems to be solved are formidable, and the materials are too expensive for applications in shallow depths where their use is not mandatory. Since currently available glass and ceramic materi- als with the ultimate depth capability have been found impractical for general use because of their high costs of fabrication and yet unsolved joint design problems, the path has been cleared for other materials with limited depth capabilities, but with an attractive cost factor and ease of applicability to large structures. Such a material is concrete which is theoretically satisfactory for buoy- ant hulls with an operational depth requirement of 3,000 to 4,000 feet. In addition, engineering estimates indicate that it is more economical than other available metallic or nonmetallic materials when used in large ocean bottom habitat structures. The depth capability of 3,000 to 4,000 feet of buoyant concrete structures which can be towed to location and submerged, permits the utilization of such structures over large areas of the conti- nental borderlands. About 12% to 14% of the ocean floor can be explored and settled using concrete as the primary hull construction material for the ocean floor installations. While the portion of ocean floor that could be made accessible through the use of con- erete hulls is small in comparison to the total ocean area, in terms of importance, it is a most critical part of the total. The reason for its importance is found in the presence of large con- tinental borderlands made up of shelves, and submerged banks in zero to 3,500 foot depth range. Since these shelves and banks are generally flat to gently sloping, and since they are also in the general vicinity of land, they make ideal construction sites. The areas of primary interest to the United States are direct- ly accessible from this country without transversing ocean floors that are under another country's sovereignty. By occupying the shelves and banks adjacent to the United States in the O- to 3,500- foot depth range, the land area of the United States could be ex- tended approximately 23%. In summary, it can be stated that although there is a need to develop materials and structures with ultimate depth capability, development of materials and structures for depths to 3,500 feet is more important in terms of national defense and natural resources. Since concrete has been in many cases the most economical construc- tion material on land, it should be investigated early in the search for undersea construction materials. It may also turn out to be one of the most economical materials for ocean-floor con- struction on continental shelves and submarine banks in the 0 to 3,500-foot depth range. 221 Stachiw APPLICABILITY OF CONCRETE TO OCEAN BOTTOM HABITATS There are several very good reasons why concrete will find application for the construction of ocean bottom habitat foundations and pressure hulls containing atmospheric shirt-sleeve environment. The major reasons are low cost of material, ease of forming double curvature shells, strength to weight ratio (Figure 1) equivalent to steels with a 45,000 psi yield point, and excellent resistance to corrosion. Other reasons also important, but considered minor in respect to the previously enumerated ones, are high elastic stability eliminating the need for rib stiffeners in spherical hab- itats for depths beyond those 100 feet (Figure 2), excellent blend- ing in with the ocean bottom making it difficult to detect the habitat by hostile personnel with standard submarine detection gear, and excellent resistance to underwater explosions or impacts created by hostile forces. The thick walled concrete hulls for ocean bottom habitats make it relatively easy to incorporate window, hatch, and feed- through penetration flanges without additional thickening of con- erete wall around the penetrations. The low heat and sound con- ductivity of concrete make it unnecessary also to insulate the interior of the structure against heat losses and noise emission which is helpful in detecting the habitat by hostile personnel. Properly formulated concrete serves also as an excellent radiation shield for nuclear power generators with which future ocean bottom habitats will be equipped. The two drawbacks that concrete possesses is its permeability to sea water and tensile strength of less than 500 psi. These drawbacks can be overcome by taking them into consideration during the design of the habitat hull, and by the use of proper steel reinforcements and waterproof coatings during the fabrication process. Although concrete will be also used in the construction of habitat foundations and columns supporting the pressure hull, all of the discussion in this paper and subsequent experimental work has been devoted only to pressure hulls, as they represent a more demanding application for concrete. EXPERIMENT DESIGN Literature search failed to disclose any previous experimental work with concrete pressure hulls under external hydrostatic loading simulating deep ocean environment. Therefore it was decided first to conduct an exploratory investigation into the use of concrete hulls for deep ocean pressure environment to determine the avenues along which it would be most profitable to direct future studies. Many avenues of investigation are open in such an exploratory study. Not only may different hull shapes be selected, but also the composition of concrete mix, the thickness of the walls and types of joints. In addition, once the hull shape has been select- 222 Stachiw ed, it can be tested for different properties, depending on the requirements of the study. Since so many alternatives were pos- sible the approach was chosen by which only buoyant concrete hulls of maximized pressure resistant shape of the simplest construction were to be considered first. The pressure hull shape chosen for the experimental study was a sphere as it represents the optimum pressure resistant hull. The spherical hull is also desirable for its inherent uniform dis- tribution of stresses. Because of this uniformity of stress dis- tribution, the strains measured at any point of the sphere's sur- face can be considered representative of the strains on the sphere. Knowledge of the maximum compressive strain found in simple spher- ical concrete hulls is extremely useful in the evaluation of future concrete hull models where the presence of inserts and penetrations will create stress risers that may lower the critical pressure of the hull. The spherical shape is also advantageous for the determination of concrete's permeability under different levels of hydrostatic pressure. Permeability of concrete is probably related to stress level, therefore uniformity of stress in the sphere eliminates those side effects that are associated with the nonuniform distri- bution of stresses. A spherical hull also eliminated any anomalies caused by edge effects of the test sample, as would be found, for example, in a flat specimen mounted in some sort of a flange. Furthermore, since in a spherical hull there is also continuity of curvature, a reasonable assumption can be made that the permeability of water through the walls of the sphere will be uniform through- out, and thus only the level of water in the sphere must be known in order to determine a nominal rate of permeability through the given concrete mix. The actual dimensions chosen for the concrete hull models were 16-inch outside diameter and 14-inch inside diameter. The outside dimension was controlled by the inside diameter of the largest vessel available at NCEL, while the inside dimension of the con- erete hull was based on the requirement that the resultant concrete hull possess at most a 0.75 weight/displacement ratio. This weight to displacement ratio was considered to be the highest allowable that would permit -the fully equipped concrete habitat hull to be either positively, or at worst, neutrally buoyant. SCOPE OF INVESTIGATION The study of spherical concrete hulls was limited, to-date, to three major phases of experimental investigation. PHASE I - Investigation encompassed the testing to destruction of twelve identical spherical concrete hulls of 16-inch outside and 14-inch inside diameters without penetrations (Figure 3): Six of the models were waterproofed and six were bare. PHASE II - Investigation centered around the testing to de- struction under hydrostatic pressure of six 16-inch external and 223 Stachiw 14-inch internal diameter concrete spherical hull models with pene- trations (Figure 4) closed by inserts (Figure 5) of different ridi- gities. Only two sizes of inserts, and three kinds of insert ma- terials were experimentally evaluated. All of the models were waterproofed prior to implosion testing in simulated hydrospace facility. PHASE III - Investigation concerned itself with the design, fabrication, and testing of two concrete habitat models (Figure 6) with 16-inch external and 14-inch internal diameters. The models were equipped with operational windows and wire feed-throughs located inside penetrations of the concrete sphere (Figure 4) reinforced by annular penetration flanges (Figure 7). OBJECTIVES AND PROCEDURES OF THE INVESTIGATION PHASE I - The objective of the tests was to touch upon as many facets of concrete hull's behavior under hydrostatic pressure as possible, rather than research any one of them exhaustively. Thus the hydrostatic tests on the simple concrete spheres were employed to explore ultimate compressive strength of concrete under short-term and long-term loading, the latter at hydrostatic pressures approaching the critical pressure. Experiments were also conducted to investigate the leakage of water through unpro- tected concrete at different hydrostatic pressure levels. Because of the exploratory nature of these tests, only one to three spheres were tested in each type of experiment. Experimental data from such a small number of test samples are considered to be indicators of the general level of magnitude of the parameters studied, but not conclusive and final evidence of these parameters. Once the general magnitude of the parameters investigated is known, an accurate plan can be drawn up for future experiments to more thoroughly evaluate and define the physical and mechanical pro- perties of concrete under the external hydrostatic pressure of seawater. PHASE II - The testing to destruction of the spherical hull models had as its objective a quantitative evaluation of the relationship between the size and rigidity of the penetration insert, and the critical pressure of the whole concrete hull as- sembly under hydrostatic pressure. The concrete spherical hull models with solid penetration inserts of different rigidity had the same dimensions and were cast from the same mix as the models without penetrations. Since it is known that the stress concen- tration around a penetration in the hull is to a large degree dependent on the size and on the mismatch between the rigidity of the penetration insert and that of the hull material, two sizes of penetrations and three types of insert materials were selected that represented a wide range of rigidity properties. The two selected sizes of model penetration inserts were considered to be representative of penetration inserts required for full size spherical structure. The 32°930' size insert simulated a penetra- 224: Stachiw tion in the hull required for man-sized hatches or windows, while the 8° size insert simulated an electrical wiring or hydraulic piping feed-through on an underwater hull structure of ten to twenty feet in diameter with personnel transfer capability. The most rigid inserts selected were made of steel, while the least rigid inserts were made from polyvinyl chloride plastic; other inserts used were made from aluminum. During the hydrostatic test- ing to destruction of the insert-equipped models, strains were measured around the inserts and compared to strains existing in the same sphere away from the penetration inserts. In such a manner some quantitative measure of the stress concentration factors produced by inserts of different rigidities could be obtained. The comparison of the critical pressures of insert equipped spherical hulls with the critical pressures of identical hull without any penetrations would also be indicative of the effect that penetra- tion inserts have on the overall strength of the spherical concrete pressure hull. PHASE III - The design, fabrication, and testing of the spherical habitat model had as its objective proving the feasibil- ity of concrete pressure hulls with usable windows, hatches, and wire feed-throughs for 3500 foot depth service. This concrete habitat model could be considered a typical example of first gener- ation concrete habitats for ocean bottom location. The concrete hull model dimensions, concrete mix composition, and method of casting was selected to be the same as in the previous phases of concrete spherical hull feasibility study. 2 In this manner, the critical pressure of the model with penetrations reinforced by flanges could be directly compared to the critical pressure of models without penetrations. The difference between the critical pressure of the working model and of the concrete spheres without penetrations would serve as a quantitative indicator of hull strength decrease due to use of the particular type of window, hatch, and feed-through flange designs. The pressure hull for the ocean bottom habitat was conceived as a monocoque concrete sphere resting on an aluminum cradle supported by three pad equipped legs. Three large window assem- blies placed around the circumference, and one located at the bottom of the sphere would permit television or photographic ca- meras to observe and record the behavior of ocean floor, hydrospace, and its inhabitants. To make the habitat adaptable to different missions, it could be selectively equipped with an array of spe- cialized subassemblies, fitting into typical large window penetra- tions. Such subassemblies, in the form of windows (Figure 8), a glass observation dome (Figure 9), diver transfer chamber, vehicle transfer hatch, or oceanographic instrument tower would make the basic concept of the concrete ocean bottom habitat adaptable to an almost unlimited number of mission requirements. The only re- quirement that would apply to all of them was that their mounting plates fit the penetration opening, and that the plate bearing lip matches the bearing lip on the penetration flange. In order to 225 Stachiw maintain the effect of the penetration flange rigidity constant, all insert subassemblies that fit inside the penetration flanges were designed to fit with a known clearance between the exterior taper of the insert and the interior taper of the penetration flange. The only point of contact between the penetration flange and the insert subassembly was at the O-ring sealing surface located on the penetration flange lip. FABRICATION OF CONCRETE SPHERES Concrete hemispheres were cast in a mold and subsequently cemented together with an epoxy bonding agent. Depending on the type of test, the exterior and interior surfaces were either left untreated or coated with a waterproofing material. The concrete mix used developed after 250 days a strength of 10,000 to 11,000 psi as determined by uniaxial compression testing of solid test cylinders associated with spheres. The treatment of the exterior surface depended upon the type of test for which the given sphere was intended. For the perme- ability tests, where the rate of water flow through concrete under hydrostatic pressure was under investigation, the exterior surface of the sphere was left untreated, the way it emerged from the mold. For the strain determination tests, on the other hand, where the prime objective of the test was to protect the electric strain gages from seawater, the external surface of the spheres was protected by a thin coat of epoxy resin. HYDROSTATIC TESTING INSTRUMENTATION - Two different types of instrumentation were employed on the concrete spheres. The permeability experi- ments required instrumentation designed to measure rate of per- meability, while the short- and long-term stress investigations needed only strain measuring instruments. Instrumentation for the determination of strains consisted of electric resistance strain gages attached to the concrete sphere, andan automatic strain switch and read-out unit. Two different approaches were used to measure the rate of permeability through concrete in the experimental spheres. One approach relied exclu- sively on electronic transducers and read-out equipment, while the other utilized only mechanical or hydraulic components. The elec- tronic water detector, specially designed for this study, operated on the principle that a rising water level in the sphere would markedly change the resistance between two separated rods placed inside the sphere cavity. As the water rose in the sphere, it would wet more and more of the two vertical rods, decreasing the resistance between them. This voltage change could be amplified, measured, and recorded to provide a resistance versus time record. The other approach used in the measurement of permeability rate consisted of tubing inserted into the sphere, through which accum- ulated water in the sphere's interior could be ejected at desired 226 Stachiw time intervals. TESTING PROCEDURE - The testing of the concrete spheres under hydrostatic pressure was conducted in the pressure vessel of the Deep Ocean Simulation Laboratory. The spheres were either placed in a retaining cage prior to testing or were attached to end closure (Figure 10) so that they would not float in the vessel and strike the end closure when the vessel was filled with water. The vessel was pressurized by air-operated, positive-displacement pumps that raised the pressure inside the vessel at a predetermined rate until implosion occurred. Pressure and temperature sensors located inside the pressure vessel permitted recording of these two para- meters on a strip chart recorder. Upon implosion of the concrete spheres, manifesting itself by a loud noise, the end closure was removed and the fragments of the concrete structure were inspected. (Figure 11). DISCUSSION OF TEST RESULTS PHASE I SHORT-TERM STRENGTH OF DRY CONCRETE SPHERES - The average ultimate compressive strength of the concrete spheres under hydro- static loading has been found to be approximately 48% higher than for the 3 x 6-inch solid control cylinders tested under standard conditions (Table 1). Table 1. Implosion Pressure, Calculated Maximum Stress in Dry Concrete Spheres; and Average Compressive Strength of Test Cylinders Associated With These Spheres. Sphere 8 mplosion pressure 3,050 psi} 3,200 psi |} 3,600 psi ompressive poreseu! 14,540 psi |16,350 psi on the interior of the sphere Compressive stressL/ 12,940 psi |L4,550 psi on the exterior of the sphere Average compressive 9,750 psi] 9,930 psi {11,200 psi strength of 3x6- inch dry test cylin- ders under uniaxial compression 1/ Stress calculated with Equation 1 (Figure 2). 227 Stachiw SHORT-TERM STRENGTH OF WET CONCRETE SPHERES - The average ulti- mate compressive short-term strength of concrete spheres permeated by, seawater has been found to be approximately 18% higher than the compressive strength of identical dry concrete in 3 x 6-inch solid test cylinders under uniaxial compression (Table 2). Table 2. Implosion Pressure, Calculated Maximum Stress in Wet Concrete Spheres and Average Compressive Strength for the Dry Test Cylinders. Sphere 5 Sphere 12 Implosion pressure 2,850 psi 2,750 psi 2,800 psi Compressive stress4/on 12,950 psi 12,500 psi 12,720 psi the interior of the sphere Compressive stress//on 11,550 psi 11,100 psi 11,320 psi the exterior of the sphere Average compressive 10,500 psi 11,060 psi 10,780 psi strength of 3 x 6-inch dry test cylinders= Wetting period at 1,500 psi hydrostatic pressure 1/ Calculated stress with Equation 1 (Figure 2). 2/ Under uniaxial compression. TIME DEPENDENT BUCKLING OF CONCRETE SPHERES - Long term pres- surization of wet and dry concrete spheres has shown that wet con- crete spheres are more susceptible to static fatigue than dry concrete spheres loaded to the same fraction of their short-term implosion pressure (Table 3). 228 Stachiw Table 3. Hydrostatic Pressure and Duration of Loading of Wet and Dry Concrete Spheres in Static Fatigue Test; and the Average Compressive Strength of the Correspond- ing Dry Test Cylinders Condition of concrete in sphere dry Hydrostatic pressure 2,000 psi 2,500 psi 3,000 psi Percent of their short- term critical pressure Compressive SErese~’ on the interior of the sphere To Die 89%, oie 9,089 psi |11,361 psi | 13,633 psi Duration of loading prior 10 minutes 3 days2/ to implosion 6 days Average compressive strength of 3 x 6-inch dry test cylinders 10,890 psi |10,610 psi | 11,200 psi / Stress c4lculated with Equation 1 (Figure 2). / No implosion occurred during 3-day test. L 2 PERMEABILITY OF CONCRETE SPHERES TO SEAWATER - The rate of seawater seepage into sphere at 750 psi has been measured to be approximately 2.5 milliliters per hour, while for a sphere pres- surized to 1,500 psi the rate was approximately 5 milliliters per hour. When the leakage rate is divided by the surface area of the sphere, it can be expressed as 6 x 1073 milliliters per hour per square inch of area per inch of thickness at 1500 psi hydro- static pressure. In both cases the salinity of water siphoned from the interior of the sphere was about 20% lower than the sal- inity of the pressurization medium. From this rather sparse data, it would appear that permeability of concrete to seawater under high hydrostatic pressure is quite low, and that some chemical or physical phenomena, which occurs in the concrete, results in a decrease in the salinity of the water that passes through the concrete sphere wall. DEFORMATION OF CONCRETE SPHERES UNDER LONG-TERM LOADING - The measured strains (Figures 12 and 13) showed that dry concrete on the sphere's interior has a time-dependent strain rate, which is very large immediately after load application, but which decreases with time. That the time-dependent strain is a function of both 29) Stachiw the compressive stress level as well as time was shown by the differ- ence of time-dependent strain rates measured on the exterior and interior surfaces of the sphere. The interior surface of the sphere, which was under a higher stress, showed a considerably higher time- dependent strain rate than the exterior of the sphere, which was under a lesser stress. The long-term hydrostatic loading was conducted for only 3 days, and thus it is not known how much the time-dependent strain rate decreases after loading duration of several months, or years. The data generated indicates that even at the 6,700 foot depth level to which the waterproofed concrete sphere was subjected, the time- dependent strain rate of dry concrete decreased to 0.01 microinch/ inch/minute after 3 days. This would lead one to believe that at lower stress levels, corresponding to 3,500 foot operational depth, time-dependent strain would not pose any serious engineering prob- lems for concrete spheres with a 0.0625 wall-thickness to diameter ratio. Upon depressurization, a time-dependent relaxation strain was observed whose rate decreased to a very small value after only 3 days. The difference between the strain level at the beginning of pressurization, and the strain after 3 days of relaxation at zero pressure shows that a nonrecoverable deformation of concrete in the sphere occurred. TANGENT MODULUS OF ELASTICITY UNDER SHORT-TERM LOADING - The tangent modulus of elasticity of concrete under short-term uniaxial compression (2,100 psi/minute loading rate) was found to decrease with increasing stress level. The axial strains on the exterior surface of solid dry concrete test cylinders under uniaxial com- pression show that the average tangent modulus of elasticity for concrete mix employed in the casting of spheres is 3.68 x 106 psi in the 0 to 4,500 psi stress range, but decreases rapidly at higher stress levels (Figure 14). What the magnitude of change is in the tangent modulus of elasticity under biaxial or triaxial stresses, as found in the sphere, is not known. The slope of the strain curve for the interior of the spheres shows, however, positively that a decrease in the tangent modulus of elasticity does take place. This makes it necessary to treat E; in equation (1) as a variable, and not as a constant. Since curves for E; of conerete under dif- ferent biaxial and triaxial stress levels do not exist at the pre- sent time, E; as determined under uniaxial compression must be used in the meantime. Since E; under uniaxial compression appears to be larger than E~ under biaxial and triaxial stress combinations, use of Et under uniaxial loading is a conservative assumption. PHASE II EFFECT OF PENETRATION INSERT RIGIDITY ON CRITICAL PRESSURE OF SPHERE - It was found that there is no significant difference be- tween critical pressures of concrete hull models with solid pene- tration inserts (Spheres No. 15, 16, and 17) and model without penetration (Spheres No. 18) as long as the rigidity of the insert was equal to, or larger than, the rigidity of the concrete hull 230 model. When the rigidity of the penetration insert was considerably lower than the rigidity of concrete (0.5 x 106 psi for polyvinyl chloride versus 3.65 x 106 for concrete) the sphere equipped with such inserts (Sphere No. 17) imploded at a significantly lower pressure than the sphere without penetrations (Table 4). Table 4. Type of Inserts Solid steel inserts Solid alu- minum in- serts Solid polyvinyl chloride inserts Short term sion test; psi/minute surization Short term sion test; psi/minute surization Short term sion test; psi/minute surization Short term sion test; psi/minute surization Implosion Pressures of Concrete Spheres With Solid Penetration Inserts. Strength Implosion of Pressure Concrete* 3485 psi 330 (11,205 psi days 3400 psi 335 11,165 psi days 2675 psi 330 11,150 psi days 3375 psi 320 11,480 psi days * Strength of concrete was determined by subjecting 3 x 6 inch test cylinders of the same mix and age as the associated concrete sphere to compression testing in an uniaxial test machine. The compressive strength shown is the average of 18 test cylinders loaded to destruction at 2100 psi/minute rate. THE EFFECT OF PENETRATION INSERT RIGIDITY ON STRAINS AND STRESSES IN THE SPHERE - The strains at the small penetrations in the concrete were not significantly different from strains measured at locations in the sphere away from penetrations. The strains at the large penetrations in the concrete, however, were significantly higher than strains measured at locations in the sphere away from penetrations only in the sphere No. 17 containing plastic inserts The strains measured there one half of an inch away from the edge of the penetration with a plastic insert were approximately 40 percent higher than the strains in the same sphere not in close proximity to the penetrations. When the strains on the in the penetrations. Stachiw interior of the concrete hull away from penetrations are translated into stresses, the measured stresses at locations remote from the penetrations are found to be of the same magnitude as stresses de- rived analytically for the interior of a thick sphere. The strains on the interior surface of the solid inserts varied inversely with the modulus of elasticity of the particular insert material (Figure. 15). Stresses, calculated on the basis of the strains in insert materials, were higher than in concrete for materi- als with modulus of elasticity higher than of conerete. Conversely, the stresses in inserts with modulus of elasticity less than of concrete, were less than in concrete itself (Figure 16). PHASE III CRITICAL PRESSURE OF OPERATIONAL HABITAT MODELS - The critical pressures of the two identical habitat models were approximately the same as of the concrete sphere with identical dimensions without any penetrations (Table 5) even though one of the models was subject- ed prior to destructive testing to its operational depth for 200 hours. Table 5. Implosion Pressures of Ocean Bottom Habitat Models Age of | Strength Implosion Pressure Held at 1500 psi 10,840 psi for 8 days, then pressurized to implosion at 100 psi/min rate Short term implo- 9,640 psi sion test; 100 psi/minute pres- surization rate Short term implo- 11,480 psi sion test; 100 psi/minute pres- surization rate STRAINS AND STRESSES IN HABITAT MODELS - The strains and stresses in the interior of the concrete habitat models were not significantly different from those found in aphence with solid steel (E = 27 x 106) or aluminum (E = 10 x 10° psi) inserts, even though the rigidity of the annular steel penetration flanges was approximately the same as of a rigid insert with a 5 x 106 psi modulus of elasticity. The experimentally determined meridional and equatorial stresses on the interior of the model were in the 232 Stachiw 4400 to 4600 psi range at 1000 psi exterior pressure. The time dependent strain rate of the model's interior under 3350 feet operational depth submersion (Figures 17 and 18) decreased from 100 microinches/hour, one half hour after submersion to 0.15 microinches/hour after 200 hours of submersion at operational depth indicating that only very little additional time dependent strain would take place in the future if the model was left at operational depth permanently. OPERATIONAL PERFORMANCE OF HABITAT MODELS - The operational performance of the habitat models was successful. No leakage occur- red through the four window assemblies, three wire feed-throughs, and the single instrumentation tower assembly. Inspection of the penetration insert assembly components after implosion of the models showed that no yielding took place in any of the components. CONCLUS IONS Findings based on experimental data resulting from testing to destruction of spherical concrete pressure hulls show that concrete is a reliable material and that buoyant external pressure hulls with a safety factor of two can be built from it on land that will per- form successfully undersea at 3350 foot depth for at least one to two week periods of time. Whether concrete hulls with a safety factor of two will perform successfully at 3500 feet design depth for periods of time measured in years will have to be experimentally established. Indications exist that they will be able to do so successfully. In view of the fact that concrete is a very economical material whose composition is well known, and that construction techniques of large concrete shells on land are well developed, more emphasis should be placed by the U. S. Navy on utilization of this material to permanent, or semi permanent ocean bottom installations in the 0 to 3500 foot depth range. 233 Stachiw REFERENCES 1. Stachiw, J. D. "Glass and Ceramic Hulls for Oceanographic Appli- cations," Second U. S. Navy Symposium on Military Oceanography, Proceedings of the Symposium Volume I, 1965. 2. Stachiw, J: D. "Solid Glass and Ceramic External Pressure Ves- sels,"" Ordnance Research Laboratory Report NOw 63-0209-C-2, Pennsylvania State University, January 1964. 3. Stachiw, J. D. and Gray, K. 0., “Behavior of Spherical Concrete Hulls Under Hydrostatic Loading; Part I - Exploratory Investigation." U. S. Naval Civil Engineering Laboratory, Technical Report R 517, March 1967. 4. Stachiw, J. D., "Behavior of Spherical Concrete Hulls Under Hydrostatic Loading; Part II - Hull Penetrations." U. S. Naval Civil Engineering Laboratory, Technical Report R (under preparation). 234 Collapse Pressure (psi) Stachiw 80 No. 7740 Glass Critical Pressure of Spheres i? » 2 300,000 psi From Selected Materials E = 8.96 x 10° psi 72 f Density = 139 Ib 13 Pyroceram 64 > > 300,000 psi E = 17.3 x 109 psi 5 6 s 3 Fiber Glass - o =75,000 psi, E =5x 10° psi, = | Density (sa ibis Density = 130 Ib/ft = 56 Alumina : L Ee Sera Titanium - o = 150,000 psi, E = 17.5 x 10° psi, 3 Sees si Density = 276 |b/f Ee Density = 238 Ib/ft3 Sa Bat > 2 Aluminum - o =60,000 psi, E = 10.8 x 109 psi, < 40 Density = 173 |b/ft? £ a 32 Steel - 7 = 150,000 psi, E = 30 x 109 psi, ry Density = 490 |b/ft3 a = 24 & Acrylic Plastic - a = 11,000 psi, E = 450,000 psi, Density =75 lb/ft 18 Steel - c =80,000 psi, E =30 x 10° psi, Density = 490 |b/ft3 . 8 Concrete - c = 14,000 psi, E =3.65 x 10° psi, Density = 150 Ib/ft3 0 OF @2 O4h O05 06 0.7 08 Weight of Sphere/Weight of Displaced Seawater Fig. 1 - Critical pressures of spherical hulls made from selected materials 8,000 Critical Pressure of Concrete Spheres 7,000 6,000 Note: E =4.0 x 106 Poisson’s ratio =0.20 S = stress on the interior surface of sphere at implosion : g : 2,000 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Weight of Sphere/Weight of Displaced Seawater Fig. 2 - Critical pressures of concrete spherical hulls as a function of their weight to displacement ratios 235 Stachiw Fig. 3 - Waterproofed and strain-gaged spherical concrete hull without penetration 236 Stachiw Fig. 4 - Hull with penetrations prior to insertion and bonding of solid penetration inserts 237 Stachiw L “iQ Sulsaaulsuz ueasg daag OL & 8 a sjiesut uotye1j0Ued pPt[OS = AsOpesOGE] UCIMD WI 238 Stachiw Fig. 6 - Concrete habitat model for 3500 ft operational depth; imploded at 7400 feet 239 Stachiw suotjerjoued [TeuIs pue od1eT TOF SOsULTT DUTDIOJUTOT UOTJeAJOUNG - L 3ta 240 Stachiw Sutr Zututejyor Moputm (fF) ‘moputm o1jAroe (¢) ‘yeas MOpuTM (7) ‘osueTy Zutorosuter uotzearjoued ({) ‘japour yeqtqey ey, r0z ATGWOSse MOPUTM - g “STq zs (8 aaa 5 = MO MOT] MebLg Mey Agee ny ayy 241 Stachiw surz sututejos ozeTd Sutareaq (F) ‘autop (¢) ‘aqyeTd Butzresq owop (z) ‘esue4pz sult 1OJULEeT UOTJeIJOUEd (T) ‘[epout zeqIqeYy oy} 1OF ATQUIESSe OWIOP UOTJeATOSGO SSeTD - 6 *B1q 242 Stachiw Beeeey Fig. 10 - Habitat model attached to the end closure of pressure vessel used in testing the model 243 Stachiw Fig. 11 - Fragments of the habitat model after implosion at 7400 feet simulated depth 244 Stachiw [T2pou jejiIqey 93919u05 pepoTtduit 9yy UI0rj Uaye} UOTIJIOS [TPM YOY, yout suo ay} jo ydeasojoydorzs1p = BT sear 245 yjdep payenurts J00F NOLO 32 TINY 23e79U09 Teotszoyds ay} JO IOTIOJUT oy} UO SUTeIjs JUopuedep-eUTy, - ZT “SIA Stachiw ("urj-ur7y) UI044S 246 (44) awiy 002 ost O9L orl OzL ool 038 09 OV 0¢ 0 —+ T ae T —T : hell T 0 { ) € 2609 ainssaid yupysuo> isd g ainsseid yuDjsuo>d isd Qgg’¢ Vv Zz 2605 | | 2609 = 00s - ———— \ 000'1- a 2. > @ ’ oO oos t- YW = o o a = c 2 o Si 3 0 a w 000 ¢2- . 2 oO cs) o Uv uo!ssaidwod |pIxp1uN Japun @ “. } isd ggz‘{, = ys5uess a4a19U05 @ | S@4DU!I pl = Jayawoip apisu| ® 1 00S‘Z- Saydul 9g, = Jasawolp apising o rsuolyDI1}12Aad5 asayds fo) mo} as 000'€ - = a 000'r- e1aydg ayaiduoy yO JOlsayuy ay} UO SUIDASS a =e fe = =e =k yjdop poyeMuUIts j99F NOLO 7e TINY 93e19U09 Teotzeyds ayy JO IOTIO4xXO oY} UO suUTeI4s JUOpUedep-sUT], - ¢{[ “3tq (ay) own 00z 081 091 orl ol 001 08 ainsseid yupysuod isd q Stachiw uoissaidwo> |pIxpiun 4apun isd 00Z‘LL = yyBuasys 8491907 S@yUl p| = Je;awWDIP apisu| S@YIU! O| = JasawDIP apisjnO iSUO14¥DI1}19adG e1ayds a ) = < 3 Ls} oO > v a oa a 1) c = oa - ° o Uv uw OV 02 ainssaid yupjsuod isd 000’€ ————_ sd 990’ 9} eunssaud jo Bulsips asaydsg aya42u07 JO 101494x UO SUIDAIS 00s‘E- 000'7- 247 Stachiw S[OpOUl TINY 9491DU0D UT posn 94919U09 jo AjtotyseTSe JO sn—TNpour yuesuey, - FL “3ty (901 x 1sd) snjnpow suabup | S v € 4 l 0 000’01- isd ggo’ZL > >} > !s4 900'01 00S ‘Z- LD) Seen SS S25 SK oS < Q © O32 KS 000'S- |Mpow jo abun a ae ; ('sd) ssaays PQRQQRLIF COV Oe, poneee VU UC Socececenes 005°2- WW SKK oS vecenas < oS SK SRK XX — > OO ene SxS 5 SxS oS uoissaidwos) jDIxy 418pus) sdapuijA> $Sa] a4219U0D g x € fO Ayia 14sD} 4O SA|NPOW juaBup | 248 Stachiw Vv € 2605 B Z 2605 e@ | 26p5 aljassoy SJIOSUT UOTJeIJOUed PTITOS YIM sT[NY 93019U00 Teotzeyds ay} JO TOTIOIUT BY} UO SUTeIS - GT “314 (Yau /saysuloisiw) UIDI4S 000'E- 000'Z- ulw/isd QO] = 2404 Uo!yoZIINSSaddq ,, Supipp4 jDI14ayds /9G°Q = JesaWDIp jiasu| , HaSUI jaa4s - a4 Hasul wnuiwniy - |v dasul S14spjd - DAd B2D}1NS a4949U0D - ZOIS :UO SUIDISS ~SHOSU] PISS YiIM Sazayds fO 4O14a4U] a4} UO SUID’ S4NSS31 jouse4yx3 (14) ¥¥ 249 Stachiw 000'0z- @ Vv 000’91- SS9d41S [DWION “UlW SSO4S [DW4ON “xDW ajjasoy S$]IOSUI UOCTJeI}OUEd PITOS YIM ST[TMY 249.19uU09 Teotsz9yds 9y4 JO TOTI94UT BY} UO S9Sssatj4g - OT “BIW (1sd) SSO4S 000'ZL - 000'8- ul /isd QO, = a4D4 UOIJDZIINSSe1q ,, SUDIPD4 [DI IWaYyds /9G°Q = JeJeWIDIP J4asu| , 000'7- 0 000'L _SHOSU] PILOS YHIM sasauds JO JOlJaju] ayy Uo Sassais 000’€ (1sd) SINSSB1Id [DUISIx> * 250 Stachiw (You! /Saysulosd IW) UID44S 000'r- 000'E- 000'7- ulW/ISd QO] 2404 UOIyDZIuNSSadgq , (Z[ 21nB14 eas) S4noy QZ 404 1Sd 0S‘ $D poulbjulbw S4NSSaid 000'1- [2POW 404!qD}] e4esDU0> JO Jolia4yuj ayy UO SUIDIYS uotsojduit 0} ro1szd Tapout yejIGqeYy 9}3aT9U0D 9Yj Fo ATOJSIY UTeITIS - pT “BIA 000'L 000’r (isd) S4INSsSaid [Ousayx3g 251 yydep poJye[Nuits JO JOoF OGEE 3e Tepow jejyIqey 9}919U09 ay} JO TOTIOJUI OY} UO SsUTeI}s yuOpuoedop SUIT], - BT “S1q (1sd Qog’L oF UOlJDZIINSSaid 4a44D S4NOY) Sul | 002 og 0c 1 08 ov ¢ e6p4 Vv B Z EJe}e9) 3) l eBoy 344ES Od Stachiw AINSS21q D1yDJSOspALy [DUIa4xF }UDJSUOT }O isd ggg’, 49pUN |ePOW +94!q0H ajasauo7) 4O JolJayuy] ayy UO SUIDIJG JUepUsded ow! | OOlL- 00C- 00E- (YoUI /S@YDUIOAD IW) UIDA4S 252 Five Years Experience with a Shipboard Oceanographic Data Processing System: Hindsight and Foresight * C. O. Bowin Woods Hole Oceanographic Institution Woods Hole, Massachusetts INTRODUCTION Several years ago the Geophysics Department of the Woods Hole Oceanographic Institution recognized the need to accomplish at sea as much data analysis and interpretation as possible concurrent with the investigations. In recent years, oceanographic instrumen- tation developments have made it practical to collect large amounts of data during a single cruise, and the reduction of these data involves the processing of similar amounts of navigational informa- tion. The use of digital computers is a natural solution to the problem of keeping pace with an accelerating data collection capability. Data which has been collected must usually undergo a certain amount of processing or reduction to be meaningful. A researcher generally cannot evaluate data concurrent with its acquisition; this task is usually performed ashore, either manually or with the aid of a computer. Thus, scientifically interesting or significant regions may be discovered only after the data has been reduced ashore, often necessitating a return trip at great expense. Computer processing at sea eliminates this delay and expense, and allows scientists to monitor, assimilate, and evaluate the data at sea. Some data may prove worthless upon reduction at the shore * Contribution No. 1937 from the Woods Hole Oceanographic Institution. 253 Bowin facility. This may be avoided, to some extent, by periodically checking important equipment voltages or parameters. With the necessary interface, the computer can be programmed to check and control scientific or navigational instruments and signal the operator when a detected abnormality exists. The scientist may thereby be released from many routine tasks to plan and direct the current research, and to interpret results. Also, by reducing the data on-line, incorrect or unreasonable data is more obvious, anda prompt check of the scientific instruments can be made. If the computer has the capability of sharing its operation between more than one task (time-sharing), general-purpose computation can be done while the main program is underway. THE WOODS HOLE EXPERIMENT Since June of 1962, two shipboard data processing and control systems have been implemented by the Geophysics Department of the Woods Hole Oceanographic Institution utilizing an IBM 1710 Computer System. The first system (System I) is described by Bernstein and Bowin (1963) and Bowin (1963). The second system (System II) is described by Bowin et al. (in press) and Bowin et al. (1967). These efforts were funded largely by contracts Nonr-1367 (00) and Nonr-4029(00) with the Office of Naval Research, Department of the Navy. These systems made it possible to automatically sample, compute, and record data concerning the ship's heading and speed, latitude and longitude, water depth, acceleration due to gravity, free-air and Bouguer gravity anomalies, and the magnetic field of the earth. System II also provided on-line plotting of bathymetric, gravity anomaly, and magnetic field profiles; computer control of gravity meter spring tension; processing of surface temperature measurements and ocean sound velocity measurements; reduction of Loran C and VLF radio navigation data to latitude and longitude; display of ship's position and numerical data at remote stations aboard the ship; and malfunction detection and alarm message generation. Experiments were also made using three input/output typewriters at remote locations on the ship. A block diagram of System II is shown in Figure 1. Presently, we are implementing a new shipboard computer system (System III) utilizing a Hewlett Packard 2116A computer. This new system gives promise of considerable expansion capability and as time, money, and personnel are available, we plan to continue development towards a multi-purpose, 254 Bowin multi-discipline facility. The goal is the development of a system capable of automatic sampling, computation, recording, and display of scientific information concurrent with the investigation (real- time); automatic control of scientific instruments and equipm ent (feedback control); and the capability of background time-sharing for compilation, assembly, off-line programs, diagnostics, and experimental on-line programs without endangering the operation of the main real-time program. How far we will be able to progress towards this goal remains to be seen. It is important, however, that our initial system (both hardware and software) incorporates the concept that someday all these objectives may be attained. HINDSIGHT AND FORESIGHT The operation and use of a prototype system obviously provides experience which can be obtained in no other way. The main purpose of this paper is to summarize the important aspects of that experience. Although the details of the Woods Hole systems are by no means necessarily directly pertinent to the needs or problems of other organizations, Many aspects have important relation to the general problems of shipboard data processing. Our experience from five years of development and use of a shipboard oceanographic data processing system has confirmed the value of such systems in the accomplishment of scientific investigations at sea. Most significant is the production by a digital plotter of profiles of bathymetry, gravity anomalies, and magnetic total intensity along the ship's track in real-time while the ship is traversing that part of the ocean These records have enabled scientists aboard to better assimilate the large amounts of data being collected and thereby more effectively utilize both the ship and their time at sea. Looking back over the laSt several years brings first to mind the interest, dedication, and creativity of the many people at the International Business Machines Corporation and the Woods Hole Oceanographic Institution who helped in the implementation of the system. Although a commitment on a personal level perhaps is not an essential ingredient of a successful program, the numerous demands on it during the system development suggest that it is important. It is amatter of some importance that the development of a system should be directed by the person who will use the collected data. Many possible difficulties and misunderstandings can be thereby circumvented. 255 Bowin A stable and reliable power supply is essential to a system. Inverters with precision frequency control were purchased for the R/V CHAIN to supply power to the computer. These inverters have several advantages. They provide a precise and stable frequency for the operation of the system. The IBM 1710 computer required a primary power frequency stability of + 0.5 hz, which could not be met by normal ship's power. Precise frequency is also important for all clocks and recorders that have synchronous motors as their basic driving motor. The inverters were modified so as to be driven by an external oscillator, which, in our case, provided frequency control that was accurate to better than one thousandths of a hertz. In addition, the inverters were powered by the ship's DC battery which in turn is trickle-charged by the ship's DC buss. This arrangement assures continued operation during possible ship's power failures lasting up to forty-five minutes. Another advantage in operating the inverters from a battery is that large voltage transients on the ship's DC buss are in large part filtered by the battery before they reach the inverters. Our system utilized one 3 KVA inverter and one 10 KVA inverter. An important aspect of a real-time computer system is the identification of the data which is collected. We chose to identify the data by the date and time at which it was sampled. That identification might prove to be a troublesome source of difficulty was in large part only recognized during subsequent data analysis. Errors that occurred included failures of the contacts in the mechanical clock, thereby resulting in invalid or erroneous values being supplied to the computer. Errors also occurred because the clock stopped whenever the computer power was turned off. Oftentimes during initialization of the program, the operator either would forget to enter the date and time or would enter erroneous values. In all these cases the errors proved very troublesome in processing the data later on shore. In the new system, we are endeavoring to eliminate these problems. To this end we have obtained a solid state electronic calendar clock that operates independently of the computer, and supplies the system with the day and month, and time in hours, minutes and seconds. This clock will also have its own battery backup supply in case of failure in the inverters. Whenever the system requires the time and date, it reads the clock, thereby always having accurate and reliable time. We expect this to be far superior to manually setting the time, and utilizing a counter within the computer program. 256 Bowin It is undesirable to subject electronic components to ambient temperatures in excess of about 90°F. Because temperatures this high, and higher, are relatively common in many parts of the world throughout much of the year, it is necessary to furnish air- conditioning for most computer systems. At the time of installation of our original system, the CHAIN was not air-conditioned. It was therefore necessary to build a room around the system in the main laboratory of the ship. Three water-cooled air-conditioners were used for cooling the room. This proved completely satisfactory. Even though difficulties with one or more of the air-conditioners was undesirably common, at least one of them was always operating We were thereby able to maintain the computer at its required temperatures. At the same time as the system was expanded in 1963, air-conditioning was supplied to the laboratories, living quarters and messrooms of the CHAIN. Following that time, we dispensed with the room in the main laboratory that had previously sheltered the computer. Although we were no longer directly responsible for air-conditioning, it nevertheless required the suspension of operations whenever the ship's main compressor unit failed. This experience suggests that some additional independent air-conditioning capability should be provided for the computer system if the ship itself is generally air-conditioned, or that more than one air-conditioning unit should be available in any event. High humidity has not been identified as a difficulty in our past systems. The only problem encountered occurred when water actually dripped on the components of the system from either an overhead vent or from condensation upon the cooling ducts in the laboratory. Both these troubles have been corrected; one by sealing off the overhead vent and the other by coating the ducts with an insulating material. Vibration is another concern of most people who are contem- plating utilizing a computer aboard a ship. Again this has not been a serious problem in our installation. We have attributed only one failure directly to vibration when it was believed that a low voltage protection relay may have been erroneously actuated by vibration. The central processor of the IBM 1710 computer system proved extremely reliable. In the years of operation we had only three failures in the central processor, and two of these were caused when the IBM maintenance man accidentally shorted printed circuit cards. One failure of a printed circuit card in the analog to digital converter demonstrated a weakness of equipment diagnostic Bowin programs. The manifestations of this failure were excessive stepping of the automatic spring tension controller for the gravity meter, water depths were changed to zero, and the digital display showed unbelievable numbers. Diagnostic programs failed to reveal any difficulties. Although adjustments of a bias voltage in the analog to digital converter helped relieve the trouble, the failing printed circuit card was not identified until the ship had returned to Woods Hole and an IBM engineer from San Jose, California worked on the system. The problem appears to have been caused by components that failed only when loaded by the computer doing multiple operations. The diagnostic tests did not reveal the trouble because they carry out only one test at a time. We now believe that most serious difficulties during the first three and a half years of operation were the result of radio frequency interference. Now, of course, we can only surmise that many of the éarlier difficulties were caused by RFI. However, the capriciousness and mysteriousness of the earlier troubles have a strong similarity to the later troubles which were positively identified to be caused by RFI and leads us to the strong supposition that this was a major source of trouble earlier. The manifestation of the interference was the alteration of values in the computer program. For example, one of the first indications of such trouble was erroneous values of ship's speed and water depth. In both these cases the trouble was identified as occurring either during the analog to digital conversion or immediately following. Somehow the interference was coming into the computer system. Many attempts were made to determine the source of these troubles; including the removal of antenna lead-in lines coming directly into the main laboratory and near the computer facility, installation of radio frequency suppression filters in the ship's radio power lines and in the 1710 system, and improvement of the system ground. These attempts usually temporarily alleviated the problem, but were not a permanent solution. In spite of these efforts, RFI problems continued which eventually developed into a mutual interference between the ship's radio and the computer system in which transmission would disturb values within the computer system. Also the units of the computer, in particular the type- writers and digital plotter, could be heard on many frequencies of the ship's radio receiver. When this interference became strong on 500 Khz (international calling and distress frequency) the problem demanded immediate correction. We tried the drastic action of removing all the personal radio receiver lines installed haphazardly throughout the ship by the scientists, crew, and officers, including the Captain's and the cook's. Older abandoned 258 Bowin lines throughout the ship were removed as far as possible or cut off where further removal became impractical... New antenna lead-in lines were then installed to avoid their approaching or intertwining with the systems cabling as they had previously. Following this action, which took place in the Fall of 1965, we have had practically no radio frequency interference either from the ship's radio trans- mitter or to the ship's radio receiver from the computer. Our second greatest source of difficulty has resulted from operator action or misaction in performing his tasks. This will be discussed in more detail later. The third most frequent source of trouble has been with mechanical components of the system; in particular the typewriters, and to a lesser extent the reed-relays that operate the digital display units. During the initial testing period the reed-relays exhibited frequent failures which were difficult to analyze because the trouble was intermittent and the displays were the means by which we were checking the values. However, once a good set of reed-relay cards were found, they functioned very well from then on. The typewriters broke springs, keys, and other parts which in most cases were easy to replace. The system operation was not Seriously affected by these difficulties because by merely throwing a switch the output could be transferred from one type- writer to another. One of the surprising performances was given by the IBM 1311 disc drives. These disc drives rotate on a vertical axis at 1500 revolutions per minute. These rotating discs act as gyros, and it was felt that during the rolling and pitching of the ship they would exert considerable force upon the bearings in which they rotate and thereby more rapidly wear those bearings. In four years of operation we had only one bearing fail amongst the three disc drives and fortunately, although the spare bearing that we supposedly had aboard the ship could not be located, a bearing amongst the ship's spare parts did miraculously fit the unit and it was able to be returned to operation. We did come, however, to the practice of securing the disc drives during very rough weather. We still do not know whether this is essential or not; our only experience being that following a particularly rough storm, in which operation of the discs was continued, oxide was noted around the outside of the disc area and several scratches existed on the disc surfaces. We do not know whether these scratches were formed during the storm or were there before and only noted upon careful examina- tion following the storm. However, this experience did cause us to 239 Bowin act cautiously in their operation during rough weather. The necessity of securing the primary recording units of a system is undesirable. We are hoping that we will be able to operate the magnetic tape drives of System III through all weather conditions encountered. At present, however, we do not know if this will be in fact the case. We have found that all inputs that are entered manually by an operator, particularly routine manual inputs are unsatisfactory. All human beings will, with varying frequency, either enter erroneous values, turn the wrong switch, push the wrong button, or otherwise make mistakes. If the operator is to make an entry at a particular time, this is another potential for error. Mistakes will also occur even if the values are entered by the person who will later make use of the data, and therefore by someone who will have the most conscientious attitude towards the task. Such errors create an enormous amount of wasted time and difficulties in the correction and analysis of the data. For example, we are finding that the correction of water depth information that was usually entered manually in Systems I and II, are taking more time to correct than was spent in gathering the data. In System III we are dispensing with all routine manual inputs. Those values that are needed and their entry is not yet automated, will be digitized later ashore using conventional digitizers. Because, as mentioned previously, time and date are the means by which we identify individual data points, we have obtained a calendar clock to provide that information to the system, thereby eliminating the necessity of relying upon a person to enter such information. Should the clock or other inputs fail, we will have provision, of course, for manual input as a backup measure. The program for Systems I and II was written in an assembly language which at the time was the only language capable of sampling analog input channels, BCD input channels, branch indicators, processing interrupts, and activating contact closures. During the development of the Systems I and II, the real-time program grew piecemeal as ideas for improvement and expansion of the computers capability developed. A complex interleaving and interconnecting of various parts of the program is a most undesirable way for the computer program to grow. In the latter stages of our utilization of System II we had reached the state where even the programmer who wrote the real-time program was afraid to make modifications or additions for fear it would jeopardize the operation of the system. One still unexplained mystery that may in part have resulted from program interaction was the inability 260 Bowin to originally have three input/output typewriters function properly at the same time. Any combination of two would work well, but as soon as the third one was added, typing of double letters and various other unexplained difficulties ensued. The situation was remedied merely by rearranging the priority numbers of the three typewriters Why priority assignments should have caused the trouble in the first place, or why a rearrangement of the assignments should have been a solution, remains amystery. All that we can do is surmise that there was Some interaction between the computer program and critical timing requirements. It is perhaps quite instructive to note that we had our most reliable operation of System II during its six months cruise to the Mediterranean immediately before being returned to IBM. The program used during this entire cruise had been checked and corrected on a prior cruise and no further modifications were made. This record suggests that program improvements are not always improvements unless they can be thoroughly tested and corrected, preferably at times when data is not of importance. The use of a system invariably leads to ideas for improvement and expansion, and identifies errors in the original program. To meet these situations, it is very desirable that modifications to the real-time program may be made with ease and with confidence that such changes will not adversely affect other portions of the program. These requirements can be facilitated using a modular construction for the program rather than a complex interweaving of operations. It is also highly desirable that the real-time program be written by the user, or at least that he be able to easily follow the coded computer program. Neither was the case in Systems I and II which led to our dependence upon the IBM programmers for even rather simple modifications. This depend- ency is particularly frustrating in the years following the completion of an original contract by which such programs may have been written, and the dispersement of the programmers to new contract efforts essentially making them unavailable for further development. The facility of a user to write his own programs or to easily understand a program written by someone else is best accomplished, to date, by writing the program in Fortran. The use of an assembly language, or worse yet, machine language, only further isolates the average user from the ability to be creative in his use of the computer. Also, the larger the core memory, the more practical and convenient will be program improvement and an expansion of the tasks conducted by the system. This will be the case whether or not the system has random access 261 Bowin storage capability. The tasks to which modern computer facilities may be directed lie in four categories: 1) numerical and logical analysis, 2) acqui- sition and collation of information, 3) display of information, and 4) control of equipment and instruments. Of these four tasks, only analysis techniques have reached a stage of implementation that could be considered young adulthood. To continue the analogy, display may be in adolescence, and acquisition and control are yet in childhood. Although notable improvements have been made in acquisition and control computers during the last five years, the implementation of these capabilities remains an effort beset by equipment specifications and timing problems; still very much dependent on the individual computer being used. The solution of these problems for one particular computer does not necessarily simplify the effort in accomplishing the same task on another computer. Developments that I expect to take place during the next several years include programming and equipment improvements that will facilitate acquisition, control, and time-sharing capabil- ities ; reduction in the size and cost of peripheral equipment; and greater stress on reliability of commercial computer systems. Digital computers now offer far more speed and computational abilities at lower cost then the machines of five years ago. The decision today is not whether to take a computer to sea, but which computer and how many tasks should be assigned to it. The main decision during the next several years will be in choosing between the alternatives of a large central computer system, a medium- sized central computer with small satellite computers, or many small independent computers for the accomplishment of a total program. Many factors will influence this decision including: the rate at which funding is expected; the future cost of large versus small systems; the memory size, computing power, and extent of peripheral devices desired; the degree to which future expansion capability is desired; and the degree of reliability that is required by the mission of the system. REFERENCES Bernstein, R. and CC. ©. Bowin, 1963: Real-time’ Digital Computer Acquisition and Computation of Gravity Data at Sea, IHEE Trans. on Geoscience Electronics, vol. GE-1, MO, by . A=i1O, 262 Bowin Bowin, C. O., R. Bernstein, HE. Ungar, and J. R. Madigan, in press. A Shipboard Oceanographic Data Processing and Control System, IEEE Trans. on Geoscience Electronics. Bowne. Orga. Oe) Nichols), i. Aldrich) Ho sh Dusan, G. N. Ruppert, and J. B. Zwilling, 1967. Second Supplementary Report on the Oceanographic Shipboard Data Processing and Control System (ODPCS) Aboard the Research Vessel CHAIN October 1963 - December 1966, WHOI Ref. No. 67-26 of the Woods Hole Oceanographic Institution. 263 AUTOMATIC INPUTS GRAVITY METER EM LOG (SHIP'S SPEED) GYROCOMPASS (SHIP'S HEADING) INPUT MAGNETOMETER PGR (AUTOMATIC DEPTH) SURFACE WATER TEMPERATURE SYSTEM REFERENCE VOLTAGES MANUAL INPUTS SWITCHES (DATA) LATITUDE/LONGITUDE FIX LORAN-C TIME DIFF. FIX VLF. PHASE DIFF, FIX QUALITY, SOURCE OF FIX OCEAN DEPTH DIGITAL DISPLAY SELECTOR SWITCHES (SENSE) PROGRAM MODIFICATION TYPEWRITER KEYBOARD LOGGING MESSAGES TIMED INTERRUPTS PROGRAMMABLE TENOR CLOCK CONVERSION EQUIPMENT Bowin 1BM 1710 CONTROL SYSTEM MULTIPLEXER A/D CONVERTER DIGITAL INPUTS CENTRAL PROCESSING UNIT ADDITIONAL MEMORY REAL-TIME CLOCK A) CONTINUOUS. UNDER-WAY OPERATION MANUAL INPUTS TYPEWRITER LATITUDE /LONGITUDE DATE & TIME LOWERING NUMBER VELOCIMETER NUMBER CRUISE NUMBER VELOCIMETER CONSTANTS OPERATOR NAME COMMENTS AUTOMATIC INPUTS VELOCIMETER TEMPERATURE SENSOR OCEAN DEPTH B) ON-STATION OPERATION INPUT CONVERSION EQUIPMENT CONTROL COMPUTER IBM 41710 CONTROL SYSTEM OUTPUTS & D.P EQPT. OATA OUTPUT PRINTERS (2) LOGGING TYPEWRITER OUTPUT (3) WATER DEPTH TYPEWRITER OUTPUT ALARM MESSAGE TYPEWRITER LATITUDE/LONGITUDE DISPLAYS(3) DIGITAL DISPLAYS (2) REMOTE ALARMS (3) DIGITAL PLOTTER GRAVITY METER CONTROLLER CARD READER/PUNCH MAGNETIC DISK DRIVES (3) OUTPUT & D.RP EQPT. DATA OUTPUT PRINTER DIGITAL PLOTTER DIGITAL DISPLAY CARD PUNCH DISK DRIVE Fig. 1 - Block diagram of the oceanographic data pro- cessing and control system aboard the R/V CHAIN 264 CORROSION OF MATERIALS IN HYDROSPACE PART I IRONS, STEELS, CAST IRONS, AND STEEL PRODUCTS Fred M. Reinhart U. S. Naval Civil Engineering Laboratory Port Hueneme, California PREFACE The U. S. Naval Civil Engineering Laboratory is conducting a research program to determine the effects of the deep ocean environ- ment on materials. It is expected that this research will establish the best materials to be used in deep ocean construction. A Submersible Test Unit (STU) was designed, on which many test specimens can be mounted. The STU can be lowered to the ocean floor and left for long periods of exposure. Thus far, two deep ocean test sites in the Pacific Ocean have been selected. Six STUs have been exposed and recovered. Test Site LI (nominal depth of 6,000 feet) is approximately 81 nautical miles west-southwest of Port Hueneme, latitude 33°44'N and longitude 120°45'W. Test Site II (nominal depth of 2,500 feet) is 75 nautical miles west of Port Hueneme, latitude 34°06'N and longitude 120°42'W. This report presents the results of the evaluations of the irons, steels, low alloy steels, alloy cast irons, metallic coated steel, uncoated and metallic coated steel wire ropes and anchor chains for six exposure periods and two nominal depths. The effect of stress on some of the materials is also reported. INTRODUCTION Recent interest in, and emphasis on the deep ocean as an operating environment has created a need for information about the behavior of constructional materials in this environment. The Naval Facilities Engineering Command of the Office of Naval Materiel is charged with the responsibility for the construction of all fixed naval facilities, including the construction and main- tenance of naval structures at depths in the oceans. Fundamental to the design, construction and operation of structures, and their related facilities, is information about the deterioration of materials in the deep ocean environments. This 265 Reinhart report is devoted to the effects of these environments on the corrosion of metals and alloys. A test site was considered to be suitable if the circulation, sedimentation, and bottom conditions were representative of open ocean conditions. A site meeting these requirements was selected at a nominal depth of 6,000 feet. The location of this site in the Pacific Ocean in relation to Port Hueneme and the Channel Islands is shown in Figure 1 as Submersible Test Units (STUs) 1-1, 1-2, 1-3, and 1-4. The SCHL ES oceanographic data for Site I are shown graphi- cally in Figure 2.~° A portion of this data collected from 1961 to 1963 showed the presence of a minimum oxygen zone (as shown in Figure 2) at depths between 2,000 and 3,000 feet. Oceanographic data obtained at other sites also showed the presence of this minimum oxygen zone regardless of depth to the ocean floor. Corrosion rates are affected by the concentration of oxygen in the environment. Therefore, it was decided to establish a second exposure site (STU II-1 and II-2) at a nominal depth of 2,500 feet. This site is also shown in Figure l. The NCEL oceanographic investigations also disclosed that the ocean floor at each of these sites was rather firm and was char- acterized as sandy, green cohesive mud (partially glauconite) with some rocks. This report presents and discusses the results obtained from exposure of irons, steels, low alloy steels, alloy steels, unalloyed and alloyed cast irons, steel wire ropes, anchor chains and metallic coated products for six peridds of time and at two nominal depths. RESULTS AND DISCUSSION Dr. T. P. May, Manager, Harbor Island Corrosion Laboratory of the International Nickel Company, Inc. has granted permission to incorporate his corrosion data (Reference 4), obtained from their specimens on the six STU structures, with the NCEL data. Surface data of some alloys from the Atlantic Ocean (Reference 5) and similar to those from the Panama Canal Zone, Pacific Ocean (Reference 6) are included for comparison purposes. Deep ocean data from the Atlantic Ocean is also included to permit comparison of the different deep ocean environments, References 7, 8, and 9. The corrosion rates of all the alloys are shown graphically in Figure 3. Water in the open sea is quite uniform in its composition ‘throughout the oceans; therefore, the corrosion rates of steels exposed under similar conditions in clean sea water should be com- parable. The results of many investigations on the corrosion of structural steels in surface sea water at many locations throughout the world show that after a short period of exposure the corrosion rates are constant and amount to between 3 and 5 mils per year 012,13 Factors which may cause differences in corrosion rates outside these 266 Reinhart limits are variations in marine fouling, contamination of the sea water near the shorelines, variations in sea water velocity, and differences in the surface water temperature. IRONS AND STEELS Corrosion The corrosion rates of low carbon steels in sea water at different locations are compared in Figure 3: a. Surface waters of the Atlantic Ocean at Harbor Island, North Carolina; b. Surface waters of the Pacific Ocean at Fort Amador, Panama Canal Zone; c. Deep Atlantic Ocean waters, Tongue-of-the-Ocean, Bahamas ;/>859 d. Deep Pacific Ocean waters, Port Hueneme, California. The corrosion rates of the steels at the surface in both the Atlantic and Pacific Oceans decrease rather rapidly with time and become relatively constant after about 2 to 3 years of uninterrupted exposure. The higher corrosion rates at Fort Amador are attributed to the difference in temperature between the two sites (27°C vs 21°C). The corrosion rates of the steels exposed at nominal depths of 5,500 and 2,350 feet in the Pacific Ocean also decreased with time of exposure and were consistently lower than the surface corrosion rates. These lower corrosion rates are attributed to the combined effects of the differences between the variables at the surface and at the two depths; temperature, pressure and oxygen con- centration. Also, the corrosion rates at a depth of 2,350 feet were lower than those at a depth of 5,500 feet. In this case the lower corrosion rates at a depth of 2,350 are attributed to the combined effects of the differences between the variables at the two depths; temperature, pressure and oxygen concentration. The above differences in the corrosion rates cannot be attributed chiefly to any one variable because of the interdependence of one variable on another. For example, the solubility of oxygen in sea water is increased as the pressure is increased at constant temperature but at constant pressure the solubility of oxygen de- ereases as the temperature increases. The corrosion rates at a depth of 5,500 feet in the Pacific Ocean were about one-third the rate of the steels at Harbor Island after about 3 years of exposure. The corrosion rates for a steel exposed by the Naval Research Laboratory at a depth of 5,600 feet in the Tongue-of-the- Ocean in the Atlantic were slightly higher than those in this 267 Reinhart investigation, Figure 3. Oceanographic data reported for the Tongue- of-the-Ocean are: depth, 4,967 feet; 4.18°C and 5.73 m1/1 oxygen. 14 Since the differences between the depths, pressures and temperatures are small the higher corrosion rates in the Atlantic are attributed chiefly to the difference in the concentration of oxygen between the two locations (5.73 vs 1.4 ml/1) with the possibility that some might be due to the difference in the currents (unknown in the Atlantic but practically stagnant in the Pacific). The difference between the corrosion rates on the surface at Harbor Island, N. C. and at a depth of 5,600 feet in TOTO is attributed to differences in depth (pressure, 0 vs 2520 psi) and temperature (19°C vs 4.2°C). Corrosion rates for steel at a depth of about 4,500 feet®>? in TOTO were practically the same as those at the surface at Harbor Island for comparable periods of time. The corrosion rates of wrought iron and Armco iron at depths were comparable with those of AISI 1010 steel as shown in Figure 4. The corrosion rate of wrought iron at the surface at Fort Amador in the Pacific Ocean Panama Canal Zone!9 after about 3 years of exposure was approximately 7 times greater than at a depth of 5,500 feet in the Pacific Ocean. The corrosion rates of all the alloy steels at depths of 5,500 and 2,350 feet in sea water are shown in Figure 5. These values are shown as shaded areas encompassing most of the values. The corrosion rates for these steels decreased similarly to those for carbon steel with time of exposure at both depths. Although the corrosion rates at a depth of 5,500 feet varied between 1.9 and 6.0 MPY after 123 days of exposure they were all essentially the same after 1,064 days of exposure (0.5 to 0.9 MPY). The performance of these same steels when partially embedded in the bottom sediments is shown in Figure 6. After 1,064 days of exposure at a depth of 5,500 feet, the corrosion rates were the same as those in the sea water above the bottom sediments. However, the corrosion rates for many of the steels after 403 days of exposure in the bottom sediments at a depth of 6,780 feet were less than 0.5 MPY; this is attributed to the greater proportion of each specimen that was embedded in the bottom sediment. The specimens of these particular steels were about 2 inch diameter discs and in all probability were nearly completely embedded in the bottom sediment. The data for all the steels was analyzed statistically. The mean curve of the corrosion rates and 95 percent confidence limits are shown in Figure 7 for the specimens exposed in the sea water. The corrosion rate curves for AISI 1010 steel and high-strength-low alloy steel #2 exposed at a depth of 5,600 feet in TOTO are also included to reveal that they are outside the 95 percent confidence limits. The fact that they are outside the 95 percent confidence limits of the corrosion rates of the steels exposed at a depth of 5,500 feet in the Pacific Ocean indicates that the environment in the Atlantic Ocean is somewhat different from the environment in the Pacific Ocean. The median curve of corrosion rates for the 2,350 foot depth is below that for the 5,500 foot depth indicating a difference in environment even though the confidence limits overlap. 268 Reinhart The median corrosion rate curves for the 2,350 foot and 5,500 foot depths are shown in Figure 8. This figure shows quite clearly that after two years the water and bottom sediment environments are alike, with regard to their effect on the corrosion of steels. It also shows that the bottom sediment at 5,500 feet is about the same as the sea water environment at a depth of 2,350 feet. After 400 days the bottom sediment at a depth of 2,350 and the 2,350 foot level sea water environments are of equal aggressiveness in their corrosive actions. Variations of from 1.5 to 9 percent in the nickel content of steel were ineffectual with respect to the corrosion rates as shown in Figure 9. Stress Corrosion Some of the steels were exposed in the stressed condition at values equivalent to 35, 50 and 75 percent of their respective yield strengths. None of these steels were susceptible to stress corrosion. cracking for the periods of time exposed at the various depths. Corrosion Products The corrosion products from some of the steels were analyzed by X-ray diffraction, spectrographic analysis, quantitative chemical analysis and infra red spectrophotometry. The constituents found were: Alpha iron oxide - Fe 903 ° H90 Iron hydroxide - Fe(0H)»5 Beta iron(IIL) oxide hydroxide - Fe00OH Iron oxide hydrate - Fe903 ° Hy 0 Significant amounts of chloride, sulphate and phosphate ions. Anchor Chains Two types of 3/4 inch anchor chain, Dilok and welded stud link were exposed at depths. The chain links were covered with layers of loose, flaky rust after each exposure. The layers varied from thin to thick as the time of exposure increased. Destructive testing of the exposed chain links showed no decrease in the breaking loads of the links for periods of exposure of at least 1,064 days. Hence, there was no impairment of the strength of either of the chains. The Dilok links all.failed at the bottoms of the sockets where the cross-sectional area of the steel was the smallest. Rust was present in all these broken sockets indicating that sea water had penetrated the joints. Stagnant sea water in these sockets for 269 Reinhart prolonged periods of time could result in destruction of the links due to the internal stresses created by the formation of corrosion products. Wire Rope A number of metallic wire ropes were exposed at various depths and for different periods of time. These were plow steel, galvanized steel, aluminized steel, stainless steel and 90 copper- 10 nickel clad stainless steel ropes and cables of different types of construction. The zine on the 0.125 inch diameter, 7 x 19 construction, lubricated galvanized aircraft cable was completely covered with red rust after 403 days of exposure at a depth of 6,780 feet. In addition, the breaking strength had decreased by 50 percent. The amount of zinc remaining on the other galvanized ropes varied from none in the case of the 0.094 inch diameter, 7 x 7 cable which was 100 percent rusted on the outer surfaces to considerable remaining on the 0.25 inch diameter, 7 x 19 construction cable which was dark gray. There was no loss in the breaking strength of any of these five cables. After 403 days of exposure at a depth of 6,780 feet the smaller diameter (0.094, 0.125 and 0.187 inch diameter) stainless steel cables lost considerable strength, 90, 86, and 96 percent respectively. These decreases were all attributed to crevice cor- rosion of the internal wires. Many pits were also found on the individual wires away from the breaks and some broken ends were pro- truding from the cables prior to testing. There was no loss in breaking strength of the three larger diameter stainless steel cables, the inside strands were chiefly metallic color with only a few localized rust spots. Two types 304 stainless steel cables clad with a 90 percent copper-10 nickel alloy were exposed for 402 days at a depth of 2,370 feet. One cable, 1x37x7 construction with a 0.3 mil thick clad layer was covered with rust on the outside but the inside wires were uncorroded. The other cable, 7 x 7 construction with a clad layer 0.7 mil thick was covered with green corrosion products on the out- side, uncorroded on the inside strands and had lost no strength. Three aluminized steel cables (7 x 7, 1 x 19, and 1 x 19 construction) with 0.6, 0.6 and 0.7 mil thick coatings lost no strength during the 402 day exposure at a depth of 2,370 feet. The 7 x 7, 0.187 inch diameter cable was covered with white corrosion products and a few light rust stains but the inside strands were dull gray in color. The outside surfaces of the 1 x 19 construction wires (0.250 and 0.313 inch diameter) were gray in color with scat- tered white corrosion products covering about 50 percent. of the surfaces. The inside strands were a dull gray color. Eight wire ropes were stressed in tension equivalent to approximately 20 percent of their respective original breaking strengths. There were no stress corrosion failures after either 751 or 1,064 days of exposure. However, the breaking strength of the 270 Reinhart Type 316 wire rope lost 40 percent of its strength after 1,064 days of exposure at a depth of 5,300 feet because of crevice corrosion of the internal wires. The breaking strength of the galvanized plow steel (0.83 oz Zn) was decreased by 17 percent. The breaking strengths of the other six wire ropes were unaffected, Although there was no loss in the breaking strength of the 18 percent chromium-14 percent man- ganese stainless steel rope there were quite a number of broken wires due to corrosion both on the outside and on the inside strands. Metallic Coatings Zinc, aluminum, sprayed aluminum and titanium-cadmium coated steel specimens were exposed at depth. The galvanized steel (1.0 oz per sq ft) was covered with a layer of flaky red rust after 402 days of exposure at a depth of 2,370 feet. The corrosion rates were 0.9 MPY for the specimens ex-= posed in the sea water and 0.4 MPY for the specimens partially em- bedded in the bottom sediment. The corrosion rate for bare steel (AISI 1010) in sea water under the same conditions was 1.2 MPY indi- cating that the zinc coating was removed within a short period of time 93 to 4 months). The difference in corrosion rates in the bottom sediment was 0.7 MPY which shows that the zine coating pro- tected the steel in the bottom sediment for at least twice as long as it did in the sea water. There was no loss in the mechanical properties of the galvanized steel. The aluminized steel (1.03 oz per sq ft) was covered with white corrosion products, spotted with a few specks of red rust after 402 days of exposure at a depth of 2,370 feet. About 22 percent of the aluminum coating was corroded from the specimens exposed in the sea water and 40 percent was corroded from the specimens partially embedded in the bottom sediment; the underlying steel had not cor- roded. Therefore, it can be concluded, on a weight basis, that 1 oz per sq. ft. of aluminum will protect steel for a longer period of time than 1 oz per sq. ft. of zine; about 4 times as long in sea water and about 2 times as long when partially embedded in the bottom sediment. A titanium-cadmium coating on AISI 4130 steel was completely sacrificed and the steel was covered with a layer of red rust after 402 days of exposure at a depth of 2,370 feet. A 6 mil thick, sprayed aluminum coating which had been primed and sprayed with 2 coats of clear vinyl sealer protected the underlying steel for 1,064 days of exposure at a depth of 5,300 feet. After removal from exposure the aluminum coating was dark gray in color, speckled with pin point size areas of white corrosion products. Cast Irons The corrosion rates for the gray, nickel, nickel-chromiun, silicon, silicon-molybdenum and ductile cast irons at the two nominal depths in the Pacific Ocean are shown graphically in Figure 10 for sea water. 271 Reinhart There was no measurable corrosion of the silicon and silicon- molybdenum cast irons at either depth. In sea water at both depths the other cast irons behaved similarly to the steels as is clearly shown by comparing the curves in Figure 5 with those in Figure 10. The corrosion rates of the austenitic cast irons in sea water are shown graphically in Figure 11. The corrosion rates of these alloys in sea water also decrease with time of exposure at both depths with the rates at 2,350 feet being lower than those at 5,500 EGE The median curves for the two groups of cast irons and the alloy steels are shown in Figure 12 for sea water and in Figure 13 for bottom sediments. These curves (Figure 12) show that in sea water at a depth of 5,500 feet corrosion behavior of these three groups of alloys was the same after 750 days of exposure. There was a slight decrease in the corrosion rates of the three groups of alloys with time at a depth of 2,350 feet and the corrosion rate of each group was lower than that of its companion group at a depth of 5,500 feet. In the bottom sediments the behavior of the alloys was some- what erratic. The lower corrosion rates after 400 days at a depth of 6,780 feet is attributed to the fact that a greater proportion of each specimen was embedded in the bottom sediment than during the other three exposure periods at the nominal depth of 5,500 feet. The corrosion rates at 2,350 feet tended to increase slightly with time for the steels and austenitic cast irons while those for the cast irons increased sharply. The type of behavior for the cast and wrought alloys can only be attributed to their proximity to the water-sediment interface or the percent embedment in the bottom sediment. SUMMARY AND CONCLUSIONS The purpose of this investigation was to determine the effects of deep ocean environments on the corrosion of irons, steels and cast irons. To accomplish this, specimens of 47 different alloys were exposed at nominal depths of 2,350 and 5,500 feet for periods of time varying from 123 to 1,064 days. The corrosion rates of all the alloys, both cast and wrought, decreased asymptotically with time and became constant at rates varying between 0.5 and 1.0 MPY after three years of exposure at a nominal depth of 5,500 feet in sea water. These corrosion rates are about one-third those of wrought steels at the surface in the Atlantic Ocean at Harbor Island, North Carolina. The corrosion rates of these same alloys in sea water at a depth of 2,350 feet were lower than those at the 5,500 foot depth and decreased with time. In general, the corrosion rates of all the alloys exposed either adjacent to or partially embedded in the bottom sediments at the 5,500 foot depth decreased asymptotically with time and became constant at rates between 0.5 and 1.0 MPY after three years of ex- posure. The corrosion rates of the alloys in the bottom sediments at the 2,350 feet depth tended to increase with time. CAUCE: Reinhart The corrosion rate of steel was not affected by nickel additions to 9 percent at either depth. Silicon and silicon-molybdenum cast irons were immune to corrosion in deep ocean environments. The mechanical properties of the alloys were not impaired. The corrosion products of the alloys were composed chiefly of alpha iron oxide, ferric oxide hydrate, ferrous hydroxide and Beta iron (III) oxide-hydroxide. Zinc (hot-dipped) (1.7 mils) and titanium-cadmium coatings failed to protect sheet steel for one year of exposure. A hot-dipped aluminum coating (4 mils) protected sheet steel for a minimum of one year whereas a sprayed aluminum coating (6 mils, sealed) protected sheet steel for three years. The mechanical properties of anchor chains were unimpaired. However, sea water penetrated the forged sockets of one type of chain as evidenced by corrosion at the bottoms of the sockets. The mechanical properties of Type 304 stainless steel cables in sizes 0.094, 0.125 and 0.187 inch diameter were decreased by a minimum of 85 percent due to corrosion of the internal wires while those of the larger diameter wires were unaffected. The breaking strength of a Type 304 stainless steel cable coated with 90 percent copper-10 percent nickel was not affected. The breaking strengths of the aluminum coated steel wire ropes were unaffected. The bare steel, zinc and aluminum coated steel and stainless steel wire ropes were not susceptible to stress corrosion cracking when stressed at 20 percent of their respective breaking loads. However, the Type 316 stainless steel wire rope lost 40 percent of its breaking strength due to corrosion of the internal wires. The breaking strengths of bare steel, zinc and aluminum coated steel wire ropes, both stressed and unstressed, were unimpaired by exposure to deep ocean environments for periods of time as long as 1,064 days. However, based on visual observations zinc coatings corroded at faster rates than aluminum coatings on the wire ropes. ACKNOWLEDGMENTS The author wishes to acknowledge the generosity of Dr. T. P. May, Manager, Harbor Island (Kure Beach) Corrosion Laboratory, International Nickel Company, Inc. for granting permission to include his deep ocean corrosion data in this report. 273 Reinhart REFERENCES Iho 10. 11. Wo U. S. Naval Civil Engineering Laboratory Technical Note N-446 "Effects of the Deep Ocean Environment on Materials - A Progress Report" by K. O. Gray, Port Hueneme, Calif., July 1962. U. S. Naval Civil Engineering Laboratory Technical Note N-657: "Environment of the Deep Ocean Test Sites (Nominal Depth 6,000 feet) Latitude 33°46'N, Longitude 120°37'W" by K. O. Gray, Port Hueneme, Calif., Feb 1965. U. S. Naval Civil Engineering Laboratory, Unpublished Oceano- graphic Data Reports, by K. O. Gray, Port Hueneme, Calif. Dr. T. P. May, unpublished data. Dr. T. P. May, personal correspondence. Naval Research Laboratory Report 5153, “Corrosion of Metals in Tropical Environments, Part 3 - Underwater Corrosion of Ten Structural Steels" by B. W. Forgeson, C. R. Southwell and A. L. Alexander, Washington, D. C., August 8, 1958. Naval Research Laboratory Memorandum Report 1634, "Marine Corrosion Studies - Third Interim Report of Progress" by B. F. Brown and others, Washington, D. C., July 1965. Naval Applied Science Laboratory Report, ‘Corrosion at 4,500 Foot Depth in Tongue-of-the-Ocean,'' SROO4-03-01, Task 0589, Lab. Project 9400-72, Technical Memorandum 3, by E. Fischer & S. Finger, Brooklyn, N. Y., Mar 1966. Naval Applied Science Laboratory Report, "Retrieval, Examination and Evaluation of Materials Exposed for 102 Days on NASL Deep Sea Materials Exposure Mooring No. 1" SF 0099-03-01, Task 1481 and (SF-020-03-06 Task 1003) Lab. Project 9300-6, Technical Memorandum 4 (and Lab. Project 9300-7), A. Anaustasio, Brooklyn, No Woy Nov I9OD- Naval Research Laboratory Report 5370, "Corrosion of Metals in Tropical Environments, Part 4 - Wrought Iron" by C. R. Southwell, B. W. Forgeson and A. L. Alexander, Washington, D. C., Oct 1959. A. C. Redfield, "Characteristics of Sea Water," in Corrosion Handbook, edited by H. H. Uhlig. New York, Wiley, 1948, p. 1111. F. L. LaQue, "Behavior of Metals and Alloys in Sea Water," in Corrosion Handbook, edited by H. H. Uhlig. New York, Wiley, 1943. peo 384, pS. 390. 274 Reinhart 13. C. P. Larrabee. "Corrosion Resistance of High-strength-low-alloy Steels as Influenced by Composition and Environment ," Corrosion, Vol. 9, No. 8, Aug. 1953, pp. 259-271. 14. U. S. Naval Research Laboratory. NRL Memorandum Report 1383, "Abyssal Corrosion and Its Mitigation - Part II. Results of a Pilot Exposure Test", by B. W. Forgeson, et al, Washington, D. Gay Dec 1962. 215) Reinhart JOSUI UT 9IN}ONIIS NLS Sy sejaBuy soz i) awauany 4410q ‘VseOD OTFIOe FJO SOITS HTS surmoys deur eo1y bioging byuDS TE osiiat 276 Reinhart Oxygen, m1/1 8 2 4 6 3 Temperature, C 6.6 6.8 7.0 7.2 7.4 p pH 33.2 33.4 33.6 33.8 34.0 Salinity, ppt 12 14 16 7.6 7.8 8.0 34.2 34.4 34.6 Fig. 2 - Oceanographic data at STU sites 277 Corrosion Rate, MPY Reinhart VAN >< - Surface, Atlantic @ - Surface, Pacific, Panama Canal o - 5,500 Feet, Pacific, NCEL s: @ - 2,350 Feet, Pacific, NCEL =a O & - 5,500 Feet, Pacific, INCO @ - 2,350 Feet, Pacific, INCO o - 4,500 Feet, Atlantic, Plate, NASL po - 4,500 Feet, Atlantic, Disc, NASI. wo - 4,250 Feet, Atlantic, INCQ-NASL 72 tT O - 5,600 Feet, Atlantic, NRI D SAI ~ | aN i ™~Y. [ees wee A all — |Z 7\\ ~~ 200 400 600 800 1000 1200 1400 Exposure, Days Fig. 3 - Corrosion rates of low carbon steels at various locations 278 Corrosion Rate, MPY Reinhart Depth, Feet 5,500 2,350 Armco Iron Wrought Iron AISI 1010 Wrought iron, surface, Panama Cana] Mild steel, surface, Atlantic 200 400 600 800 1000 1200 Exposure, Days Fig. 4 - Corrosion rates of wrought ironand Armco iron 279 Corrosion Rate, MPY f Ww y 4 200 NY ~ 400 Reinhart >~€- Mild Steel, Surface, Atlantic o - AISI 1010, NCEL 4 - A387, NCEL 0 - HSLA No. l, VY - HSLA No. 2 W - HSLA No. 2 O- HSLA No. 4 @ - HSLA No. 4 % - HSLA No. 5 ® - HSLA No. 5 Q- HSLA No. 7 @-HSLA No. 8 9 -HSLA No. 9 NCEL NCEL 5,600 Feet, Atlantic NCEL INCO NCEL. INCO- INCO INCO INCO. - HSLA No. 10, INCU ©-HSLA No. ll, e@ - A36, NCEL INCG @ - AISI 4340, NCEL a@- AISI 4130, NAEC V/,-5,500 Feet WW - 2,350 Feet 800 Exposure, Days 1000 1200 1400 Fig. 5 - Corrosion rates of steels in sea water 280 Reinhart SJUSUITPeS UWI0}}10q 9} UT STee}S JO Sa}JeI UOTSOAIOD - 9 ‘BIA 00ZcT 72207 OSE*Z 7207 006‘S THON “9CV THON ‘OEY ISIV OONI ‘TI °ON WISH OONI ‘OL °“ON WISH OONI “6 °ON VISH OONT ‘8 “ON VISH OON1 ‘Z ‘CN WISH OONI ‘¢ °ON WISH THON ‘GS °ON WISH OONI ‘y °ON WISH THON “4 “ON WISH THON ‘Z °ON VISH THON ‘I °ON WISH THON ‘/8EV THON ‘OTOL ISIV mE, ee * Tp skeq ‘ainsodxqg OOOT 008 009 007 002 ey, i UH) -@ - B -&a a, =o -» -0 -@ -Q -~ = (9 -B = [0] = \v - oO AdW ‘99B8y uotsoi109 281 19}jeM 9S UT ST90}18S O0cT Ui Me™ LLIN QOOT 00 /, IOF ‘S}IUIIT DDUepTyU om, | Lp, 09 jus0red Gé ‘soa 00” 00 ? ie Uy, [Ds YORE ‘aainsodxq Uf RMI { ( ames EaaN) Ino [eomst1e1g - 2 SIA V y 282 Reinhart SJUSWITPSS WI0}JOG SY} UT PUe TOJeM LOS UL S[994JS TOF SOAIND TeOT}SsT}e}Ss UeIPSyT - g -3T_ skeq ‘oinsodxg ca N AdW ‘93eY UOTSOII09 Vv OSse*z & O 00S ‘°¢ JUSUIT pes 197eM qoog t “yadeq 283 I9}JOM LIS UT [99}S JO 9JeI UOTSOIIOD 9Y} UO TOYOTU FO 4D9FFA - 6 “B14 quaoisg ‘124° IN Reinhart qeog oLe‘z ‘shea zov =O qeeq ooe’s ‘skeq 7901 VV qeeq 0V9'G ‘shea 16L O OONI eS, RdW ‘92eyY UOTSOI1I09 284 hart Rein I9}JeM CIS UT SUOII 1SeD FO SO4eI UOTSOIIOD - OT 381A sXeq ‘oainsodxq eleval O0cT OOOT 008 009 00” (S > \ Z “ON eTTIONG T “ON eTFIONG “ON WNTUWOTYD-TeXOIN “ON uMTUIOTYD-TeXOTN T24TN fe) Kery OSG5 2) 00GeS OONI qeeq ‘yzdeq N RdW ‘a3eY uotso1109 Up Zz 0 0 a iL 285 19jeM 29S UT SUOAT JSe9 OT}TIUSISNe FO So9jJeET UOTSOLIIOYD - []T “SI q sfeq ‘ain 008 009 | re | Ty |,