NAVAL POSTGRADUATE SCHOOL Monterey, California THESIS ESTABLISHMENT OF HYDROGRAPHIC SHORE CONTROL BY DOPPLER SATELLITE TECHNIQUES by David H. Minkel June 1984 Thesis Advisors: L. D. Hothem D. Puccini Approved for public release, distribution unlimited 7223007 SECURITY CLASSIFICATION OF THIS PAGE (Whan Data Entarad) REPORT DOCUMENTATION PAGE I. REPORT NUMBER 2. GOVT ACCESSION NO 4. TITLE (and Submit) Establishment of Hydrographic Shore Control by Doppler Satellite Techniques 7. AuTHO«r«> David Henry Minkel » PERFORMING ORGANIZATION NAME ANO AOORESS Naval Postgraduate School Monterey, CA 93943 t. CONTROLLING OFFICE NAME ANO AOORESS Naval Postgraduate School Monterey, CA 93943 READ INSTRUCTIONS BEFORE COMPLETING FORM 3- RECIPIENT'S CATALOG NUMBER 5. TYPE OP REPORT 4 PERIOD COVERED Master's Thesis June 1984 6. PERFORMING ORG. REPORT NUMBER 8. CONTRACT OR GRANT NUMBER^,) 10. PROGRAM ELEMENT. PROJECT TASK AREA 4 WORK UNIT NUMBERS 12. REPORT DATE June 1984 13. NUMBER OF PAGES 14. MONITORING AGENCY NAME 4 ADORESSf// dlltaranl from Controlling Olltca) 15. SECURITY CLASS, (ol thlt report) I5«. DECLASSIFICATION/ DOWNGRADING SCHEDULE '« DISTRIBUTION STATEMENT (ol thlt Report) Approved for Public Release, distribution unlimited 17. DISTRIBUTION STATEMENT (ol lha abatract antarad In Block 30, It dlllarar\t horn Raport) IB. SURRLEMCNTARY NOTES 19. KEY WORDS (Contlnua on ravaraa aid* II nacaaamry and Idantlty by block numbmr) TRANSIT, hydrographic shore control, hydrography, point positioning, relative positioning, GE0D0P V, MAGNET, DOPPLR, MX 1502, datum shift NOS Accuracy Standards, IHO accuracy standards, translocation 20. ABSTRACT (Continue on ravaraa alda II nacaaaary and Idantlty by block numbar) The methods of Doppler Satellite surveying, as applied to establishing hydrographic shore control, are presented and evaluated. Both methods, point and relative positioning, are defined procedurally with the advantages and disadvantages of each included. The field operations of two Doppler surveys (Monterey and Lake Superior) are reviewed with regard to require- ments and procedures. A cost breakdown of the Lake Superior survey illustrates the high cost effectiveness of satellite techniques. The DO I JAM 7J 1473 EDITION OF 1 NOV «S IS OBSOLETE S'N 0102- LF-014- 6601 1 SECURITY CLASSIFICATION OF THIS PAGE (Whan Data Kntarac SECURITY CLASSIFICATION OF THIS PAGE (Whit Dmta EntarMQ results of four Doppler data reduction programs sin A «z = [N(l-e2)+h] sin 0 Conversion of x, y, 2 to (ft, A, h. Formula by B . R. Bowring p = (x2 + y2) / tan u = (z/p) (a/b) tan 3 5*3 CO- =te ~**" o\J co- co- s«s — k3 <-3 ^3- -- co 5 IN! ^3 *jy 1 — j 051 CO- "»B b1^ ^3 W- CO- >JB ^3 O c CO ug ^3 CO- =Ss U3 ^-3 ^^ 8*8 co- «3 -3 =SS: 0^ co- -3 ^3 ~- CO- 0^ S-2 % s/n 3] SURVEY 0PI -3 S*L CO CO- ia Jj- CA- CO u3 -^ % CO- CO ta «a w SCT CO co- ■a i— l ^it co- V) >a /n 168 MONTEREY CO CO. 3 ■■a r^ co< CO 5 col co •^3 OCT ^ co) CO «3 CO m T-R3 aw cud as 48 [Ref. 34,35] warned of possible difficulties due to multi-path interference at stations near water. Since the broadcast signals are relatively high frequency they could reflect off the water's surface. One station was selected inland to the east of the other stations occupied. This station was occupied so that the effect of network configuration could be evaluated with regard to multi-station solutions. The preceeding described the criteria used to determine whether a station was worth the effort required to attempt recovering the station. The above criteria yielded a list of approximately twenty-five stations. Some of these stations were recovered (or an attempt made) and then evaluated on the following site considerations: accessability, visibility, security, and power. At some sites, approximately seven, it was obvious from the general location that the station would not be suitable for occupation. No attempt was made to recover these stations. The major consideration was accessablity. The receivers required routine servicing in the form of changing data tapes, changing batteries, and checking the status of the receiver. Because the survey was to be performed by the author alone, accessability was a prime virtue. The equipment is not readily transported by one person in a single trip. The second consideration at a location was the horizon at the station. An obstructed horizon would cause a reduction in passes tracked. A horizon clear of obstructions 5° above the horizon in all quadrants was the preferred condition. This condition was not met at all sites. One station (504&4) had blockage to the east as high as 15 to 20° above the horizontal. This horizon criteria is a standard requirement for Doppler stations and is therefore not unreasonable. A data set which lacks an equal amount of passes 49 in each quadrant may cause a bias in the height and/or longitude of the station. The final consideration for site suitability was security. Due to the small, portable design of the receivers, they are easily stolen. The problem of security at a site was solved by one of two solutions. Either the unit was locked within one of four covered trailers leased from a local U-haul dealer, or stations were occupied on weekends when the sites could be camped on with the receivers. The station (50464) where the three reference marks were established was selected because it was extremely secure, and had 110 v AC power available. This site was used to verify that the receivers were in fact operating correctly. B. SURVEY OPERATIONS Four MAGNAVOX MX-1502 Geoceiver Satellite Surveyors were used to collect Doppler data at the various survey stations. One receiver was leased, the other three were on loan from NGS, MAGNAVOX, and the Maryland Dept. of Natural Resources. The period of the survey was from April to June 1982. All four receivers were not available for the entire survey period. The MX-1502 is a portable, 12 v DC, geodetic Doppler receiver designed for field use (Figs. 5 & 6) . Pass tracking is controlled entirely by an onboard microcomputer. The receiver is initialized and controlled via a key pad on the face of the instrument and data is displayed on a LED display window. The MX-1502 has various diagnostics for system status, and commands which allow the operator to determine the quality of the data being recorded. As a satellite pass is tracked, it is read into memory; after the computations 50 are performed (or attempted), the pass is recorded on a cassette tape. If a position computation was possible, the solution from the computation is also recorded. The cassette is not standard in that there is a clock track recorded on the back side of the tape. Data is only recorded on one side of the tape. As the data is being recorded, it is read back and compared to memory, bit by bit, to verify that the recorded data is correct. Approximately 70 passes can be recorded on a single cassette. In Monterey, approximately 3*5 days were required to obtain 70 passes. During operation at a site, the MX-1502 maintains two types of position based on passes tracked. The 2-D position is the position solution based only on the last satellite pass, only latitude and longitude are computed. The 2-D solution holds the height (input during initialization) fixed. This is the same form of computation that is performed in navigation type receivers. If a pass meets various criteria, such as: pass elevation, number of iterations in the 2-D computation, number of Doppler counts, and standard deviation of the residuals of the 2-D solution, it is used in the 3-D position solution. The position displayed is the culmination of all passes accepted for the 3-D solution. The update of the 3-D position is performed via a sequential adjustment using each newly accepted pass. The position computations are actually performed in X, Y, and Z; these values are converted to latitude, longitude, and height using the WGS-72 ellipsoidal parameters and stored geoidal map, and then displayed. The number of 3-D passes collected is a safe indication of how many satellite passes will be accepted for post-processing software packages. Therefore, it is a simple means of specifying the number of passes to be collected at a site. However, the criteria are specific to the MX-1502 and may not be similar in other receivers. Additionally, the 51 Magnavox MX 1502 Figure 5 52 1 Indicates voltage 2 Connects internal or external battery to meter 3 Indicates internal temperature 4 Desiccant absorbs internal moisture 5 External battery power switch 6 Operate/standby power switch 7 Fuse 8 Enters the code or data displayed 9 Numeral keys 0 thru 9 for entry of codes and data 10 Clear key 11 Change sign key 12 Space key 13 Back space key 14 Tape cassette transport 1516 character alphanumeric display Key to MX 1502 Figure 6 53 residual limit can be changed by key pad entries. If the number of 3-D passes is used to specify the number of passes to be collected, the alterable criteria should also be specified. Use of the tripod supplied with the unit would have been cumbersome since the tripod has no provisions for leveling the head, or for horizontal movement of the antenna. By use of an adaptor, antennas were mounted on surveyor's tripods, using a conventional tribrach. This allowed for quick leveling and plumbing of the antenna over the station marks. The unit comes normally with ten and twenty meter antenna cables, a connector is also included which allows joining the two cables. At one station a sixty meter cable was used, the cable was made by NGS for use with its unit. As stated before, the unit requires 12v. DC power for operation. During the survey, power was supplied by either using two 12 volt batteries in parallel, or by using a single 12 volt battery connected to a self-regulating battery charger (where power was available). On stations where the author camped with the units, a portable gas generator was used in conjunction with a battery charger to charge a single battery. The unit does have two internal batteries (gel cells) which are used to maintain memory and keep the oscillator on power. When a power failure did occur no data was lost (in memory), only passes available for tracking during the power failure were lost. The unit is designed to shut down when a minimum voltage is reached. It is important to note that this survey was conducted entirely by one person (the author) and consumed an average of 8 hrs per day. This points out the low man power requirements for surveying by satellite methods as compared to conventional methods. Units were visited and maintained on an 54 after class basis. Batteries were usually changed every two days, and data tapes changed every three to four days. The schedule was very tight, and did not allow for the monitoring of a pass with every visit to a station. This was not by choice, as monitoring of passes while tracking can indicate possible problems. Even so, little data was lost due to receiver failures during the survey. 55 VII. LAKE SUPERIOR SURVEY The Lake Superior Doppler survey was performed to establish hydrographic control for upcoming NOS surveys scheduled for the near future. The lake area is a splendid example of an area well suited for using Doppler satellite techniques. The area is densely wooded; with the forest beginning at the water's edge in most cases. The shoreline is rugged, generally rocky, and has occasional cliff faces rising up to 150 ft above the water. Accessability from the interior to the shore is poor on both the north and south shores. It was estimated by an advance party that to establish hydrographic control using conventional methods would require at least a full year with a crew of 8 to 12 men. It was at this point that alternate methods of establishing control were investigated by NOS personnel. In July 1982, the decision was made to establish the needed control via Doppler satellite methods. The survey was to be performed by the NOS Atlantic Marine Center (AMC), Operations Division. In late July a planning meeting was held at the AMC. The purpose was to qeview the project area and required sites, and discuss considerations which would have to be kept in mind during the reconnaissance stage of the survey. The meeting also served as a question and answer session since most of the personnel scheduled to perform the survey had no Doppler experience. Due to the dimensions of the survey area, and the requirement for good relative uncertainty (_± 1 m) within the control network, it had been decided to use four receivers (MX-1502) simultaneously. Two of the four would be 56 located on established first order stations, while the other two would be used to establish the needed shore control stations. In this way, all stations would be tied to one another through a few, common base stations. This common tie would allow computation of the relative uncertainties of all stations to one another. Based on a desired positional uncertainty of one meter or better, it was decided that at least 30 useable passes would be recorded at each of the stations to be established. Useable was defined for this project as a pass which had been accepted into the 3-D solution of the MX-1502 using the default residual limit values (0.25 m) . The 30 pass figure was used so that the desired postional uncertainty could be obtained from a point position solution (using the PE) if need be. The survey began in late August with the field unit (four men) conducting some of the needed reconnaissance. Some of the station marks were set at this time also. The author arrived on the evening of the 26th of August to replace one of the survey party members who had to leave, and to assist in starting the survey. The units were received and put on power on the evening of the 27th. The 28th was used to familiarize the other three men of the field party with the operation of the receivers. Since the survey party consisted of four men, four vehicles, and four receivers, when neccessary, each man could be relatively independent of the others. Independence was sometimes forced due to the size of the survey. The two fixed stations were on each end of the survey (approximately 200 miles) with each unit tended by an individual. The two mobile units were maintained by the remaining two men usually working together. These two worked together for efficiency and safety's sake. The two fixed units were set up and needed only tapes and batteries changed. Field operations commenced on the 29th of August 57 and ran continuously until the 28th of September. Baaed on the schedule of the upcoming hydrographic surveys, priority was given to the north shore and the area around Duluth, Minnesota. Operations started at the most easterly station on the north shore, and progressed westerly to Duluth, then easterly along the South shore, terminating on the Keewenaw Peninsula. Stations Finland and MCM 91 were used as fixed control stations through most of the survey. With Finland in the northwest corner of the survey and MCM 91 in the southeast corner the survey area was well bracketed (Fig. 7 & Table 2). As the work progressed to the south shore (Apostle Island area) it became necessary to start reconnaissance of more station sites. The original reconnaissance of the area had been done while the survey unit was working on another project, on a time available basis. The additional reconnaissance was performed by the three men working on the south shore. The normal daily schedule was to check the operation of the receivers in the morning, recon and/or set station marks, then return and recheck the receivers. At the latitude of the survey, it took approximately two days (an average of 44 hours) to track and record 30 useable passes. This allowed party members to perform two tasks (reconnaissance and receiver operation) at the same time, since at least every other day the units would not be moved. During the 31 day period of the survey, Doppler positions were established on 25 survey stations, covering approximately 420 miles of shoreline. With only one exception, all stations had a minimum of 30 3-D passes before the receiver was moved. 58 ■Kllta JlVi CM o> O III ? I'll1 ill' a • il ii !!!! I JJ i I ill! ! Li! I ii Si) iiilliilllililli ! II i 59 _, n - 0O ON C ■— cNc>ONa. crio>oc c ( c CT c 0 c CN CN m lt u- lO to un lO a" 1/1 IT) l/IU" LP LT LP u- LO it ILZ — i 1 ; , 1 =Sa S^ " *•**■ t^ i i =**» 5-S "'l* 6^ | COf <^3 =S* S^S r ■ rw — -v>r 1-0 , 'SI S^ ^ <^( ^ =* 1 !o^ l3 1 -a ">S v> O 193 =st 1 s^s ia 1 CO ^ 1 <0< s>« v> ■41 1 ' 1 -Hi ; i N isr U3 ! ON 0 1— I H ** nV \v> w3 =j» 5^ i :V 0> U5 ; =5* i c CjJ uy C-i 95Z =4 J-i > SJ =st 1 .1 i N zq 1 1 1 1 U5 5^ 3 i-3 u3 UJ > :*ti U( o1^ =*a iC ^ 0 CO CNI -:^t '-^> S>! •» • ■ =** ^3) • &vj as W ^^t «a • ^3 kJ TS3 *!* ^3 -^42 0 3 H 3 i ^5 5^9 (A « l3 ?3 H =** B^ L vfl 03 S frs<» ia =t» *g u3 o^J CO ^ ^8 ^3 1 sga fN <0 0 9 ^s B>4 CO -3 9*3 =55* n-i ■-S O J ^f ftvi 5< SQ -3 i 3 | >-3 <-3 c s3 W 1 ■-3 Si<0 M 0) • =at i« * a > w w 2 < -" :s; >-3 --" j— 1 oej ll: <£ < •z z < Z Kl J-< 3- ON < (33 cc 1— ( Q f- 03 0 H sc 0 a m k H > X *-r 2; 3 rn H a rn rs M 2 K O CO C < CO IE -< J J r" H Ed H c S Q Ed J < U 00 > O IK ON H < P-i 2 H Z ^ 0 c E§£ Z J t . tJ 3 C^ 1— 1 1 — Ph S CC Q C < u <: U ►_ M O BACH, pg. 4-25 Where: G = drms of vessel position P a = standard deviation of range y = angle of intersection of ranges a = (a2 + a2)1"2 t v r s Where: a = standard deviation of station position s in a coordinate axis a = standard deviation of range measurement r .2^2 then: a = /2 (a + a ) 2 esc y p r s To determine if incorporating the station error would have a significant effect on the ship's position, we solve for a with a set to .7m (the proposed specification) and set to zero. The difference will be the error in the ship's position due to the station position error. a = ( 1//2 a sin y) 2 - a r p s Setting: a = 10m (.1 mm at scale of survey, assuming 1:10000) a = . 7m s y = 150 (worst case) a = 3.47m r With a = 0.0m s a = 3.54m r Error Propagation Figure 8 65 approach to data processing. The interferometric method is to solve for the phase difference between a single signal received at two locations. This method requires that one receiver location be known therefore it can only be used in a relative positioning mode with two receivers. Preliminary results with program SADOSA, in the interferometric mode, show baseline differences with an RMS of + 1 8 cm. These measurements were made on a 39 meter baseline with two passes per solution [Ref. 37]. Further program testing , with data collected on longer baselines (up to 100 km), is not expected to show any significant difference with the preliminary results^. Because the interferometric mode requires a pass on each side of the observer's meridian, three or four passes may be required before an East-West pair is tracked. Therefore, this method of data reduction could reduce the required observation period, based on the specification proposed in this paper, to one third (8 hours or less) of the time presently required. C. IHO STANDARD This specification was also written to conform to the new (Nov 19 82) shore control standards of the International Hydrographic Organization (IHO). The IHO standards state. that when the shore control survey is extensive, the relative positions of control stations will not be in error by more than one half the plottable error at the scale of the survey [Ref. 38]. Using the proposed specification of 70 cm. in any coordinate axis, the relative accuracy of two Doppler stations is .99 m (1 sigma) or 1.07 m (CMAS ). The 9Personal conversation with Sz. Mihaly, Satellite Geodetic Observatory, Hungary, 1983. Circular Map Accuracy Standard 66 allowable relative error on a 1:5000 scale survey is 1.25 m CMAS. Therefore, the proposed specification, in the worst case (point position), will meet the IHO standards for shore control on surveys of 1:5000 or smaller (Fig. 9). The IHO standard further specifies that satellite (or astronomic) methods should be used to establish a point of origin for the geodetic network when there is no existing network. The requirement is that the origin should have a probable error of less than 60 m. The point of origin can be established by occupying the point for the period specified required by the 70 cm. specification. The resultant point position (either PE or BE) would meet the 60 m. requirement. A PE reduction would be preferred. It is not the purpose of this paper to recommend a new standard for all N0S hydrographic shore control. However, in the opinion of the author, the next logical standard would be a statement of acceptable positional accuracy based on the variance of the station position in any coordinate axis. The difficulty arises in that the present FGCC standards for geodetic control, do not address station error in this manner. An example of a classification system which does incorporate the error ellipse of a station into the accuracy clasif ication -is the system used in Canada (Appendix D). With the upcoming adjustment of the North American Datum, this type of classification system would be much easier to implement since much of the distortion in the current network will be removed. Until the positional error of a geodetic position can be inferred by its order of accuracy it will not be possible to specify all hydrographic shore control by an acceptable positional error. •67 c - a = standard deviation of control station position in a coordinate axis = . 7m (proposed specification) O = standard deviation of relative position a = (a2 + a20)h r xl x2 a = (.98) 2 = .99m (1 sigma) a CMAS = 1.0 73 a r r a CMAS = 1.062m r 2a (97% confidence level) = 1.98m r a = ^(plottable error) x (scale of survey) a - Js(.5m) x (5000) P a = 1.25m P Note: The assumption has been made that the standards are based on CMAS. No specific statement was made in the IHO standards in regard to the confidence level of the position. * Circular Map Accuracy Standard, 90% confidence level IHO Standard Figure 9 68 IX MONTEREY SURVEY RESULTS A. PROGRAM DOPPLR Redaction of the Monterey Doppler data was performed by the NGS, Astronomy and Space Geodesy Section using program DOPPLR (version NGS-03). The reduction was performed as a standard production run, no special procedures or options were used. All data collected during the survey was input for reduction. The reduction was performed with ephemerides for all five satellites. It should be noted that the ephemerides for all five satellites may not always be available. Table 3 is a summary of the datum shifts observed at the six triangulation stations which were occupied during the survey. The datum shifts shown are the origin shifts from the PE (NSWC 9Z-2) system to the local datum (NAD-27). If the PE spatial coordinates are converted to WGS-72 spatial coordinates, then differenced with the NAD-27 coordinates, the result is the datum shift from WGS-72 to NAD-27 for this area. Comparing these values to the commonly quoted shift values yields the difference in local datum shifts from the quoted mean values. This was done for station 50459 (Fig. 10) and yielded the following differences: ddx = -6 m, ddy = +6 m, and ddz = +6 m. Use of the predicted mean datum shift values in the Monterey area to perform a transformation would cause a position shift of 10 m from the local datum position. A difference of this magnitude would cause significant errors if a transformed position (to NAD-27) were used with already existent geodetic control to position a hydrographic survey; If point positioning techniques 69 are to be used to establish hydrographic shore control, the datum shift for the survey area must be determined by occupation of existent geodetic control. Table 4 is a summary of the point position reduction for the Monterey survey. Observations refers to the number of 30 second Doppler counts11. The geocentric coordinates shown are derived from the PE and are nominally earth centered. The height shown is the ellipsoidal height, not elevation of the station. The high rejection rate observed on station 50 462 is due to an error in the data collection. Some data (8 passes) observed at 50463 were erroneously marked as from 50462. When the data was processed, these passes were rejected due to the position misclosure. Removal of this data would bring the rejection rate to 4% . Station 50466 and 50467 both have two solutions summarized due to different occupations. In both cases, the antennas were not re-established close enough to the original antenna height to allow reduction as a single station. FGCC specifications require the antenna height be re-established within + 0.005 m of the original antenna height. All reduction programs reduce the data at the phase center of the antenna, then correct the final position to the survey mark. This means the data set must be subdivided if there are multiple "antenna heights otherwise the solution will have a high RMS. Table 5 is a summary of the station positions in the local datum. These are the positions as determined from conventional methods. Stations 50465, The observations RMS is the root mean square of the ranges which are computed from the 30 second Doppler counts. This value can be used as an indicator of the quality of the 'data. A value of .30 m or higher would indicate a poor data set. •70 £C — _ = X E - c-> O o in x ~ XZ — o ^r ri o o 2 i ps Ps os o c» ro M3 PS sO O — — Z os tcl s> <=> C^ — LJ — •— f-l •» CO (M r-i CN — ri r-i -O X ~o Ul 1 CM V y*> p>« p*. _ C-l ri p^ Ps o f-l co o PI CM p- Ps Ps os Ml a <=» CO c~ r^ CM o p-> rj ro Ps ^ i/) — os -o. o ■"■ rs. rs. n ►"■ O o Ul UJ •^ ^ ri UT p^ a CO Cs c- IDs Os uj w ITS r-» r-) ro ri P-) P- CD rj CI rj a o 03 Ml CO m CO UT Ml o- m CJ CM po o- o> c- o c- r j T u CP p-> o- r>i c~ U1 ^* o »- — © — Ul M cm <=> o o Psl CM 1— — UJ Z M3 CM Ps- r-j a s» «T n ^ UT rj r-j ri CI r-j r i r-i o >— u o -o -o M> -o sO M3 MD sO a ro p-) n ri ri PI PI PI ri PI PI bl CO -o CI , P"> o> CO c— Os _ — rs. (*M p^ ■v o — © Ps UT UT o- CO CS P» P» „ _ r-l r-l Ps Ps Ps o U1 rsi ■o wn CJ CM CJ CM pm se in <=> Pv ■v Ml Ul Ul UT UT Ul UT on w UT ~o ro n CO r. Ps rs. rs. rs. rs. Ul os CO wi ^ ~o 'O Ml Ml -O so >— CD rs. Pv p% p^ rs. rs. rs. P^ Ps rs. «r pi r-1 n n PI n r-l pi r-l PI PI z i— ■ Q O O r-> r« PI CM T » *H r-l r-> oc -o r-l p^ Os s^ -- >- x: OD CO O^ rs C4 r-i rj r-i (_> w 09 T O *— ■»— o O © © o © u-> VI p^ ao ~o --3 -o M3 •o -s* OS n n n n PO ri n PI r-l r-J r-l z i i UJ u ca T— 00 ■^ r-l — Ps *■ UT so Ps a ul U*l U) O ~0 r-l CO PM rs rs. 'O ^o CD C* — in CO m © >o CO O CO CO Cs CO X XT o- -o 'O P-* ■v UT CO P- O sO >o sO M3 sO sO CO o o pv ~o r» P» PN p> Ps Ps Ps Ps Ps PM CM 1 *>• *^* *^ ,v' p^ •^ *^ CM 'T- cn to -» s2 XT = o ce — ec Ml _ ^ P^ CO Ps Ps -a v V v rsi >* z o ■■ Os »■> ■ >» «a a ■=> ~s. ~s ^ ^ \ ~s \ ^s ^s UJ UJ 1 CO -o P^- o* CP tr-4 Ps. ^r CM CO co in i Ps o CO ao u-> CP Os ul a 3 i rv CO r^ Ps Ps CO r-i Ps T O i UT r-j C< *~ CI Ps r-j ~- — O r-i IT) ro CP o o- u-> Ps __ -o to Z ©■ n -O >o - i 09 ca CO CO CO OS CO 00 00 CO CO i=i o i Ss ^ ^^ ■^ ^^ ^s s s s N S o a: o i CP ps ^ o Pv n Os Os Os P-l IO to u-l ■^ Ul u-> u-> r-i Ul PI ut or >- i UJ C-J CM Ps ^r Ml CM >— U. I r-l ■^ PM p^ ■r CM CJ ro ^ CM Ul »— i i CM — rj os o _ c-t PO ■ -o -o ~o ^o M3 -o Ml -O Ml Ml w-> w» M-> UD — • o Os. n j c: o * j o p x - w O Z O CD P « U — Ps « C (— m cn uj j* mi 17. UJ h- UJ *- OC *>1 II X H O •ec O 3 ■— I CO — UJ UJ O l_J ui oe CD. ■ C£ 3B CO ul uj O E ul CO u. ul uj 3 uj X P- OC Q u 4 » — PM DZ Z UJ 0 3 PM o sC Ul Q « p- uj =1 i rM CO 3 C/J 73 «r> CO CO o 1— > »-l ' Ol O ^o o -o •c i o. rsi Pv T CN* -O ' 1-1 CD U-J rv* T 1 W-3 u-> ro n CO r. e> CO Wl *- ■« z N* ■v \n «i o- c~ b*> ca -o c~ -o «T OS o — | rv -a rN rs. o a: in r*j rv n c*. <_>>-— <=> •*r o O ■_ ro o- r»j n o z — u") bT r» o •o «r ro *-l n i-^ r-> v-i *r (N « O O- m O* u *■* •— CO rM -o o H K (N *r (M m UT X w O- o- -O ■^ m CO r^ o~ >o CO <=> o r. ■O rs r\ r^ <-* rsi CN CM rx rj a o •— r^ CO -o •» m to — CO m •o r. cs u. o -» — — r: — »-> i% o i-n rj _l UJ ~ r-j o* » —I = r^ UJ Z t— i CO T CO (N ~o rs* M — 1 C U3- n n n r) to n *— ■ z — r) «-» rj r>« r>j UJ ~- CO u o o o *=> o ■^3 r* r^ CO ton rsi -O -O Ol ^ ~ O m j Z _ 1 Q i C-l CM r*i c« CI O C*» ew -o ^ <-> 1 CO •-• at i -o ^ w ft M D UJ UJ *o i (ni O CO ' UJ — c i CO m rx ~o n ^o O "- i tan ^ m rj »— i «r ~o -o -o -a -o -o _l o CC i .. z z z z z z tn i3 > ce; o o o «o o <_i j o oe o o CJ r-» — UJ b"> -o ~o -o ~o ~o — CO -*• •■r ^ ■*" •*■ ^ 1 c o o o o o o h- 3 i ta-> m 111 lO ta? m to z i 0) M J-) c o en 0) jj CO c •H -o o o u CO Q CO a o >1 u CO cO H 74 indicator of the quality of the local geodetic network. Doppler derived positions can be expected to have an internal precision of approximately 30 cm (1 sigma) when observations were performed simultaneously and sufficient passes (30 or more) were observed [Ref. 39]. If one observes high variation in the coordinate difference values, the local network lacks internal precision. Changes of sign with the associated magnitudes seen in Tables 6 and 7 indicate the local geodetic network in the Monterey area lacks internal precision. This lack of precision is due, in part, to the stations not having ties to one another. This lack of consistency in the coordinate differences would not generally be found in geodetic stations which had all been established with the same survey. Table 8 can be used to determine if there is a scale difference between the Doppler coordinate system and the local geodetic system. By computing the baseline differences in parts per million for each baseline, and meaning these differences, one can detect scale difference. The standard deviation of the mean should also be computed to determine if the mean is realistic. Doppler (NSWC 9Z-2) and NAD 27 have a scale factor of about -0.5 + 0.04 ppm [Ref. 40]. A scale factor this small would produce negligible differences on a survey as small as the Monterey survey. B. PROGRAM MAGNET Reduction of the Monterey Doppler data with program MAGNET (version HP 80256) was performed by Mr. Robert Skeans, MAGNAVOX Corporation. Therefore, the procedures and options used are not as well known to the author as those used for the other reduction programs. The following discussion is based on the program output and program documentation supplied by Mr. Skeans.' 75 uj — -* — . ^3 3= u_ •— i CJ j_ |_J - X _i <=> — ■ o o en o= ua :c O =3 CO c-j o CM UT — O •o O o <=> o LO ro CO cm -0 o O 111 CO in Cv» ro CM -o ro cm >o ro •o -O -o ro T o ITJ rO o o o O b*3 o © o ■z. zz. © o © ^O o in o o- r-N r\ f--l T CO — r--. © M ^o CO -3 ci 1! rvi T IN ro O CO Li ro UT u-j t © — cm rv '."■J u-i » <3> <=> o O © © t- cm r\ o> o o n r^ U0 O UT CO ^O o o r j •>— © © © o» <=> « -o to — ' x= => UJ CJ O uj o: z: 3= s -l i— l i— < © <£ cc: XI C2 <=> T CO ^ m T 11 0) UJ ^ r^J »— < Q O a ce: Li "J ^3 ii •>- & L_ "O ZoO irt IU I - • C r-> :r © i a en ii CO ca "z. o CJ ! — o i— i _J -D — * — - 2; CJ ro uJ en _) CO UJ — ' CJ z oJ JC ■— » Lu XI u_ uj CJ i— O OS ►— _J ~ CJ O CO a: i=» ic r— co a: cn o ^j O Ch O o o o LO 'jO o CO -0 •T © 00 r»^ n i 1 o CD T if) > T4 o —— ~o »— ro •— o ro q CN T *— o o o O O O CI O rv CN CO •=r •T ro — u~> CO in Pv -o in T ro «~ •o -O -o -o ro ro ro ro o o o © — ro CM o CO r-^ *- ,— «r T — o- o O O © © o O o> O O o o O o CN .ro ^* in -o •c -C ■o o ■V «r ■^ ^ «r ^ © O o © © © UT U"J llO m m wo -O O- r^ ro uT CN CI I -o •w XI CO © © 1 ■o ao rv. •- r-j UT rs «T > ^ o o> n .— © © © © o o i O CO — O UJ •— i x: => UJ CJ a t—t i— i=a uJ OC C C x: CO B CM) DIFFERENCES (DOPPLER fllNUS OT-HER DDX DDT DJZ DB (M> (H) ifP (M) -28973.35 ■28972.7? -6298.05 -6298.00 2345.92 2345.4,' -5289.44 -5290.53 1431 .46 1431 .38 22675.30 22674.79 31319.27 31318.26 23683.91 23682.26 30404.81 30404. 1 7 8643.98 8643.47 35369.32 35368.74 31980.78 31980.54 53013.77 5301 4.06 63079. 44 63079.76 41334.29 41334.68 -3388.54 -3388. 18 17644.45 17645.32 27710.1 1 27711 .01 5964.96 5965.94 21032.99 21033.50 19570. 1? 19569.99 31819.6? 31819.37 6 2151.46 62151 .37 66971 .60 66971 .08 48027.3? 48027.39 1 224?. 50 12249.38 42581 .27 42581 .39 47401 .41 47401 .09 28457.20 28 457.40 30331 .77 30332.00 49733.65 49732.83 45551 .38 45551 .00 81723.73 31723.84 92153.07 92152.98 63381 .41 63331 .66 25994.26 25993.71 55726.01 55725.80 59796.92 59796. 4i 42069.53 42069.34 37909.39 37909.74 -.56 .05 .4t 1 .09 .06 .51 1 .01 1 .65 .64 .50 .58 -.40 -.36 -.98 .51 00 .32 3? .10 .55 4V .1? 35 50461 50463 DOPPLER 1003 1003 OTHER 50461 50464 DOPPLER 1003 1003 OTHER 50462 50463 DOPPLER 1003 1003 OTHER 50462 50464 DOPPLER 1003 1003 OTHER 50463 50464 DOPPLER 1003 1003 OTHER 1008.62 1007.48 7729.52 7729.38 -7635.36 -7436.00 -914.44 -914.09 6720.90 4721 .91 31098.66 31099.20 9353.51 9354.12 10065.66 10065.69 -1 1679.49 -11679.38 -21745.15 -21745.07 35151 .91 35151 .71 16207.70 16208.02 4820.14 4819.71 -14124.07 -14123.98 -18944.21 -18943.69 46944.65 46944.84 20246.56 20247.05 13522.20 13522.43 18350.37 18350.21 29612.58 29612.42 1.14 .13 .64 -.37 -1.01 -.54 -.61 -.03 -.11 -.08 .20 -.3: ,43 .09 .18 -.4? .23 16 ARITHMETIC MEAN STANDARD DEVIATION (RMS) H = 15 SPREAD MAXIMUM MINIMUM .39 -.35 .05 .09 .71 .43 .29 .37 2.66 1 .56 1 .04 1 .31 1 .65 .58 .52 .32 -1 .01 -.98 ■.49 Comparison of Baseline Vectors Table 8 78 To maximize the amount of data used in the reduction, a pass only had to be tracked at two stations to be accepted into the solution. In areas where all stations have good horizon visibility this would have little effect on the size of the final data set. In the Monterey survey, a noticeable portion of passes could have been excluded if the requirement had been that three stations must track a pass. This is due to two factors: 1) only four receivers were used in the survey and many times one was being transported 2) stations 50464, 50465, 50466, and 50467 had poor visibility to the east and would not "see" low level passes in that direction. The maximum RMS value for the errors of a posiiton fix was set at 17 cm. The position fix is the range from the satellite based on the 30 second Doppler count. The RMS of the six or seven 4.6 second Doppler determinations could not exceed 17 cm without the 30 second Doppler count being rejected. The frequency drift of the oscillators was not computed in the reduction. This condition was imposed because more than one receiver had been used on some stations. To accurately solve for receiver characteristics such as frequency drift and receiver time delay, station data sets must be subdivided into single receiver data sets. This would require that the subsets be processed as separate stations. MAGNET adjusts three parameters of the satellites' orbits. These orbital biases were constrained to 24 m, 4 m, and 9 m; for along track, height and cross track, respectively. Pass cutoff was set at 5° (above the horizon); no Doppler data below a 5° elevation is used. Furthermore, a pass was not used if the maximum elevation did not reach 14.5° [Ref. 41]. All other reduction programs with which the author is familiar use a 7.5° cutoff. As mentioned before, this cutoff value is specified to help minimize error due to tropospheric refraction. 79 The one sigma estimates of latitude, longitude, and antenna height are shown for each station in Table 9. These values are the uncertainties of the station positions relative to the other stations. The estimated standard deviation of unit weight is not output, so the validity of these estimated accuracies is not strictly known. Based on experience with other reductions these values do seem realistic. Baseline lengths determined by this reduction are compared with baselines determined by the other reduction programs in Table 22, section XIV. The results of this reduction may not be optimal. The major change in the reduction which should improve solution accuracy would be division of data sets into subsets of a single occupation. The improved uncertainties would be due to each subset having data from only one receiver obtained at a single antenna height. The errors induced by combining all Doppler data observed at a single station may have been somewhat reduced by the large size of the entire data set. Division of the data sets would have forced two reductions to be performed due to the station limit (10) of MAGNET. C. MX-1502 TRANSLOCATION PROGRAM To evaluate the accuracy and ease of performing field computed positions three data sets were reduced using the MX-1502 Field Translocation option. The processing was not performed while the unit was on site tracking. Instead, the computations were performed at a later date, after the data collection phase had been completed. In brief, the computation procedure is to input, into the MX 1502, the final 3-D positions, determined in the field via point positioning, of both the remote and control stations. The first acceptable seventeen passes are 80 -J —i r-» — • O O u-> CO — O* ON ON CN o m •-3 oo CN o in en a M e I CO J3 a £ O — IOX - 3 C tn O o at u. a IT) — © — r^ ♦ IO •» — 00 CM o UJ a u -o — ©» rv — ■» O IV — u-s — co cu •u CO C •H TD O o u ©- — *• o W» — u cu rH a a o Q -o CO •— cm cm •— CM o 9- CO ««■ ©» II UJ NMC C3) CM CD »- UJ — >- O CO I (J) CM ■ o a. O CO c CO 1-1 H - UJ UJ -I z ce O- UJ -O H- -O u. -O u. • U 0D » not - CM t/> a. o a cm ^ ~ -o cn uJ ea X 3E D Q£ X 3 x CD «r ►— O O _l - o c*"> o en o — ^ -O I—) — CO CN O o in — — ~> f^ — CM — CO c •H -a M O O U 0) LTl rv — CVI — -}■«- — cm — cm — m f\ ^ CN ^0 •*■ - UJ >— — CJ u, • U I C tf U1 O LkJ -o u. =c • UJ => o ae i— n CD -T PM rsi — r» — m — oo — ro — »r »-> — ro cm as o in 0* o> co m i> 0» n (N -o CO m in «■ -o O- O- pn to ro ITS o* i »S 0» — o» — • o pn in -o tn in P\. rv m po CM -O -o -o -o -O -o -o -O -O (M (N o »rv •O o- n -o GO o OD b- CO _-, o> PO a9 p** -O P»J »— ■o -— O r«l !•> in -o o rs. cm m o ▼ -O rs «r CD -O O CO HD (N ▼ m cm m ^. po PO pn ro ro m M ro *0 — -o l>1 — — — m rv. — p^ p^ p^ o o o •o "O ■o -o •c -o ~o ■o ~o r> o P-* CO rx o pn Pv CM CO CD m rs m © •»p CM ■o o co © ~0 CO o CO CM CO m o O- m p^ 1 in i (X i o- 1 o ■ o- 1 1 TM CO i o p^ C*J — O a. l-l UJ — fin rt *r m CO -r ■ o> CT> II CN >- ac o — v_^ u> aJ en cn O 0) • •u 03 H C X •H Ofi T3 •, • M ■o Lw U. O ^ in - <=> o — -i • s P— CJ (N = ■z> p^ »— !.) UJ — •« u rH H UJ UJ OJ -Q PM p^ ex. p-f 01 Q cm CD UJ a H 9. _ — £ o cm - UJ u-> -a u —J o ■x m UJ _l - H ro C*l — *T cr — — => x i: 1/5 — — - ~ <-> x: o — <£ UJ o ■« •St o o* _ CD CO m — — to-> — — O r-» ro r-j CO BT3 r j ui «r S3 OB o Ifl r^ «r — — > -» in <-n co « II ■O ~ •O I » SS -» Z -• u on «— » _l CO CO CO CO 14-1 x: en r» UJ E= rj .. •^ 3 o ~ |3 CO <4-| CC c •• O »o iO o CJ '_) <=> >, «r n ,, U T- co CO — C- £ ■a — C£ b O <» 3 O Ul CJ — co U — i — J. : •» 108 baselines of 500 or more kilometers will be affected by network configuration '3 . In either case, large or small survey, network configuration must be considered to obtain the best tie to the local geodetic control. B. ADDITION OF MORE DATA The accuracy of a Doppler position is inversely proportional to the number of passes observed. The solution usually reaches convergence at 30 to 40 passes for point positioning reductions. In most parts of the contiguous 48 states, 40 passes can be observed in 3 days or less. Comparing the results from the GEODOP V reductions shown in Table 20 with the results of the equation quoted from [Ref. 44], SIGMA = 150/(N(S-1 ) ) 1/ 2 one sees little difference (Fig. 18). The GEODOP V results shown are not optimum but are acceptable for this comparison. Therefore, one could use this equation to determine how many passes are needed to obtain a specified uncertainty. There are two factors which do need to be considered if this equation is to be used for predictive purposes; 1) only common passes are used in the equation and 2) the approximation assumes good data. In the field, one can either add additional passes as a safety margin or review the data at all sites to verify the number of acceptable common passes. The number of 3-D passes (MX-1502) could be used as an indication of the number of acceptable passes. The accuracy of the prediction may go down as one deals with smaller data sets (10 or less passes); especially as the number of observing stations decreases. 13 Personal conversation with J. Kouba, Geodetic Survey of Canada, 1983 109 FROM 2Q 50459 50460 50459 50463 40 SIGMA = 150/(N(S-1))1/2 N = 40 passes S = 3 stations SIGMA = 16.77 cm d* dy. .dz 17.13 15.10 11.59 17.33 15.33 11.81 FROM Jp_ U <£s dy 50222 50222 50231 306 91 14 28.91 27.65 23.31 23-91 SIGMA = 28 .35 cm Uncertainties (dx, dy , & dz) are in < centimeters d^ 18.93 19.10 Comparison to Estimate Figure 17 To estimate the obtainable precision of point position, precise ephemeris reductions with respect to the number of passes observed one can use the equation in Fig. 18. This equation is used in program CLASSI to determine the a-priori accuracy estimates of Doppler positions weighted to the number of passes [Ref. 45]. The weighting is based on results from 30 pass data sets. This equation also assumes high data quality; passes with range rate sigmas greater than 0.30 m cannot be used. The associated sigma values are based on experience at NGS in performing adjustments. There are two groups of values for the sigmas presented. The first group is for use when observations are going to be performed in a relatively short time period 110 (leas than 1 year). The second set is used when observations are going to be performed over a longer period. The higher value is required due to the long term variation in the PE coordinates as reported in [Ref. 46]. During the GEODOP V reduction of the Lake Superior survey, five 15 station reductions were performed. These reductions were primarily performed to test the program since it had been recently adapted to the NOAA Univac. The results are included only to show that there is a point where added passes do not significantly improve the solution. Comparing the sigmas of position differences between stations 50281 and 50299 (Fig. 19) to the sigmas of the first solution shown in the FGCC data set (Table 21 ) one observes little difference. The solutions are based on 255 and 78 passes in common, respectively. In fact, little difference is seen between the first and second cases of the FGCC results. Solution convergence for a point position solution is obtained at 40-50 passes. These results indicate the same is probably true for relative positioning, obviously it is reached at 70 passes. Due to periodic, systematic variation of the BE it is best to reduce data sets of 100 passes or less . Larger data sets could be affected by this variation. The best technique for reduction of large surveys is to reduce the data sets in small groups, then perform an adjustment of the entire survey using the subset solutions. The most practical method for subdivision of the survey data set is to reduce each group of stations which were observed simultaneously. The size of these data sets will be determined by the number of receivers used in the survey. One must bear in mind that to perform the adjustment all stations must be linked to one another through 14 Personal conversation with J. Kouba, Geodetic Survey of Canada, • 1983 111 Based on a-priori estimates of 30 pass solutions, for long term data sets, (1 year or more), where: a30 = 60, 80, and 100 cm (cf>, A, & ht.) passes aA a^ a 10 103 139 173 15 85 113 141 20 73 98 123 25 66 88 110 ax 103 139 85 113 73 98 66 88 a CT30 w (Total passes/30) 2 a w V V & ah a3o is the standard deviation in a coordinate axis based on 30 passes, J is the resultant estimated standard deviation in thecoordinate axes For short term data sets, (J30 = 30, 40, & 40 cm. passes a, a, a, (J> A h 10 50 70 86 15 42 66 70 20 37 49 49 25 33 44 44 Note: All values are based on a reduction with an 8 cutoff, and at least on pass in each quadrant. Point Position Accuracy Estimate Figure 18 112 d2 dz 7.62 7.46 8.17 7.43 From To passes dx 50281 50299 255 11.46 50222 50231 78 12.48 Uncertainties (dx, dy, & dz) are in centimeters Sigma Differences Figure 19 one or more stations. These linking stations can and should be the base stations. Additionally, all mobile stations should be moved at the same time. If on a four receiver survey two remotes were moved on alternate days, one could not subdivide the data sets into 4 station groups without cutting a station's pass data in half. 113 XII. STATION ELEVATION DETERMINATION The compuation of station elevation from Doppler measunnents is dependent on an accurate knowledge of the geoid. The geoidal slope of the area should also be known for the survey area if the results are to be optimum. This is the reason for the emphasis on occuping base stations which have ties to the NGVD. Because Doppler positions yield an ellipsoid height at each station one can determine the geoidal height to a reasonable approximation. The accuracy of the approximation is affected by the accuracy of the tie to the NGVD. By subtracting the elevation at a station from the ellipsoidal height one computes the geoidal height. Comparison of all geoidal height values at base stations allows one to infer the geoidal slope of the survey area. This inference can be degraded by large changes in the topography of an area. Assuming one can use the geoidal slope information obtained for the survey area, each station in the area can be corrected to yield an estimated elevation above MSL. These elevations can then be differenced to yield height differences between stations. This method (Fig. 20) is not accurate enough to replace geodetic leveling but should suffice for correcting the slope ranges of the hydrographic positioning system to horizontal ranges. As suggested earlier (sect IV. B. ) if base stations with ties to the NGVD can not be found, occupation of bench marks will yield this geoidal slope information. Bench marks should be selected to bracket this Doppler survey so that the inference of the geoidal slope is best suited for the survey. 114 Ellipsoidal height - Elevation = Geoidal height STATION 1LJL Elev. ^Ju 50459 -1.87 31.93 -33-80 50464 -12.95 21.25 -34.20 50461 -27.68 dGh = -34.20 - (-33-80) dGh = -0.40 m Baseline distances; 50459 to 50464 = 63383.45 50459 to 50461 = 45552.98 distance from 50459 to 50460 is approximately 72$ of distance to 50464; then, assuming a constant change in the geoidal slope: dGh = (-.40 m) (.72) = -.29 m Gh (50461 = Gh (50459) + dGh Gh (50461) = -33.80 + (-.3) = -34.1 m Elevation (50459) = Ellipsoidal height - Geoidal height Elevation (50459) = -27-68 - (-34.1) = 6.4 m Note: This method was used since there were two stations with known elevations. If there is only one station with an elevation, one must either assume a constant (level) geoid for the survey area or use some other means to determine the geoidal slope in the survey area. Station Elevation Computation Figure 20 A geoidal contour map can be used to indicate areas where there might be major variation in the geoid. However, most geoidal height maps do not have sufficient resolution to allow their use for obtaining geoidal height information. 115 Another method of obtaining geoidal height information is with NGSf program MCANAL. The program accepts latitude, longitude, and elevations for points of interest and outputs geoidal height information for these locatons15. Personal conversation' with M. Chin, Gravity, Astronomy, and Space Geodesy Branch, NGS. 116 XIII. FGCC TEST NETWORK RESULTS During the FGCC test of the Motorola Mini-Ranger Doppler Satellite Survey System in May, 1982 observations were also performed with three MX-1502's. The data that was collected by the MX-1502's was reduced with GEODOP V and is presented here as an example of attainable precision. The MX-1502 data was used for this reduction only because there is no input subroutine for Motorola data. The three station data set was processed using the procedures outlined in Appendix C. Simultaneity of pass data was enforced in all solutions presented. This did not cause much loss of data since all stations had good horizon visibility so nearly all passes were tracked at all three stations. Numerous runs were performed to optinize the results, selection of the representative solutions was based on the formal statistics of the solutions. Meterological data was not input. The baseline distances between the three solutions are known to a high accuracy. A conservative estimate of the estimated accuracy between the stations is 1:500000 (2 sigma) [Ref. 47]. All three stations have been tied (with first order methods) to the Transcontinental Traverse (TCT) network which has an estimated accuracy of 1:1000000. The conservative figure of 1:500000 yields an uncertainty of about 8 cm for the 42 km baseline, 7 cm for the 35 km baseline, and 4 for the 19 km baseline. Table 21 shows the number of passes in the solution, the sigma of the position differences, and the differences between the terrestrial standard 117 and the GEODOP V derived baselines. Also included is the estimated standard deviation of unit weight (SO) for each station. The optimum .solution would be with each station having an SO of 0.95. Some of the results should not be taken to be representative of attainable accuracy based on the number of passes. Specifically, the 5 pass solution; time and resources did not allow more research into representative accuracies for such a small data set. The solution shown was the only acceptable solution, based on the formal statistics, out of approximately 15 runs. The magnitude of the baseline differences for this set should not be considered typical for such a small data set. The magnitude of the baseline differences of all solutions are reflected in: 1) the proximity of the station SO's to .95 and 2) the spread between the SO's. Based on the SO's and the sigmas of the position differences the third, fourth, and sixth solutions would need to be redetermined before the results would be acceptable. The sixth solution appears, based on the SO's, to be an acceptable solution. Comparison with the standard shows otherwise. The variance for the solution is high based on the number of passes. This tends to indicate a weak solution and would be sufficient cause to rerun the reduction. Bear in mind that the worst error in this data set yields a proportional accuracy of 1:95000 (first order is 1:100000). The worst case presented in Table 21 shows a proportional error of 1:55000, this would be acceptable as second order, class I. Again, these are the worst cases and would have been reprocessed, based on the statistics, if time and resources would have permitted. 118 o o on — i 0 O O o tNI O o o —l o o o o j O o ON rH 0 o o o o o o (N| O o Q NT tN '0200 o o o o o O 0 o o o o o tN o o 0 tN O o o a o oo 0 -i pn en tN o ^T tN en o 00 m 3N en 5 NT ~T tN NT o NT tN NT tN NT oo o er- go o - tNI O ON tN tN o ON pn —i Pn O o 1 o o 1 o 1 O o 1 O o 1 oo o CSI tNI o -J tNI go en m O NT vO en NT en tN o o 1 o o 1 O O o 1 o o fN 00 en fN| tN| tN 00 ON en oo 00 o 00 NT o o 1 o ? o O o 1 o 1 o 1 en NT oo in pn o in m en m o p~ en -o en pn ON en on en ON o in 00 ON en rN pn O - u-l in < _ jl tn« connection t>«t-r**i*i vai'O""1* irn immtrntfot i.rj « mv%t 0« 'Ml tft»n r-C io-O 21 cm wnere C ri*» i»-g t »ss>gn«a lO' ir*« oravr SPECIFICATIONS Figure 1 Horizontal control surveys are classified as first, second, third or fourth-order according to standards of accuracy. The statistical concepts of standard deviation and confidence region are used to define standards of accuracy. These statistical concepts replace the concept of maximum anticipated error used in the Branch specifications issued in 1961 (See Appendices A and B) A survey station of a network is classified according to whether the semi-maior axis of the 95 percent confi- dence region, with respect to other stations of the network, is less than or equal to r = C (d + 0.2), where ns in centimetres, d is distance in kilometres to any station, and C is a factor assigned according to the order of survey. An ellipse bounding the 95 percent confidence region is shown in Figure 1. For first-order, the value assigned to C is 2. This means that for a station to be classified as first-order, the semi-maior axis of the 95 percent confidence region must be less than or equal to r = 2d + 0 4. Ellipse Showing the 95% Confidence Region of One Station Relative to Another (the area within which there is a 95 percent probability of the true relative position being situated). TABLE I VALUES OF C FOR HORIZONTAL CONTROL SURVEYS ACCORDING TO ORDER. USING r = C(d + 0.2) (r is in cm. d is distance in km) ORDER C 1st 2 2nd 5 3rd 12 4th 30 For two stations 10 km apart, r = 20.4 cm. For these stations to be classified as first-order, the semi-major axis of the 95 percent confidence region of one station relative to the other must be less than or equal to 20.4 cm. The values of C assigned to various orders of survey are shown in Table I (Figure 2 is a graph of r against distance See also Table II). As noted in Table II, the use of r = C (d+0 2) causes the parts per million (ppm) and ratio values to change significantly with distance, for short lines; this reflects practical considerations Experience shows that with most modern methods of establishing closely-spaced control, the overall pattern of error propagation — the combination of instrumental and centering e/rors. the effects of network configuration, and a host of other contributing errors, most of which defy individual identification — is not proportional to distance. The errors of measurement contributing to this pattern can be divided into two groups; those proportional to distance and those that are independent of distance. As lines become shorter, the second group becomes dominant. For the commonly used short-distance measuring instruments, the first group is dominant above three kilometres, and the second group is significant within the range zero to three kilometres. Therefore, these specifications are useful for surveys with points either closely or widely spaced or with a mixture of both 147 ' — _ __ __ -+- - — - - - j--- - - - --- f- - . [- — — V — !•-- - - _L ■ — ■4- - ■ - — j— / /. y ,< -^ ^ -- / ■v - 60<-n v r _/ - / Z y z -7? s / — /" s' > / ! ,* • '' ^x 5s: / lOcm vV> Z i *" LZ_ *"* '==~ ^ z_ ___rd_ ;z: 6cm .1? i ^ S, ■ "7 c^ *™ «*! • : ►5S2 , q j2^ r°< *^j ._ >(Z. __. -f - j *>• *VX " • * • «£>* icm .rtf ^ ^^ ! I!/ r • T " — ■ ■-- . A _ 1 i NOTE •*•» ;*10« i ogaf'""** 60m OOm 300m 600m I km 3hn d'dlstance between any two stations of the survey Figure 2 Accuracy Standards for Horizontal Control Surveys (bnsed on r C(d * 0 ?' \:itr re r ,<; ,r, , •■ niul -1 n krr.l 148 Where a survey network is disror*ed by constraint (inaccuracies in positions held fixed) examination of the ad|ustment results should be made beyond merely observing whether the error ellipses are within these accuracy standards This examination should include a study of the residuals and the relative shift m positions between free and constrained adjustments In computing standard error ellipses for networks under constraint, the computed standard deviation for unit weight from the adjustment should be used. Sometimes this means that stations, which would be classified as a first-order survey by an unconstrained adjustment, must be classified as lower-order until a general readjustment removes the distortion. Guidelines on network design and measurements are given in "Network Design" to assist in achieving the various orders of accuracy. However, it is stressed that by merely following the guidelines one does not ensure the achievement of the order of accuracy desired. The order can only be confirmed by an analysis of the survey results. NETWORK DESIGN The size and shape of the confidence region is dependent not only on the accuracy of the field measurements but also on the configuration of the control network. For a network to fulfill its basic role as a strong and reliable reference framework, it must be homogeneous, feature a reasonable number of redundancies, and the individual figures should be well-shaped Stations should be as evenly spaced as possible, and all a^.acent pairs of stations m the network should preferably be connnected by direct measurement The ratio of the longest length to the shortest should never be greater than five and usually should be much less. A basic principle of control surveys is to work from the large to the small, therefore, the spacing of higher-order control stations should generally be greater than that of lower-order stations In addition there should always t>e a sufficient density of higher-order control to govern the establishment of lower orders Frequently, these ideals cannot be realized Reality is often a network that has adjacent points which cannot be conveniently connected, that has large variation m lengths, and that has been measured with various instruments with significantly different accuracies The surveyor must design the network with these factors in mind. To design a network to achieve required accuracies, good a priori estimates of the accuracies of various instru- ments used with various techniques must be available. These estimates must reflect not only the consistency of several measurements of the same quantity by the same instrument, over a short interval of time under ideal conditions, but must also reflect normal random errors likely to occur in normal field use. under normal operating conditions by personnel who take only normal precau- tions. In addition, the estimates must take into account systematic errors that may not be evident in a normal survey, for example, an uncorrected zero error m Electronic Distance Measuring (EDM) instruments, sys- tematic meteorological errors due to imperfect measur- ing techniques, etc Appendix E lists typical standard deviations that may be expected under normal cir- cumstances and which may be used to compute weights in network design programs. Higher accuracies should be estimated if extraordinary precautions are taken m calibration and measurement. The accuracy of a horizontal control survey can be assessed properly from the results of a rigorous least-squares adjustment of the measurements Since this assessment can only be made after the field work has been completed, something more helpful is needed for those who wish to design networks and prepare measurement guidelines, and who require some reason- able assurance that a particular order of accuracy will be obtained when the field work is done. TABLE II ACCURACY STANDARDS FOR HORIZONTAL CONTROL SURVEYS (showing the variation in proportional accuracy over short distances) ORDER SEMI-MAJOR AXIS OF 95% CONFIDENCE REGION, r = Cld + 0 2). WHERE d IS THE DISTANCE BETWEEN ANY TWO STATIONS (or d = 0 03 km for d = 0 1 km lor d = 0 3 km for d = 1 0 km cm ppm ratio cm ppm ratio . cm ppm rano cm ppm ratio cm ppm (or d ■ 3 0 km (of d = 10 km rano 1 2 05 153 1 /5500 06 2 5 1 2 383 1/2600 1 5 3 12 28 920 1/1 100 36 4 30 69 2300 1/430 90 60 1/16700 1 0 1 50 1 .'6700 2 5 360 1/2800 6 0 900 WIIOO 15 0 33 1 /3000O 2 4 83 1/12000 6 0 200 1/5000 14 4 500 1/-000 36 0 24 1/41 700 6 4 60 1/16700 160 144 1/6900 38 4 360 1/7800 Ofio 21 1/46900 20 20 1/50000 53 1/18800 50 50 1/20000 128 1/7800 120 120 ' '8300 T>n i .imo "loo inn i -i inf- 149 The best c .rse of action s tc : "^.ate the pr&ooaea network in a suitable computer Digram such as GALS" using a prior; estimates for ;he standard deviations of the proposed measurements (see Appendix £) The results of such a simulation study, tempered with the wisdom of practical experience, usually provide a reliable indication of the accuracy likely to be obtained m the field For those not able to conduct computer simulation studies, some aids are provided in this publication • Appendix C provides measurement guidelines for the conventional methods — tnangulation. traversing and tnlateration — based on practical experience, and the results of computer simulation studies of simple idealized networks At best, these guidelines are a general guide only and must be treated with caution The reader should pay particular attention to the characteristics of the idealized networks depicted therein, to determine whether extrapolation can reasor.aci^ be n ajp from 'he gui'l' s 'o the : • at r~and The -eider sr^ uid a>so pa\ ittont.on 'o the explanat rv notes which follow the network sket hes in Appendix C Appendix D demonstrates some simple calculations that can be of benefit in estimating the accuracy of points m a network Appendix E lists typical standard deviations, stemming from practical experience, for distances, directions, azimuths and position differences measured using various instruments and methods of observation PHOTOGRAMMETRIC METHODS On occasion, horizontal control can be densified effec- tively using photogrammetnc methods (see Appendix F) *A Geo'lP'ic Si. 150 LIST OF REFERENCES 1. Umbach, M. J., Hvdrographic Manual Forth Edition. U.S. Dept. of Commerce, 1976. 2. Stansell, T. A., The TRANSIT Navigation Satellite System. Magnavox, 1978. 3. Jenkins, R. E. and Leroy, C. F., "Broadcast" Versus "Precise" Ephemeris - Apples and Oranges? Proceedings of the Second International Geodetic Symposium on Satellite Doppler Positioning, 1979. 4. Stansell, T. A., The TRANSIT Navigation Satellite System. Magnavox, 197 8. 5. Specifications to Support Classification. Standards of Accuracy, and General Specifications of Geodetic Control Surveys. U.S. Dept. of Commerce, 1980. 6. Stansell, T. A., The TRANSIT Navigation Satellite System. Magnavox, 1978. 7. Hothem, L. D., Vincenty, T. and Moose, R. E. , Relationship between Doppler and Other Advanced Geodetic System Measurements Based on Global Data. Proceedings of the Third International Geodetic Symposium on Satellite Doppler Positioning, 1982. 8. Ibid. 9. Hoar, G. J., Satellite Surveying, Magnavox, 1982. 10. Ibid. 11. Kumar, M. , An Unbiased Analysis of Doppler Coordinate Systems, Proceedings of the Third International Geodetic Symposium on Satellite Doppler Positioning, 1982. 12. Hoskins, G. W., Navy Navigation Satellite System Status. Proceedings of the Third International Geodetic Symposium on Satellite Doppler Positioning, 1982. 13- Ibid. 14. Boal, J. D., Guidelines and Specifications for Satellite Doppler Surveys, Canadian Geodetic Survey, 1980. 15. Chamberlain, S. , The MX 1502 Satellite Surveyor - Description and Use. Magnavox, 1980. 16. Brunnell, R. D., The JMR-2000 Global Surveyor. JMR Instruments, 1982. 17. Hothem, L. D. and McCune J., Report on Test and Demonstration of Motorola Mini-Ranger Doppler Satellite Survey System. Instrument Subcommittee Federal Geodetic Control Committee, 1982. 151 1 8. Proposed Revisions to the Federal Geodetic Control Committee Specifications to Support Classification. Standards of Accuracy and General Specifications of Geodetic Control Surveys, 1983. 19. Ibid. 20. Hothem, L. D. and Strange, W. E. , Doppler Satellite Positioning of Offshore Structures. International Hydrographic Review, 1977. 21 . Specifications to Support Classification, Standards of Accuracy, and General Specifications of Geodetic Control Surveys, U.S. Dept. of Commerce, 1980. 22. Stansell, T. A., The TRANSIT Navigation Satellite System. Magnavox, 1978. 23. Hothem, L. D., Robertson, D. S. and Strange, W. E. , Orientation and Scale of Satellite Doppler Results based on combination and comparison with other space systems. Proceedings Second International Symposium on Problems Related to the Redefinition of North American Geodetic Networks, 1978. 24. Hothem, L. D., Report on Test and Demonstration of Semi - Short - Arc Translocation Firmware for the Magnavox MX1502 Satellite Surveyor. Instrumentation Subcommittee, Federal Geodetic Control Committee, 1980. 25. Hothem, L. D. and McCune J., Report on Test and Demonstration of Motorola Mini-Ranger Doppler Satellite Survey System. Instrument Subcommittee Federal Geodetic Control Committee, 1982. 26. Brunnell, R. D., The JMR-2000 Global Surveyor. JMR Instruments, 1982. 27. Hoar, G. J., Satellite Surveying. Magnavox, 1982. 28. Jenkins, R. E. , Merritt, B. D., Messent, D. R. and Lucas, J. R. , Refinement of Positioning Software (DOPPLR). Proceedings of the Second International Geodetic Symposium on Satellite Doppler Positioning, 1979. 29. Ibid. 30. Kouba, J. and Boal, J. D., Program GE0D0P. Geodetic Survey of Canada, 1975. 31. Archival, B. A., A Comparison of Geodetic Doppler Satellite Receivers. The Ohio State University, 1982. 32. Ross, W. T. , MAGNET Magnavox Network Adjustment Post Processing Software. Proceedings of the Third International Geodetic Symposium on Satellite Doppler Positioning, 1982. 33. Hatch, R. , Chamberlain, S. and Moore, J., MX 1502 Doppler Survey Software. Magnavox, 1979. 152 34. Hoar, G. J., Satellite Surveying. Magnavox, 1982. 35. MX 1502 Satellite Surveyor Operation and Service Manual. Magnavox, 1982. 36. Umbach, M. J., Hvdrographic Manual Forth Edition. U.S. Dept. of Commerce, 1976. 37. Fejes, I. and Mihaly, Sz. , Interferometric Approach in the NNSS Data Processing. International Astronautical Federation XXXIV Congress, 1983. 38. IHO Standards for Hvdrographic Surveys and Classification Criteria for Deep Sea Soundings. International Hydrographic Bureau, 1982. 39. Proposed Revisions to the Federal Geodetic Control rinmnii i-.t^e Specifications to Support Classification. Standards of Accuracy and General Specifications of Geodetic Control Surveys, 1 983 - 40. Hothem, L. D. , Vincenty, T. and Moose, R. E. , Relationship between Doppler and Other Advanced Geodetic System Measurements Based on Global Data. Proceedings of the Third International Geodetic Symposium on Satellite Doppler Positioning, 1982. 41. Network Adjustment Computer Program MAGNET, Magnavox, 1980. 42. Vincenty, T. , Determination of the North American Datum 1 983 Coordinates of Map Corners, Dept. of Commerce, 1976. 43. Meade, B. K., NWL-10F Versus WGS-72 Doppler Results and Broadcast versus Precise Ephemeris Coordinates, Proceedings of the Third International Geodetic Symposium on Satellite Doppler Positioning, 1982. 44. Ross, W. T. , MAGNET Magnavox Network Adjustment Post Processing Software. Proceedings of the Third International Geodetic Symposium on Satellite Doppler Positioning, 1982. 45. Hothem, L. D., Vincenty, T. and Moose, R. E. , Relationship between Doppler and Other Advanced Geodetic System Measurements Based on Global Data. Proceedings of the Third International Geodetic Symposium on Satellite Doppler Positioning, 1982. 46. Strange, W. E. , Hothem, L. D. and White, M. , Time Variability of Doppler Results at Uklah Latitude Observatorvf Proceedings of the Third International Geodetic Symposium on Satellite Doppler Positioning, 1982. 47. Hothem, L. D. and McCune J., Report on Test and Demonstration of Motorola Mini-Ranger Doppler Satellite Survey System. Instrument Subcommittee Federal Geodetic Control Committee, 1982. 48. KalafU3, R. N. , Synopsis and Recommendations of the TSC Workshop on Differential Operation of NAVSTAR GPS. Dept. of 'Transportation, 1 983. 153 INITIAL DISTRIBUTION LIST NO. COPIES 1 . Director 1 Atlantic Marine Center, NOAA 439 W. York St. Norfolk, VA 23510 2. Commanding Officer 1 NOAA Ship WHITING Atlantic Marine Center, NOAA 439 W. York St. Norfolk, VA 23510 3. Commanding Officer 1 NOAA Ship PEIRCE Atlantic Marine Center, NOAA 439 W. York St. Norfolk, VA 23510 4. Commanding Officer 1 NOAA Ship MT. MITCHELL Atlantic Marine Center, NOAA 439 W. York St. Norfolk, VA 23510 5. Director 1 Pacific Marine Center, NOAA 1801 Fairview Avenue East Seattle, Washington 98102 6. Commanding Officer 1 NCAA Ship RAINIER Pacific Marine Center, NOAA 1801 Fairview Avenue East Seattle, Washington 98102 7. Commanding Officer 1 NOAA Ship FAIRWEATHER Pacific Marine Center, NOAA 1801 Fairview Avenue East Seattle, Washington 98102 8. Commanding Officer 1 NOAA Ship DAVIDSON Pacific Marine Center, NOAA 1801 Fairview Avenue East Seattle, Washington 98102 154 9. Director Charting and Geodetic Services NOAA 6001 Executive Blvd Rockville, MD 20852 10. Chief, Program Planning and Liaison (NC2) NOAA 6001 Executive Blvd Rockville, MD 20852 11. Chief, Nautical Charting Division (N/CG2) NOAA 6001 Executive Blvd Rockville, MD 20852 12. Chief, Hydrographic Surveys Branch NOAA 6001 Executive Blvd Code: N/CG 24 Rockville, MD 20852 13. Director Defense Mapping Agency HTC 6500 Brookes Lane Code: HO Washington, D.C. 20315 14. Director Defense Mapping Agency Bldg 56, U.S. Naval Observatory Code: PPH Washington, D.C. 20305 15. Director Defense Mapping School: Code: PSD-MC Ft. Bel voir, Virginia 22060 1 6 . Commander Naval Oceanography Command NSTL Station Bay St. Louis, Mississippi 39529 17. Commanding Officer Naval Oceanographic Office NSTL Station Bay St. Louis, Miss. 39529 155 18. Commmanding Officer Naval Ocean Research and Development Activity NSTL Station Code: 302 Bay St. Loui3, Miss. 39529 19. Commanding Officer Oceanographic Unit Four USNS CHAUVENET Fleet Post Office San Francisco, CA 96662 20. Hydrographer of the Navy Ministry of Defense Hydrographic Department Taunton, Somerset TA129N England 21 . Dominion Hydrographer Canadian Hydrographic Service 615 Booth Street Ottawa, Ontario Canada K1A0E6 22. Magnavox Advanced Products and Systems 2829 Maricopa Street Attn: R. Skeans Torrance, CA 90503 23. BLM Denver Service Center Cadastral Mapping And Training Stop 411 Attn: B. Ball Denver, CO 80225 24. Dept. of Energy, Mines, & Resources Earth Physics Branch Gravity, Geothermics, & Geodynamics Br. Attn: J. Kouba Ottawa, Canada K1A 0Y3 25. NOAA 6001 Executive Blvd Attn: N/CG142 - L. Hothem Rockville, MD 20852 156 26. IHO/FIG International Advisory Board Avenue President J.F. Kennedy Monte-Carlo, Monaco 27. Director Office of Marine Operations NOAA 6001 Executive Blvd Attn: N/MO Rockville, MD 20852 28. Chairman Dept. of Oceanography Naval Postgraduate School Code: 68Mr Monterey, CA 93943 29. Cdr. G. Schaefer Dept . of Oceanography Naval Postgraduate School Code: 68 Monterey, CA 93943 30. Cdr. D. Puccini Naval Postgraduate School SMC #2294 Monterey, CA 93943 31. Director Naval Oceanography Division Naval Observatory 34th and Massachusetts Ave. NW Washington, D.C. 20390 32. Chief of Naval Research 800 N. Quincy Street Arlington, VA 22217 • 33 • Chairman, Oceanography Dept. U. S. Naval Academy Annapolis, MD 21402 34. Commanding Officer Oceanographic Unit Five USNS HARKNESS Fleet Post Office New York, NY 09573 157 35. Dept. of Natural Resources State of Maryland Tidewater Administration Tawes State Office Bldg. Attn: C. Marple Annapolis, MD 21401 36. Lt. D. Minkel NOAA Ship WHITING 439 W. York St. Norfolk, VA 23510 37. Defense Technical Information Cente! Cameron Station Alexandria, VA 223 1^ 38. Library Naval Postgraduate School Code 0142 Monterey,, CA 93943 158 13 375 Oil i'Thesis IM5992 c.i Minkel Establishment ( hydrographic shore ( llte techniques. 0373 of ■re con- / 373 Thesis M5992 c.l Minkel Establishment of hydrographic shore con- trol by Doppler Satel- lite techniques.