US, FL INY Coast, Eng: : Os Reeth ; MISCELLANEOUS PAPER CERC-88-12 Res a COASTAL PROCESSES AT SEA BRIGHT ‘3; OF eneiecme TO OCEAN TOWNSHIP, NEW JERSEY VOLUME |: MAIN TEXT AND APPENDIX A by Nicholas C. Kraus, Norman W. Scheffner, Hans Hanson Lucia W. Chou, Mary A. Cialone, Jane M. Smith, Thomas A. Hardy Coastal Engineering Research Center DEPARTMENT OF THE ARMY Waterways Experiment Station, Corps of Engineers PO Box 631, Vicksburg, Mississippi 39181-0631 August 1988 Final Report Approved For Public Release; Distribution Unlimited Prepared for US Army Engineer District, New York 26 Federal Plaza New York, New York 10278-0090 Destroy this report when no longer needed. Do not return it to the originator. The findings in this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents. The contents of this report are not to be used for advertising, publication, or promotional purposes. Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products. TN 0 0301 0089758 3 Unclassified SECURITY CLASSIFICATION OF THIS PAGE Form Approved REPORT DOCUMENTATION PAGE OMB No. 0704-0188 Exp. Date: Jun 30, 1986 la. REPORT SECURITY CLASSIFICATION Unclassified 2a. SECURITY CLASSIFICATION AUTHORITY 2b. DECLASSIFICATION / DOWNGRADING SCHEDULE 4. PERFORMING ORGANIZATION REPORT NUMBER(S) Miscellaneous Paper CERC-88-12 6a. NAME OF PERFORMING ORGANIZATION USAEWES, Coastal Engineering Research Center 6c. ADDRESS (City, State, and ZIP Code) PO Box 631 Vicksburg, MS 39181-0631 8a. NAME OF FUNDING / SPONSORING ORGANIZATION US Army Corps of Engineers 8c. ADDRESS (City, State, and ZIP Code) 1b. RESTRICTIVE MARKINGS 3. DISTRIBUTION / AVAILABILITY OF REPORT Approved for public release; distribution unlimited. 5. MONITORING ORGANIZATION REPORT NUMBER(S) 6b. OFFICE SYMBOL (If applicable) 7a. NAME OF MONITORING ORGANIZATION 7b. ADDRESS (City, State, and ZIP Code) 8b. OFFICE SYMBOL | (If applicable) 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER 10. SOURCE OF FUNDING NUMBERS PROGRAM f PROJECT TASK ELEMENT NO. } NO. NO. 11. TITLE (Include Security Classification) j Coastal Processes at Sea Bright to Ocean Township, New Jersey; Volume TI: Main Text and Appendix A; Volume II: Appendixes B-G 12. PERSONAL AUTHOR(S) See reverse. 13a. TYPE OF REPORT 13b. TIME COVERED 14. DATE OF REPORT (Year, Month, Day) |15. PAGE COUNT Final report rrom_Jan 85 to Aug 86] August 1988 297 16. SUPPLEMENTARY NOTATION WORK UNIT Washington, DC 20314-1000 ACCESSION NO See reverse. 18. SUBJECT TERMS (Continue on reverse if necessary and identify by block number) Beach erosion Sand transport Shoreline change P| TT ey Coastaliprocesses Sandy Hook ferred | NeW Jersey Sea Bright 19. ABSTRACT (Continue on reverse if necessary and identify by block number) This report describes a study of coastal processes along the Atlantic Coast from Sea Bright to Ocean Township, New Jersey. Predictive tools were developed and a data analysis made to assist in evaluation and implementation of comprehensive shore protection plans for this highly utilized stretch of coastline. The study was divided into four broad areas as: (1) deepwater wave climate analysis and nearshore wave refraction, (2) numerical modeling of long-term shoreline change, (3) numerical modeling of storm-induced beach erosion, and (4) development of stage-frequency relations for the back bay and ocean coast. A literature review of previous studies and results of the four tasks are given in Volume I of the report. Volume II contains listings and interpretations of coastal processes data assembled in the study, including data on profiles, shoreline change, sediment sizes, and incident waves. The study represents an integrated attempt to quantitatively evaluate long and short-term coastal processes on a regional scale for use in engineering design. 21. ABSTRACT SECURITY CLASSIFICATION Unclassified 22b. TELEPHONE (Include Area Code) | 22c. OFFICE SYMBOL DD FORM 1473, 84 MAR 83 APR edition may be used until exhausted SECURITY CLASSIFICATION OF THIS PAGE All other editions are obsolete. 20. DISTRIBUTION / AVAILABILITY OF ABSTRACT GY UNCLASSIFIED/UNLIMITED ([] SAME AS RPT 22a. NAME OF RESPONSIBLE INDIVIDUAL OD oric users Unclassified Unclassified SECURITY CLASSIFICATION OF THIS PAGE 12. PERSONAL AUTHOR(S) (Continued). Kraus, Nicholas C.; Scheffner, Norman W.; Hanson, Hans; Chou, Lucia W.; Cialone, Mary A.; Smith, Jane M.; Hardy, Thomas (Volume I). Kraus, Nicholas C.; Gravens, Mark B.; Mark, David J. (Volume II). 16. SUPPLEMENTARY NOTATION (Continued). Volume II (Appendixes B-G) was published under separate cover. Copies of Volume I (main text and Appendix A) and Volume II are available from National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161. Unclassified SECURITY CLASSIFICATION OF THIS PAGE PREFACE The coastal processes study reported herein was requested by US Army Engineer District, New York (NAN) as part of a comprehensive plan of shore protection for Sea Bright to Ocean Township, New Jersey. This investigation was conducted by personnel of the US Army Engineer Waterways Experiment Sta- tion (WES) Coastal Engineering Research Center (CERC) during the period January 1985 to August 1986. Messrs. Silvio Calisi and James C. Urbelis were NAN Technical Monitors during the study, and Mr. Joseph Vietri and Ms. Lynn M. Koeth were Technical Monitors during the final review and publication phase of the report. Results are contained in two volumes. Volume I presents the main narra- tive, including problem statement, background information, procedures, and principal results. Volume II contains data sets that were processed and/or generated in the course of the study, including offshore and nearshore wave information, shoreline change statistics, and plots of beach profiles. Volume I was written by Dr. Nicholas C. Kraus, Dr. Norman W. Scheffner, Dr. Hans Hanson, Ms. Lucia Chou, Ms. Mary A. Cialone, Ms. Jane M. Smith, and Mr. Thomas A. Hardy. Dr. Hanson, Lecturer, Department of Water Resources Engineering, University of Lund, Sweden, was visiting researcher in Research Division, CERC, during the period August 1985 to February 1986. Volume II was written by Dr. Nicholas C. Kraus, Mr. Mark B. Gravens, and Mr. David J. Mark. Ms. Kathryn J. Gingerich assisted in preparation of the final draft of the report. Work performed in this study was under the general supervision of Mr. H. Lee Butler, former Chief, Coastal Processes Branch, Research Division and presently Chief, Research Division; Dr. Steven A. Hughes, former Chief, Coastal Processes Branch; Dr. Edward F. Thompson, Chief, Oceanography Branch; and Dr. James R. Houston and Mr. Charles C. Calhoun, Jr., Chief and Assistant Chief, CERC, respectively. Commander and Director of WES during publication of this report was COL Dwayne G. Lee, EN. Dr. Robert W. Whalin was Technical Director. CONTENTS Page IHINGD. Goddowoon od bdo OOOO COO OOD OO BDOD OOOO MOU CUCOD CU COU ODO GD O DOO O0D00N 1 RIESE OAM NHS he OOS Gb ob Gobo ooo OU O DUO UO OC ObUD OOD DOU ROOD DOOD OO m0!0 4 [ATS HORSE GURESHeaeteta crcterereraeiedeneicbeterotehe herve elclielekohelievolel cjeuelateponciencrclcienoremeieNeneneicne = 4 CONVERSION FACTORS, NON-SI TO SI (METRIC) UNITS OF MEASUREMENT.......... 1 PART I: IQKHUOMMUICHILO sabe ooooeucoobunadoooDdHOoD DDO CD DDD DD DDC O00 0b D00C 8 Sfefajare) Che Mol 5 Gobo anducoeObdoG0b000000000000000000000000000000.0000 8 Organiizatilonvofathils wReporeCryuscnctierevelelcierelereloloncielenereNavcleiereel-nehaterRonevolan 9 Units Of Measurements elererel nel ore cletale: clevelolevelevsienercdctellesetelonnenorencnenens 10 PART II: HESTORECAM VANDEL Xd'S ein GaCONDIGKONSvererercicrercisielekenereretcreicienenorenenenen 11 Orientation to the Study Area.... occ cree sess ensseens 11 Revilewvof (Previous) SCUGMESH sister cccielel eve shelchelic clieehercpenelislicl sel oltel sli) sicwouenons 13 PART leis SaWAV EMRE RRA CGTHONMANATAYS US cy varsia) cccpsl oleltel elelokeletebelenefelieheieheneichoreiehenonenonenere a4 Wave Hama Gasits. eieisycss. isscecho ea) sre) ehicla bio ceedice eheleicyahield sb wremebele Glove elie: esa tenereitells) aiets pl Nearshoremrebractions Simulation icvencicienel cioietelsueicheloloneloiotclicletekenerensteloneners 31 Refraction Over Beach Fill Borrow Sites...........ccecccccccccsone 39 PART IV: EONG=TERM) SHORE TINE CHANGE c's ccs: cic cle © «cle telcle ercieiciereisiencuctencioneiereienene 49g Lif ae) eX6¥0 [BY SH Ba Lo) alice inset) la edule a nis A MIL NR Ea ner cn INDE ells el a cS at es os cy 4g Shoreline Change Numerical Model.........c0...ccccesoccccensovccoe 50 Models Caliibrationirands Vieriticationiiicc.ciecielelcncelercielelcuctencichoneionetelsnoiotone 56 PART V: ‘STORM=INDUCEDEBEACHPEROSTONR a aciareciecsncec cao eee 65 INE EOAUGE TOM bcscysrsccus ovete apace eyskouere tabs iabaie ojala ute tole auc ue Rueecoee teh ITER ore ioree 65 BeachsErosiony Numerical Modellay sists cerarets cia cia eee oie 66 ModelsCalnibrationandiVieritficationicin sas. ccteyeete siete che eieerorl eee otis 71 SCO MES mulated omaecarcvtves casvs ye vaievons gaevanaus ie vaute Yong tare totrolevos ic ree ee oe 78 Evaluatvonnof Existing Conditions..... oc sc cml ciate cient 81 Evaluation of Alternative Beach Fill Designs.............seeseeeee 6 92 Variability sha Ctoric cies. tye cic) s cicis axsrsiercdate clei svaletel shspenoleccusteteiel wenermeneneteremonerers 98 PART. Vile STAGE=EREQUENGYRELATTONSHERS ey eccier cue cheleleveloreletaiecleusieroneiens icine 101 TmCrOdue cd Ombeysie a wievenenciele:ycecrevsivevaltercdaneane vateve eleileleiemecs bors: cuctecene wale ceneu en erateroneteiee 101 Revslewo fe ChemP EMP eS tudiyin cic c1rcicrersieeiie love: aiiorcleilerevereierevone rere eke) ovevereneiseoreta vere 103 Adapting FIMP Ensembles and Results to the Present Study.......... 106 SCORMESUNGEMMOMe De ee ia al arre. teu oneavels lala celignesere Coreh olla ieheteteveraiore reretetetletciometenenone 109 SUiMUAtTONVOL TS COMMS iene lovey ehaia wooilo le ove iolever ere lel eon eee 123 Devellopmentiof Stage-Prequency (Curves = oi). ci). «ier eee 1 « eletele\clelerctelelevels 125 REE ERENGES i, arsectesnctete: capatcuetedenaicrs cokcdsvor shatcuctalivs outa. shgkavere Seale ones core behe ROTO Peron 135 APPEND UM WAlcOmINO TATION yore tsnchcttesi cicets:tava vestevre lteers safe leliecay alee re aslo CTE SES TN eRe ETA aE ete Al APPEND EXE cae Wilt Sip HiLND CAS Tut'S UMMAIR Vt ep n stele tates tes er Joc une a aa ei Ue ee B1 * Appendixes B-G (Volume II) were published under separate cover. Copies are available from National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161. APPENDIX C: APPENDIX D: APPENDIX E: APPENDIX F: APPENDIX G: WAVE REFRACTION PATTERNS eoceoeoee eee eee eee eee oe eee oe eee eee eee MEASURED SHORELINE AND BEACH PROFILE CHANGE ............... OPEN-OCEAN BORROW SITES WAVE REFRACTION CHANGE AT BORROW SITES ............--eee00e BACK-BAY STAGE-FREQUENCY CURWIES jcqococooc bode DD DOO MUO D D008 Table Fw wow ona (S)] 12 13 LIST OF TABLES Summary of Selected Yearly Statistics and Properties of the WIS Hindeast for the North New Jersey Coast.........-eseeeee Summary of Frequency of Occurrence and Wave Height Characteristics from the WIS Hindcast for the North New Jersey Coast 6. ice men nccene sclcisielelss cicee «es Selected Deepwater Wave Periods and Directions................ Northern Beach Fill Borrow Site: Wave and Bathymetry Condaltilons Hor wRetractGilonMRunS\1.1. yo ens iorekorsyororoleyee 600600000000 Southern Beach Fill Borrow Site: Wave and Bathymetr Condiitionsmror RefiractilonwRunSkye)reieteleerteiocrcloneleletoretenedel kel oterenene Comparison of Calibration and Verification Results............ Measured and Computed Volume Changes (cu m/m)...............-- Computed and Measured Elevation Changes Revere Beach.......... Measured Changes in Contours Between Surveys Made in the Summer of 1953 and Immediately after the 6-7 November 1953 Storm (after Caldwell 1959)............... Construction of Synthetic Hurricane Ensemble for FIMP Study.... Parameter Values of Synthetic HurricaneS........cccceecccsceee Historical Northeasters Used to Generate Synthetic SULZCMEMUS Help CHEVICIIC Srey eneierey slolelienenelehaelevicrelene clei eel nelelofeielel bei Rolet- Flood Mark Comparison for the March 1962 Northeaster.......... Flood Mark Comparison for Hurricane Donna, September 1960..... LIST OF FIGURES Eocationemaperornmtne ss Gudymarea cry. -rcjcvesuerenocienereNoreiotortekeler Keleedelelsne Seawall and groins along Highland Beach looking nonthmcowanrdeSandyaHOoksrcrteney -ialeieserclveterelnonotenrcnonenorcicneheleelenelchet- Tele Seawall and groin at Sandy Hook (looking south)............... Aerial view of groin field at North Long Branch............... Impoundment at functioning groin in study area.............20% Historical longshore sediment transport rates (thousands of cubic yards/year) (after Caldwell 1966).................. Locations of phase stations for shallow-water wave information along the Atlantic coast, Region Z.............. Wave angle bands defined by WIS relative to the shoreline trend along Sea Bright to Ocean Township.............eeeeeee Grid used in the nearshore refraction calculation............. Nearshore bathymetry based on NOAA nautical chart No. 12324.... Comparison of calculated (a) wave height and (b) wave direction at the nominal 3-m depth alongshore for original bathymetry and bathymetry updated with 1985 DEOL Me RSURVEYS artterlenerolsierarcedokoerelcloreleroenen nenorclokonereteneeloncNhencess er ehe Location and horizontal dimensions of the northern BOGOW: SUCC aac, ere a crotencee eastereuecale ie cramrelC lett sucitone teste rsbenederenopeltoneweus tens Location and horizontal dimensions of the southern DOPPOW! Sue er ek es ieee ae alae recur el eet tetieme Tonevene were) eiiolieneie te renerensterehs 40 44 Figure Page 14 Change in wave properties at the nominal 3-m depth for waves refracting over the northern borrow site: (a) change in wave height, (b) change dine WAVERGIEC ET OMe sees eRe ezctereseece creterieee re ieis Ie ue ae GRU Siele 45 15 Change in wave properties at the nominal 3-m depth for waves refracting over the southern borrow site: (a) change in wave height, (b) change ANY WAVEKGIT CGtT Ome reper creme taeelete crete itentrah dace auaselts fous celerelats atelete'e aiebe «es 46 16 Results of calibration 1971-1982 (variable depth OL LCROSUME)) | cits agoususyeceisueisveveisholeletere eve cNeretle aie oleate ers felts odes RUSE ce aleve 59 17 Results of calibration 1971-1982 (maximum constant depth of closure) ives ee et cherie a uclues polis os ueeeecuepeee toes so ke uh a 60 18 Results of verification 1932-1953 (variable depth OL MCTOSURE Wests vese st Caeenats -cierets oie vabe sia iaeeteerelc: ata vole te te 5 setorshemeltehenes elec 61 19 Results of verification 1932-1953 (maximum constant Gepthy-of, CLOSUME) Gio esi eroncvecetve tone desevustebels he tees sents ois ietant Babe ace oes 61 20 Calculated transport rate capacities using a variable and a constant maximum depth of closure...........ccccccceees 63 21 Schematized) modell sareawns sens. .ei sist sie serene sR OR Shieh Ga es 67 22 Shape parameter, A, versus sediment diameter..........eceeee00. 68 23 Conceptuale seawall ner hecesi taste cteteltcuevefelekehs 9 sec Lacks long periods Typical year Typical year Many long periods Typical year Lacks long periods Lacks long periods Typical year Typical year Typical year denote average and maximum significant wave height, T, denotes the peak spectral wave period. p Table 2 Summary of Frequency of Occurrence and Wave Height Characteristics from the WIS Hindcast for the North New Jersey Coast Percent Occur. Average H, (m) Maximum H, (m) Notation: NNE NE ENE 0.0 {fot 0.00 0.37 0.00 Assit E 15.0 0.83 6.87 H, denotes significant wave height. 27 ESE 9.9 0.62 4.60 SE SSE S (ORG erred oe =malsi5 0.61 0.49 0.24 Wea 43209" eli 10 SANDY HOOK SHELTERED ENE (7.7%) E (15.0%) SEA BRIGHT -7.5° ESE (9.9%) -52.5° (10.8%) -67.5° (13.3%) SSE (27.9%) MONMOUTH MODEL BEACH UNSTABLE Figure 8. Wave angle bands defined by WIS relative to the shoreline trend along Sea Bright to Ocean Township 28 39. From Table 1 it is seen that the average significant wave height is 0.45 m and the average maximum yearly significant wave height is 3.70 m. The largest wave height, 6.87 m, which defines Event 1, occurred during the noto- rious 3 March 1962 "5-High" storm, which had a duration that lasted through 5 high tides. 40. Table 2 (see also Figure 8) shows that 61.9 percent of the hindcast waves originated out of the southern sector (from S through ESE), 15 percent were out of the east, and 7.7 percent were out of the east-northeast. The total is 84.6 percent. The remainder, 15.4 percent, represents the occurrence of completely calm wave conditions, for which neither sea nor swell existed. The zero occurrences in Table 2 for most of the northern sector are a result of wave sheltering by Long Island, New York, discussed in the next paragraph. Due to the bias for waves to originate from the south, the predominant direc- tion of longshore sediment transport at the project area is expected to be from south to north. This is in general agreement with the commonly accepted direction of transport along this coast. The local orientation of the shore- line and wave refraction over the irregular bottom must be incorporated for more quantitative discussion of longshore sediment transport. 41, Wave Sheltering. The WIS Phase III wave generation technique allows for wave sheltering by large land masses. In the present case, Long Island restricts the fetch of winds and propagation of waves out of the north directed toward the north New Jersey coast. The directional distribution of the potential wave population is modified in two ways if sheltering enters the hindeast. For wind seas, the energy within discrete direction bands is removed (zeroed) if the orientation of the sheltering land body would preclude propagation of waves in the band. For the swell component, all energy in the geometric shadow zone of the land mass is removed. For the north New Jersey coast, the sheltering effect of Long Island is greatest at Sandy Hook and decreases with distance to the south. The single WIS hindcast specifically performed for this study was assumed to be valid for the project reach; i.e., the differential effect of shadowing with distance along the coast was assumed to be negligible. Selection of representative wave conditions 42, The shoreline model (described in Part IV) requires input of repre- sentative wave conditions. Since the purpose of the model is to simulate 29 shoreline change occurring over several years, unusually high wave energy and low wave energy years in the hindcast were avoided. The consequences of severe storm events are treated with the beach erosion model, discussed in Part V. 43. The representative wave data set is to be used for both calibration and verification of historical (measured) shoreline change and prediction of future shoreline change. Since wave data are not available in either case, a deterministic interpretation of the input wave data set is not possible. However, one can obtain a representative data set with which certain para- meters can be varied to estimate the range of variability given by the model (the wave and shoreline models). To keep computer file sizes within manage- able limits, 3 years of data were selected from among the 20 years of avail- able hindcast data to produce the representative wave data set. 44, From consideration of Table 1, the consecutive years 1973 through 1975 were judged to be suitable to represent the total data set. The average significant wave height for these 3 years is equal to the average wave height for the 20-year total. One year, 1973, has an average significant wave height above the 20-year average, and each of the selected 3 years contains two of the highest ten wave energy events occurring among the 20 years. The average peak spectral period for these years ranges from 6-8 sec, which gives a spread of values around the commonly occurring 7-sec peak period. Since wave refrac- tion and shoaling are to a large extent controlled by wave period, this spread is a desirable feature. 45. As mentioned above, the chosen 3-year wave set possesses an average wave height equal to the average of the available 20 years of hindcast data, a characteristic compatible with the function of the shoreline change model. A complementary characteristic is that the data set includes slightly more high- wave energy events than average to account for realistic stormier conditions. It is noted that the 3-year record employed is much longer than records typi- cally reported in shoreline modeling literature, in which only a small number of representative conditions, such as seasonal averages, are repeated numerous times over the simulation period. 30 Nearshore Refraction Simulation Wave transformation model 46. An estimation of wave transformation from the nominal 18.29-m (60 ft) depth to the nominal 3-m depth (10 ft) along the coast was made by application of the Regional Coastal Processes Wave Model, RCPWAVE (Ebersole, Prater, and Cialone 1986). RCPWAVE was specifically designed for use in projects with large spatial extent, such as in the present case. This model is superior to classical wave ray refraction procedures in that energy propa- gation along wave crests due to irregular bathymetry is accounted for in addi- tion to energy propagation in the direction of ray travel. The model is also more efficient than wave ray models since the governing equations are solved directly on a user-specified depth grid in the horizontal plane (by an itera- tive finite-difference solution scheme) rather than by ray shooting and inter- polation to the grid. 47. Basic assumptions used in RCPWAVE are: Mild bottom slopes Linear, monochromatic, and irrotational waves Negligible wave reflection 1a 10 Io |p Negligible energy losses due to bottom friction or wave breaking outside the surf zone 48, These assumptions are common to most numerical models used for engineering application. Results from the model are expected to be adequate for estimating longshore sand transport rates. 49. Model Grid and Boundary Conditions. The main model grid (Figure 9) is rectangular with a mesh of 106 cells across-shore (positive x-axis directed offshore) and 240 cells alongshore (positive y-axis directed north). The cell spacing alongshore was set at 150 m; the cross-shore cell spacing was 75 m. This grid spacing is fine-meshed for wide-area modeling, but it was considered the minimum necessary to resolve any systematic irregularities in the breaking wave pattern that might be induced by unusual bottom features in the area. 50. The main grid covers the coast from Sandy Hook navigation channel on the north to a point just beyond Lake Como. Across-shore, the grid extends from inland of the present shoreline seaward to approximately the 20-m con- tour. The grid was extended well outside the Sea Bright to Ocean Township project area to eliminate possible contamination from the lateral boundary conditions. 31 S) c o Z < = DX =75m DEAL LAK EXTENSION Figure 9. Grid used in the nearshore refraction calculation S2 51. At a later stage in this study, the grid was extended on the south to permit simulations of refraction over potential beachfill borrow sites located off the township of Belmar. The 106 by 39 grid cell extension began at Lake Como and ended at Manasquan Inlet. This extension made possible the incorporation of three profile lines from the 1985 survey not located on the main (original) grid. Revision of the original grid to incorporate the 1985 profile survey information is described below. 52. The grid was overlaid on National Oceanic and Atmospheric Adminis- tration (NOAA) nautical chart No. 12324 (Edition 22, dated January 1984) to assign an average depth to each cell, interpolating as necessary. The depths at the centers of grid cells were transferred to a computer file for use as model input. A 3-dimensional plot was made of the bathymetry to provide a visual check (Figure 10). 53. Wave height, direction, and period as determined from the Phase III WIS hindeast provided the required offshore boundary values for RCPWAVE. The default lateral boundary condition is a "no-flow" condition equivalent to specifying a plane beach at the sides. RCPWAVE was modified to save values of a transformation coefficient related to wave height (described further below) and wave direction at fixed points alongshore at a nominal 3-m depth for input to the local wave refraction and breaking simulation routine employed by the shoreline change model. Model runs 54. Test runs were made to verify proper operation of the model. During the testing phase, it was noted that on some runs, waves propagating northward, approximately parallel to shore (out of the south, Table 2), caused model instability. Examination of this condition indicated these waves would refract offshore at certain locations, a possibility not accommodated by RCPWAVE and of no physical significance. In order to represent waves from the south without concern for rare exceptions producing model instability, input Phase III waves out of the south were reassigned to the southeast sector. 55. Production refraction model runs were made in an innovative way to eliminate the expense of making a computer run for each offshore wave condi- tion in the WIS 3-year time series. Direct use of the time series of wave height, direction, and period at 6-hr intervals would require 8,760 runs (sea and swell). The cost to make this enormous number of runs and to store the results would be prohibitive. The procedure would also be misleading since 33 SEA BRIGHT BATHYMETRY ELEVATION (FT) MLW -100 Figure 10. Nearshore bathymetry based on NOAA nautical chart No. 12324 (Edition 22, dated January 1984) 34 the accuracy of the shoreline change model is not compatible with such detailed information. Instead, 56 deepwater wave conditions were selected to represent the wave climate. The 56 conditions are defined by all possible combinations of wave periods and directions given in Table 3. The directions given in Table 3 are with respect to an imaginary line normal to the trend of the shoreline. Thus, a wave direction of 0 deg would correspond to waves incident normal to the coast. Table 3 Selected Deepwater Wave Periods and Directions Pentodare i ia a i Direction (sec) (deg) 4.0 22.5 6.0 eo (35 -7.5 BE5 -22.5 9.5 -37.5 10.5 -52.5 W245 -67.5 12.5 Note: Directions defined in Figure 8. 56. A unit (1-m) wave height was used for each calculation period and direction combination. Since the model RCPWAVE is based on linear wave theory, the transformed unit wave height can be interpreted as the product of combined refraction, diffraction, and shoaling coefficients (called a trans- formation coefficient here). The actual value of the wave height at a partic- ular grid point is obtained as the product of the transformation coefficient and the deepwater wave height in the WIS time series. Thus, although deep- water wave period and direction were described by a limited number (56) of combinations, the wave height appearing in each wave condition in the 3-year time series was utilized. 57. The output of the production runs consists of the transformation coefficient and wave direction at the nominal 3-m depth at each of the 240 longshore grid lines. In linear wave theory, employed by RCPWAVE, the wave 35 period does not vary in the refraction process. The results from all model runs were compiled into one random access file keyed on input wave period and direction. Knowledge of the deepwater wave height associated with each set of WIS wave conditions allows rapid calculation of nearshore wave properties. Plots showing the results of selected refraction model runs were made to further verify proper operation of the model and to give a visualization of the results. Selected plots are contained in Appendix C. Time series processing 58. A program was developed to link the 3-year time series of the hindeast wave results at the 18.29-m depth to the results of the wave model runs to create a time series of wave height, period, and direction at the 3-m depth at each longshore grid line. The program reads one record of WIS data (height, direction, and period of sea and swell components) and defines a key, based on input period and direction, for both the sea and swell component. The keys are then used to enter the random access file and extract transformed wave conditions. The transformed wave height at each grid line at the 3-m depth is obtained as the product of the transformation coefficient and the deepwater wave height in the WIS record. The sea component is limited by the calculated TMA limit (Hughes 1984), which is the maximum wind sea that can occur at a given depth and period. (The acronym "TMA" derives from locations of field experiments used to verify the theory.) 59. The shoreline change model requires input wave conditions at 6-hr intervals. Therefore, every other record of the hindcast time series was analyzed. To conserve computer memory and file space, only the component (sea or swell) with the greater energy flux per unit length of shoreline was saved in the output file. If both components had wave heights of less than 20 cm (0.66 ft) at the seaward boundary of the grid, no results were saved because neither component is expected to have sufficient energy to produce significant longshore sediment transport. In the shoreline model, this condition was treated as a calm wave condition of effective zero wave height. This tech- nique of imposing a threshold, although somewhat arbitrary, resulted ina sig- nificant reduction in calculation time. Analysis of the original and pro- cessed time series showed that the energy flux of the processed series was 24 percent less than that of the original series due to the removal of the nu- merous low wave energy events. Field experiments performed by CERC subsequent to the present study support the concept of a threshold in longshore sediment 36 transport of the magnitude employed in the model (Kraus and Dean 1987). 60. The final output from this program is a sequential file that con- tains a 3-year time series of effective wave heights, periods, and directions at 6-hr intervals at the nominal 3-m depth for each of the 240 longshore grid lines. This file constitutes the principal wave input for the numerical model of shoreline change. Revision of nearshore bathym- etry and additional refraction runs 61. At the request of CENAN, in May and June, 1985, staff of CERC's Field Research Facility (FRF) in Duck, North Carolina, made beach profile surveys in the project and neighboring beach area. Many of the control points used in the 1953 (CE 1954) and 1961 surveys were reestablished. In coordina- tion with CENAN and CERC personnel involved in the numerical modeling tasks, profile surveys were made at a density of approximately four lines per mile of shoreline in the stretch between the north end of Sea Bright and the south side of Shark River Inlet. In all, 68 lines were surveyed by the FRF in 1985. These lines extended from the control point well onshore, typically landward of the seawall, to a nominal depth of 9 m (30 ft). Elevations were measured relative to the National Geodetic Vertical Datum (NGVD). Sixty-five of the 68 lines are located on the main refraction grid. The positions of the survey lines were converted from the original Easting and Northing coordinates to latitude and longitude and transferred to NOAA chart No. 12324. Depths from the surveys were read on to the grid and the bathymetry updated. Limited extrapolation and interpolation were done to update grid values immediately adjacent to the survey lines. 62. The 1985 profile data became available after the processed time series and 56 refraction runs had already been generated. In order to deter- mine whether it would be necessary to repeat this considerable effort for the updated bathymetry, five refraction runs were made with unit wave height for the pre-1985 (original) and 1985 (updated) bathymetries. Selected pairs of wave period and direction were 8 sec, -67.5 deg; 10 sec, -52.5 deg; 8 sec, -7.5 deg; 5 sec, -22.5 deg; and 8 sec, -22.5 deg. Wave transformation coef- ficients and directions for each of these period-direction pairs were output at a nominal 3-m depth for the 240 longshore grid lines of the main grid. 63. Example comparisons of calculated wave height and direction are given in Figure iia and Figure 11b, respectively. In these figures, the 37 CHANGE IN HEIGHT (m) MAIN MODEL GRID WAVE HEIGHT COMPARISON | INCIDENT WAVE ANGLE: -7.5 deg INCIDENT WAVE PERIOD: 8.0 sec | | 0.0 ORIGINAL BATHYMETRY: UPDATED BATHYMETRY: ~—~-~~--- | € 1.5 fas La 4 eis oO 1.0 Ww ac W os .¢ = 0.0 -2.0 0.0 50.0 100.0 150.0 200.0 750.0 CELL NUMBER w.0 MAIN MODEL GRID WAVE ANGLE COMPARISON INCIDENT WAVE ANGLE: —-7.5 deg INCIDENT WAVE PERIOD: 8.0 sec | q 4 0.0 4 ORIGINAL BATHYMETRY: | UPI DIATE DA BAT HiyiMiEsts Riviere > ao = W J o = ¢ WW > o-¢ = 0.0 50.0 100.0 180.0 200.0 250.0 — CELL NUMBER Figure 11. Comparison of calculated (a) wave height and (b) wave direction at the nominal 3-m depth alongshore for original bathymetry and bathymetry updated with 1985 profile surveys 38 CHANGE IN ANGLE (deg) lower plots show actual calculated quantities and the upper plots show the difference in quantities. Since the bathymetry in the study area is irregular and the shoreline jagged, wave height and direction vary appreciably along- shore. However, trends are seen to be similar for the lower plots on each figure and, importantly, the difference in calculated quantities is random, that is, no systematic change in wave properties resulted from the updated bathymetry. Therefore, it was concluded that use of the wave time series and refraction results for the original bathymetry is justified. The updated bathymetry was employed in the later stages of the project for calculating refraction over beach fill borrow sites. Refraction Over Beach Fill Borrow Sites 64. For an open-ocean coast, breaking wave height and direction are considered to be the primary factors controlling longshore sediment transport and subsequent beach change. The pattern of breaking waves is determined by the properties of the incident waves in deep water (wave height, direction, and period) and the bathymetry over which the waves propagate and transform. Alteration of the nearshore bathymetry due to removal of sediment at an off- shore beach fill borrow site may alter the breaking wave characteristics. The sediment transport rate along the beach could, in principle, be modified to such a degree that the naturally occurring evolution of the beach plan shape would be changed by an amount sufficient to have engineering significance. 65. Two open-ocean sites are under active consideration as borrow sources for the project beachfill. The locations and configurations of the potential borrow sites are indicated on Figures 12 and 13. Herein, the borrow area off Sea Bright and Sandy Hook is called the northern site, and the area off Belmar is called the southern site. A description of the two sites and the available deposits is given in Appendix E. The bathymetry in the vicinity of the southern site is highly irregular (Figure 13), whereas regular bottom contours could be drawn for the northern site (Figure 12). 66. The sites lie relatively close to shore and in water depths rang- ing from approximately 25 ft (8 m) to 60 ft (18 m) MLW. An 8-sec linear wave at these depths has a length of approximately 213 ft (65 m) and 282 ft (86 m), respectively, and the corresponding depth to wavelength ratios are 0.12 and 0.21. These values are much less than 0.5, the ratio at which waves are 39 (MIN 43 UT Sunoquood yydap) a4TS MOJJOq UJeYyQUOU 944 JO SUOTSUDUTpP TeqUOZT4uOY pue UOTAeOOT °2| aunty 999¢ ——— | 134 0092 (0) ea 9992 ave LHOIYS VAS wINOZ 1VOILIYD.,, SGNV1THDIH OOH AGNVS 40 (MIN 33 UT 4Qdap) a41S MOJI0g Uday NOS ayy JO SUOTSUSIITP TeqUOzZT4uOY pue UOTIeOOT “€| SunsTy 14 0052 0 00Sz 09 LG co yYVvW1sd YFIAIY AYVHS OWOID INV LS vg LS LE 9g 6E LS 0s Lv cS LS vv LG LS LS 8g 8V z9 6V Or LS cv £G LS €S LS LITNI NVNOSVNVW 44 traditionally judged to "feel" the bottom. Therefore, an investigation into the effect of the borrow sites on the wave refraction pattern was made. Emphasis was given to calculations for the northern site as it contains more suitable and more plentiful quantities of usable fill material, and is the probable major source. 67. Previous studies. The phenomenon of wave refraction over dredged holes has received little attention. Only two technical papers could be found (Motyka and Willis 1975; Horikawa, Sasaki, and Sakuramoto 1977) that treat the subject. In addition to investigating wave refraction over dredged holes, both studies employed a simple version of the shoreline model described in Part IV to calculate shoreline change. These studies were of a preliminary nature but do provide some practical guidance. As expected, the refraction pattern was found to notably change with increase in dredged depth at a given water depth and with increase in wave period. Motyka and Willis (1975) state that, at the time of their study, beach mining in the United Kingdom was restricted to depths greater than 18 m (60 ft). 68. Calculation conditions. Two subgrids derived from the main depth grid were used in calculation of refraction over the borrow sites. The sub- grid for the northern site ran from Long Branch to Sandy Hook Lighthouse, and the subgrid for the southern area ran from Manasquan Inlet to Deal Lake. Table 4 (northern site) and Table 5 (southern site) list the calculated com- binations of deepwater wave conditions and dredged depths. 69. As described in Appendix E, the northern site cannot be dredged uniformly because the usable sediment source lies in irregularly-shaped lenses of varying thicknesses. For greater dredged depths, the borrow hole was "dug" on the grid at allowable depths in the appropriate regions. Three borrow hole configurations were simulated for the northern site and two for the southern site. Calculation results 70. The 20 refraction run conditions for the northern site are listed in Table 4, and the 15 conditions for the southern site are listed in Table 5. A wave direction of 7.5 deg corresponds to a wave out of the east, the quad- rant associated with much higher than average wave heights in the hindcast (Table 2). For this direction, waves of three periods (4, 8, and 12 sec) were run to investigate the effect of wave period. The effect of incident wave direction was investigated by specifying waves out of the east-northeast 42 Run No. mn £& WN — Table 4 Northern Beach Fill Borrow Site: Wave and Bathymetry Conditions for Refraction Runs Wave Angle (deg) Yan A AN nnn Ww 1S) ©) OO an WT WI UW He Se Ole is wl Wave Period (sec) 4 or ££ + foe) Cc © © OO Oo 43 Dredged Depths no (ft) dredging G6 10, 10 10, 20 dredging 6, 6 10, 10 10, 20 dredging 6, 6 10, 10 10, 20 dredging 6, 6 10, 10 10, 20 dredging 6, 6 10, 10 10, 20 Table 5 Southern Beach Fill Borrow Site: Wave and Bathymetry Conditions for Refraction Runs Wave Angle Wave Period Dredged Depths Run No. (deg) (sec) (ft) 1 5 4 no dredging 2 fod 4 6 3 U5 4 10 4 5 8 no dredging 5 05 8 6 6 (5 8 10 7 Te 12 no dredging 8 38) 12 6 9 fs 12 10 10 -22.5 8 no dredging 14 -22.5 8 6 12 -22.5 8 10 13 22.5 8 no dredging 14 225 8 6 15 22.5 8 10 (22.5 deg) and east-southeast (-22.5 deg) for the fixed, average wave period of 8 sec. 71. Full results are displayed in graphical form in Appendix F. Two examples are presented; one for the northern site in Figure 14a,b and one for the southern site in Figure 15a,b. These figures show alongshore distribu- tions of calculated wave height and direction at the nominal 3-m depth. The upper curve in each figure is the difference between the quantities calculated with and without the dredged hole of specified characteristics. The lower curve plots actual values of the quantity for the dredged and original bottom conditions. The horizontal axis is given in local model grid cell numbers. Refraction model cells are 150 m in length. For the northern borrow site, grid cell 1 is at Long Branch and cell 100 is located just north of the critical zone at Sandy Hook. The lateral limits of the northern borrow site are at cell numbers 75 and 93. For the southern borrow site, grid cell 1 is yy NORTHERN BORROW SITE WAVE HEIGHT COMPARISON 1.0 INCIDENT WAVE ANGLE: 7.5 deg INCIDENT WAVE PERIOD: 8.0 sec og 2 Fa eS 2 ORIGINAL BATHYMETRY: eo DREDGED BATHYMETRY: ~~ ~~ ~~~ z DREDGED DEPTH (ft): 6,10,20 5 2 < = -1.0 x = oO a =x So wi e5 Ww > < = -2.0 0.0 20.0 40.0 60.0 80.0 100.0 120.0 CELL NUMBER NORTHERN BORROW SITE WAVE ANGLE COMPARISON 10.0 INCIDENT WAVE ANGLE: 7.5 deg INCIDENT WAVE PERIOD: 8.0 sec 0.0 o lw a ORIGINAL BATHYMETRY: uw -! DREDGED BATHYMETRY: _______ = “10.0 < DREDGED DEPTH (ft): 6,10,20 ire) 2 oO 2 * < oO x = oO Wy -20.0 _l Oo 2 < [te} > < o.5 = c.0 52-0 0.0 20.0 40.0 60.0 60.0 100.0 CELL NUMBER SOUTHERN BORROW SITE WAVE ANGLE COMPARISON 10.0 INCIDENT WAVE ANGLE: 7.5 deg INCIDENT WAVE PERIOD: 8.0 sec eo ene HS 0.0 o Ww a ORIGINAL BATHYMETRY: w a DREDGED BATHYMETRY: _______ $ -10.0 < DREDGED DEPTH (ft): 10 Bs wi oOo = 390.0 2 rea < a =x — 35.0 ° ee} -20.0 — SO) 2 < 200 < = 30.0 0.0 20.0 40.0 60.0 60.0 jw /aMer CELL NUMBER Figure 15. Change in wave properties at the nominal 3-m depth for waves refracting over the southern borrow site: (a) change in wave height, (b) change in wave direction 46 at Manasquan Inlet and grid cell 100 is at Deal Lake. The lateral limits of southern borrow site holes are at cell numbers 25 and 52. 72. Classical geometrical wave theory predicts waves will diverge behind a hole. Wave height will be lower directly in the shadow of the hole and greater at the sides, where waves that are refracted outward by the hole and waves that are directly incident will combine. Since waves diverge out- ward from the center line of a hole, the directions of waves that have passed over the hole will be of opposite sign with respect to the center line. 73. The results shown in Figures 14 and 15, and in Appendix F exhibit the general characteristics of refraction over a hole as described in the previous paragraph. In Figure 14a, the change in wave height at the approx- imate center of the borrow hole has a maximum negative value as expected, and lobes appear on each side. The side toward Sandy Hook has two major lobes caused by the irregular plan shape of the hole. The changes in wave direction at the 3-m depth, given in Figure 14b, show waves positively directed on the south side of the hole and negatively directed on the north side. The lobe with a positive change in direction (on the south side) is higher because the waves are incident out of the east. 74. From Figure 15, it is seen that the two southern holes approxi- mately act as one unit. By comparison to Figure 14, the smaller areal extent and smaller dredged depth of the southern site limit changes in the wave properties near shore. Another mitigating factor is the greater ambient depth in which the southern borrow holes are located. 75. The following list gives a summary of results obtained from the wave refraction analysis: a. A (divergence) region of lower than average refracted waves exists at a location roughly along the centerline of the borrow site hole(s). A major lobe of higher than average waves exists at both sides of the divergence region. The refracted wave direction goes through zero in the divergence region and is opposite in sign on either side. Io The region of significant change in refracted wave height and direction is relatively insensitive to deepwater incident wave direction. This region remains localized directly landward of the hole and has lateral (longshore) dimensions of the order of the lateral dimensions of the hole. Changes in refracted wave height and direction are small for short period (4-sec) waves and increase with the wave period. The rate of increase of changes in wave properties decreases as the wave period increases, however. Thus, the change is larger 10 47 between 4-see and 8-sec waves than between 8-sec and 12-sec waves. 1. Changes in refracted wave properties increase with the dredged hole depth. The dependence of the changes on dredged depth is similar to the dependence of changes on wave period, i.e., the change is larger for relatively small dredged depths and there- after only gradually grows with increases in depth. 76. Discussion. Several hypothetical dredged bottom configurations were examined in this task. For the northern site, the plans with sediment dredging below 6 ft produced notable changes in refraction for waves of periods greater than about 8 sec. However, these plans also contained more beach fill volume than is necessary for the initial proposed beach fill. A smaller dredging area (for a given depth of sediment dredging) would have a less pronounced effect on the nearshore wave pattern. Also, dredging beyond a specified water depth at the deeper contours of the site or seaward of the site would mitigate the influence of the altered bottom topography on the waves. 77. The present study has indicated that random dredging in the shal- lower portions of the northern borrow site may have an impact on the shore- line. The natural infill sediment transport rate by waves and currents is expected to gradually fill and smooth the borrow pit, reducing influence of the hole with time. 48 PART IV: LONG-TERM SHORELINE CHANGE Introduction 78. A central task of this study was to develop a numerical model for simulating long-term shoreline change along Sea Bright to Ocean Township and to apply the model to evaluate alternative plans for protective beachfill configurations. For an open-ocean coast, where wave action typically is the dominant factor producing sediment (sand) movement, shoreline change occurring in the time frame of several years or decades is believed to be controlled by the transport of sediment alongshore. The basic budget analysis technique commonly used in coastal engineering and geology is an arithmetic balance of beach volume changes with inputs and outflows of sediment at the landward, seaward, and lateral boundaries of the region considered. Regardless of the quantity of sediment gained or lost at the onshore and offshore boundaries of the study area, it is the longshore transport of sediment that ultimately determines the long-term plan shape (horizontal pattern) of the beach. A numerical model of shoreline evolution is a highly systemized and quantified implementation of the budget analysis method, in which the change in beach volume is calculated as a function of the time-varying wave conditions. 79. As discussed in Part II, the budget study of Caldwell (1966) as well as subsequent studies have concluded that longshore sediment transport is the dominant process controlling long-term shoreline evolution of the north New Jersey coast. A numerical model of shoreline change for the project area is expected to be a valid extension of previous work, providing a useful tool for examining the course of beach change resulting from proposed shore protec- tion projects. 80. Structure of this chapter. An introduction to the shoreline change numerical model is first given. Representation of structures in the model is discussed in some detail since a long seawall and numerous groins figure prom- inently in shore protection along this coast. The position of the shoreline is to a significant extent constrained by these coastal structures. Calibra- tion and verification procedures are then presented. 49 Shoreline Change Numerical Model Background 81. The introduction of numerical models of shoreline change to applied coastal engineering research and practice began in the mid-1970's in England (Price, Tomlinson, and Willis 1973; Willis and Price 1975; Motyka and Willis 1975) and Japan (Sasaki 1975, Sasaki and Sakuramoto 1978) and, later, in the United States (e.g., Perlin and Dean 1979). Although CE has sponsored devel- opment of numerical models of shoreline evolution (Le Mehaute and Soldate 1980, Perlin and Dean 1983), this technology has had limited use in District and Division projects. 82. GENESIS. The numerical modeling effort for this project was jointly carried out with ongoing shoreline change model development activities performed by research work units at CERC. This allowed considerably more resources to be applied than were available for the project alone. As prog- ress was made, the Sea Bright to Ocean Township modeling task became a detailed case study using the recently completed shoreline evolution suite of models called "GENESIS," for GENEralized model for SImulating Shoreline change (Hanson 1987, Hanson and Kraus in preparation). It is anticipated that GENESIS will be released for Corps-wide use in 1989. 83. GENESIS is an integrated set of computer programs developed to cal- culate wave refraction and diffraction under simplified conditions, breaking wave height and direction, longshore sediment transport rate, and shoreline change. A wide range of boundary conditions, numbers and types of coastal structures, beachfill locations and volumes, and other common situations influencing shoreline change can be simulated with relatively moderate levels of operator effort. Input of wave conditions from an external source (as from data or another computer program such as RCPWAVE as in the present project) is also possible. One-line model 84. The shoreline change model portion of GENESIS is classified as a "one-line" model. In one-line model theory it is assumed that beach contours remain parallel over the course of the simulation period. Therefore, one line or contour, conveniently taken as the shoreline, can be used to characterize beach plan-shape change. GENESIS was developed from site-specific one-line models that have successfully described long-term shoreline change measured at 50 long groins, detached breakwaters, and seawalls (Kraus 1983; Kraus and Harikai 1983; Kraus, Hanson, and Harikai 1985; Hanson and Kraus 1986), both in the field and in laboratory physical models. 85. One-line theory assumptions and equations. Principle assumptions in one-line model theory are: Nearshore bottom contours move in parallel. A depth of closure exists beyond which longshore sediment transport does not take place. Ion |p 10 The volume of beach material in the littoral system, assumed to consist mainly of sand, is conserved. 86. Comparisons of available beach profiles separated by a recent 32-year interval and by shorter time intervals indicate that the slope of the profile along the project site is remarkably stable (cf. Part II, and Appen- dix D paragraph 14). In general, beach profile slope adjacent to a groin is expected to be milder than average on the updrift side and steeper than aver- age on the downdrift side. However, groins along the project shoreline are not long and sand bypassing can easily occur during episodes of high longshore sediment transport, minimizing the potential offset in slopes. Since the one- line model has been successfully used to simulate beach change at groins of much greater length at other sites, assumption a is considered to be well sat- isfied. A predictive expression for the depth of closure needed to satisfy assumption b has recently become available and is described below. Assumption ce is necessary for quantitative implementation of the budget analysis technique. 87. The basic equation of the one-line model is: wv, tS. (1) c where y = shoreline position t = time D, = depth of closure Q = volume rate of longshore sand transport x = distance alongshore 88. Depth of closure is difficult to determine and becomes, in effect, part of the calibration process. In calibration trials, the depth of 5 closure, D, , was either a constant (for example, 6 m or 8 m) or calculated from an expression given by Hallermeier (1979, 1983): Dy = 2.28 Ho > 10.9 (2) a lant in which H, is the significant wave height in deep water and L, is the deepwater wavelength. Equation 2 was developed for extreme wave conditions; H, and L, were originally defined to be the average of the respective quan- tities for the highest waves occurring during 12 hours in a year. Kraus and Harikai (1983) argued that for use with the one-line model, Equation 2 can be reinterpreted to hold as a function of daily input wave conditions. Numerical values for the depth of closure are given below in discussion of model cali- bration and verification. 89. The predictive formula for the longshore sediment transport rate is taken to be Hp b i os Q=— GaIe a) K, sin (20, .) SP 20K a cot(8) (6.5) (3) where Hj, = breaking wave height Cob = wave group velocity at breaking S = ratio of sand density to water density (S = 2.65) a = sand porosity (a = 0.4) 8p5 = breaking wave angle to the shoreline tang = average nearshore beach slope The quantities K, and K5 are transport parameters determined in calibra- tion of the model. 90. The first term in Equation 3 corresponds to the CERC formula (Shore Protection Manual (SPM) 1984, Chapter 4) and describes sediment transport produced by obliquely incident breaking waves. The second term describes transport produced by a longshore variation in breaking wave height. The first term is always dominant, but the second term will provide a significant correction if diffraction enters the problem (Ozasa and Brampton 1980, Kraus 1983, Kraus and Harikai 1983). 91. The SPM recommends a value of Ky = 0.77 for root mean square wave height in Equation 3, and the coefficient K5 has been empirically found to 52 lie in the range Ky = 0.5 K, to 1.5 K, (Kraus 1983). Numerical solution scheme 92. GENESIS allows selection of either an explicit or implicit finite difference solution scheme. Because of the locally jagged nature of the proj- ect reach with its numerous, irregularly-shaped revetments, seawall, groins, and other structures, the implicit solution scheme was chosen to be run at 6-hr intervals in order to minimize problems with instability. The program remained stable for all runs. 93. Grid and boundary conditions. An alongshore grid axis was set parallel to the longshore axis of the wave refraction grid. In shoreline modeling, the axis along the trend of the shore is customarily denoted as the "x-axis," and the axis orthogonal to it and pointing positive offshore is denoted as the "y-axis." This convention was maintained. Grid spacing along- shore was set at 50 m, giving three shoreline cells per wave model refraction cell. The finer spacing in the shoreline model is necessary to adequately resolve shoreline features in groin compartments. 94. The grid had to be extended well beyond the project area on both sides to obtain termination points that would provide adequate boundary conditions. The south boundary was placed at Shark River Inlet, and the north boundary was placed at the northern end of the Sandy Hook critical zone (cf., Part II). The south boundary condition did not allow sediment to be trans- ported north across the boundary except under high wave conditions, but did allow sediment to be transported south under those wave conditions indicating transport to the south. This boundary condition represents Shark River Inlet as an almost complete sediment sink. 95. The magnitude of the longshore sediment transport rate within the grid area is controlled in part by the lateral boundary conditions. In the present case, the predominant transport direction is believed to be from south to north, and model results become sensitive to the boundary condition imposed on the south side, i.e., at Shark River Inlet. This boundary condition cannot be specified with certainty, however. Reconnaissance of the site indicated that the inlet does function as a sink of transported sediment due to the channel and jetties. 96. From inspection of available shoreline data, it appeared that the northern end of the Sandy Hook critical zone had moved little over the past 50 years. Therefore, a fixed beach boundary condition was implemented on the 53 northern end of the grid. This type of boundary condition allows sediment to move freely alongshore across the boundary from either side. The full grid consists of 537 cells and spans a coastal reach of 26.9 km (16.6 miles). This is the longest known extent to be simulated with a shoreline change model. Representation of structures in the model 97. As discussed in Part II, a massive seawall and numerous groins have been constructed on the north New Jersey coast in an attempt to reduce erosion and control the position of the shoreline. To accurately simulate shoreline change, the influence of these structures on the longshore sediment transport rate and shoreline position must be represented in the model. 98. Seawall. A seawall functions to prevent landward retreat of the shoreline. Although only portions of the project coastline are backed by an actual seawall, roads and buildings located immediately landward of the pres- ent shoreline serve as an effective seawall (transport restrictor) since erosion would not be permitted beyond such facilities. Topographic sheets from the 1985 shoreline survey indicate the position of an effective seawall for the full length of the modeled area. A seawall introduces a constraint on the longshore sediment transport rate, in addition to constraining the possible position of the shoreline. Implementation of the seawall constraint in the model is complex; the report by Hanson and Kraus (1986) should be consulted for further details. 99. The seawall constraint is imposed at the same level of approxima- tion as the assumptions used to derive the one-line model. Wave reflection, scouring, and flanking are not simulated. This description is believed to be reasonable, provided the beach slope in front of the seawall does not appreci- ably deviate from that of the neighboring beach. This restriction is equiva- lent to assumption a of the one-line theory. 100. Groins. The positions and lengths of groins were obtained directly from 1985 aerial photographs and corresponding topographic maps. Based on careful inspection of the photographs, 91 groins were placed on the full grid with, for example, 19 groins lying in the 8-mile-long stretch from Sea Bright to Monmouth. Six groins lie to the north, between the project's north terminus and Sandy Hook Gateway Recreation Park. Only groins judged to be efficient at trapping sand were entered in the model; very short groins and nonfunctioning groin remnants were not included. Classification of a groin as 54 functioning or nonfunctioning was easily determined by visual inspection. 101. Bypassing at groins. If only longshore sediment transport is considered, in principle, a high groin extending seaward well beyond the surf zone Will completely block the movement of sediment. In practice, the surf zone often extends beyond the end of a groin, allowing sediment to move past the structure. Rip currents and complex horizontal circulation patterns also act to transport sediment around a groin. During high tide and high wave con- ditions, suspended sediment moving alongshore may overtop a groin, i.e., the groin will function as a weir. If a groin contains voids, sediment tran- sported alongshore can pass through it. Recent groin inspection (Coastal Planning and Engineering, circa 1985) indicates that most functioning groins in the study area can be considered sediment tight. In the present report, longshore transport of sediment around the end of a groin is called bypassing, and transport of sediment over and through a groin is called transmission. There are no data sets available to directly estimate groin bypassing and transmission. 102. Formally, GENESIS incorporates algorithms to represent sediment bypassing and transmission at groins. Transmission is represented by speci- fying a "permeability factor" ranging between O (no transmission) and 1 (com- plete transmission). Since the groins in the modeled area are mainly built of heavy, grouted stone, the permeability factor of functioning groins was set to 0. It is recognized that a limited amount of overtopping will occur during high tide and high wave conditions, but specification of that effect is not possible at the present tiie. 103. Bypassing of groins definitely occurs in the project area and must be represented in the model. In a theoretically complete analysis of the amount of sediment transported around a groin, both the cross-shore distribu- tion of the longshore sediment transport rate (Q in Equation 3) and the hori- zontal circulation and transport pattern must be known. Knowledge of the latter is beyond the present state of the art. For the former, there is not sufficient field data available to estimate the distribution of the transport rate. Although theoretical expressions exist to predict the cross-shore distribution of the longshore sediment transport rate, all pertain to ideal- ized conditions and none have been verified. In light of these circumstances, the simplest assumption that produces reasonable results is appropriate. 104. In GENESIS, a “bypassing factor" is defined which ranges between 55 0 and 1. The value 0O describes complete blockage (no bypassing) and the value 1 describes complete bypassing (no groin). The bypassing factor, B , was defined through use of a rectangular distribution of the longshore transport rate and the depth of closure pertaining to the wave conditions at a particular time step: in which Dy is the depth at the seaward end of the groin and Do is the depth of closure given by Equation 2. A rectangular distribution of the transport rate provides a good approximation to available field data sets (Kraus and Dean 1987) and is also easily calculated. Summary 105. In summary of the above, three kinds of information are required for shoreline simulation using GENESIS: a. initial conditions, such as initial position of the shoreline, positions and characteristics of structures, duration of time to be modeled, grid spacing, etc. Io wave conditions as a function of time to calculate the long- shore sediment transport rate. boundary conditions at the lateral ends of the beach. lo Model Calibration and Verification Introduction 106. The general calibration procedure for GENESIS is to determine the transport parameters K, and Ky by reproducing measured shoreline change that occurred at the target site between two surveys. If sufficient data are available, the calibrated model is then run to simulate observed shoreline change that occurred in a time interval not spanned by the calibration, to verify that the calibration constants are independent of the time interval. Since wave data for these time intervals, which may cover years, are rarely available, it is common to use hindcast wave data. 107. As discussed in Part-II, considerable shoreline position survey data are available for this coast. However, the standard calibration and 56 verification procedure could not be followed because of the enormous amount of unrecorded shore protection activity that has taken place and the recorded erratic movement of the shoreline position (cf., Appendix D). If the time history of all construction and beach fill activity were known, GENESIS could, in principle, incorporate it. As an acceptable alternative, calibration and verification were made for the critical zone of Sandy Hook, which is conti- guous with the north end of the project. This section of beach has exper- ienced minimal human intervention, except for occasional placement of beach- fill. Dates of major beachfills are believed to be known and could be avoided. Major beachfills were documented in the early 1960's, November 1982 through August 1984, and in 1984 (Slezak et al. 1984). Thus, these intervals were not included in the calibration and verification. Phillips (1985) tenta- tively concluded that shoreline change at the critical zone could not be explained by any known process information. However, as described below, shoreline change at the critical zone occurring in recent times was readily amenable to quantification by GENESIS. 108. In addition to the direct calibration and verification procedure, two indirect tests of model prediction were made. One was a qualitative check to ascertain that calculated and observed angles and sizes of groin fillets were in general agreement. Typical values of updrift fillet angle obtained from the model were 10-16 deg, whereas measured angles were in the range of 14-16 deg. This is a qualitative verification of the sediment bypass algo- rithm. The other test (described in paragraphs 121-124) was calculation of the net longshore sediment transport rate at specified locations along the coast. 109. In the calibration and verification procedure, visual comparisons were made by plotting calculated and measured shorelines. In addition, a dif- ference indicator Ydiff was devised to provide a more objective fitting cri- terion. The difference indicator is defined as the sum of the absolute dif- ferences between measured and calculated shoreline positions at each grid point: N Ydiff = )/ |Ymeas - Yeale| , (5) =H where the summation runs over all grid cells, and Ymeas = measured shoreline position Yeale = calculated shoreline position 57 grid cell number N = number of grid cells 110. If Ydiff is multiplied by the grid cell width (50 m), the devia- tion between measured and calculated shoreline positions can be expressed as a total net plan shape area. Calibration 111. Calibration of the GENESIS model was performed using the measured shorelines for 1971 and 1982, with 1971 data representing the initial shore- line position. After 11 simulated years, the calculated 1982 shoreline posi- tion was compared to the measured position. 112. By assigning different values to the transport parameters, K, and Ky in Equation 3, and making numerous calibration trial runs, a minimum difference indicator of Ydiff = 597 m was found between the measured and simulated 1982 shorelines. The resultant optimized values of the transport parameters K, and Ks, were 0.41 and 0.10, respectively. A plot of the mea- sured 1971 and 1982 shorelines, and the simulated 1982 shoreline, are shown in Figure 16. These optimal values were obtained for a variable depth of closure (Equation 2) given as a function of the wave conditions at each time step. 113. The average value of the depth of closure from Equation 2, excluding zeros associated with calm days of no longshore sediment transport, was found to be 1.35 m for the 3-year wave hindcast data set. However, in the long term, the beach profile is active to a much greater depth. Kraus and Harikai (1983) have discussed alternative definitions of the closure depth. In the present case, a representative value of the maximum depth of closure calculated from the wave data was found to be 6.0 m. A larger value of the depth of closure requires a larger value of Ky, (and, possibly, K5) in the calibration to maintain the same time scale of shoreline change. A larger value of K, produces a higher longshore sediment transport rate (Equation 3). 114. In order to check the sensitivity of the model to depth of clo- sure, the calibration procedure was twice redone, once with a constant depth of closure of 1.35 m, and again with a constant depth of closure of 6.0 m. These values represent the average and approximate maximum depths of closure, respectively, calculated from the input wave time series and Equation 2. In the case of a constant depth of closure of 1.35 m, the 1982 shoreline was sim- ulated with Ky = 0.28 and K = 0.15 . The difference indicator for the 58 SANDY HOOK CRITICAL ZONE CALIBRATION: K,=0.41 K2=0.10 Do=VARIABLE MEASURED 1971 MEASUREDEIS8i2i 22 =a == CAECUEATIE DUNS B20 icccccicelslsle 450 aa SEAWALL 350 300 Sees SS ewovweveuecucccc ces ueeutS SHORELINE POSITION Y (m) x o ee cae owe es . oO soo 1000 1s00 2000 LONGSHORE DISTANCE (m) Figure 16. Results of calibration 1971-1982 (variable depth of closure) measured and simulated 1982 shorelines was 591 m. In the case of a constant depth of closure of 6.0 m, the position of the 1982 shoreline was optimally reproduced with K, = 0.80 and K, = 0.10 , and the difference indicator was 575m. 115. A plot of the measured 1971 and 1982 shorelines, and the simulated 1982 shoreline calculated with a constant maximum closure depth, is shown in Calculated 1982 shorelines shown in Figures 16 and 17 both show an This result Figure 17. acceptable visual fit to the measured 1982 shoreline position. indicates that use of either a variable or constant maximum depth of closure will provide a calibrated model for estimating shoreline change in the project area, 116. Although the calibration procedure reproduced the 1982 measured shoreline successfully under both conditions, the longshore sediment transport rate predicted by the two calibrations differs by a factor of about 2. Since the optimal K, transport parameter changes from 0.41 to 0.80, the transport rate will increase by the same multiplier to replicate shoreline change over 59 SANDY HOOK CRITICAL ZONE CALIBRATION: K,=0-80 Ko=0.10 Dc=6.0 M MEASURED 1971 MEASURED I S98:2)05 = === = CALCULATED 1982 ...cceccceee SHORELINE POSITION Y (m) 100 t) soo 1000 1500 2000 2500 3000 LONGSHORE DISTANCE (m) Figure 17. Results of calibration 1971-1982 (maximum constant depth of closure) the same time interval. It should be noted that a value of K, = 0.80 is close to the empirical design value of 0.77 given for the CERC formula in the SPM (Chapter 4, p 4-96). Verification 117. Verification was performed using the measured shorelines for 1932 and 1953. Simulation of shoreline change for this 21-year time period was accomplished in order to provide independent confirmation of the calibration. Transport parameters K, and Ky were determined in the calibration with the three aforementioned procedures for specifying the depth of closure. 118. The three resultant simulated 1953 shoreline positions were then visually compared to the measured 1953 shoreline position and the acceptabil- ity of fit indicator Ydiff calculated for each case. Figures 18 and 19 show results obtained for a varying depth of closure and maximum depth of closure, respectively. A minimum Ydiff of 919 m was found between the measured and simulated shorelines for a variable depth of closure. The result using a constant maximum depth of closure was about 10 percent higher. 60 SHORELINE POSITION Y (m) SHORELINE POSITION Y (m) SANDY HOOK CRITICAL ZONE VERIFICATION: K,=0.41 Ky=0.10 D.=VARIABLE MEASURED 1932 MEASURED EI S/S:3 Men pee tana is CALCULATED 1953 Solelersielersiolelsiec 450 400 soa sen~me LONGSHORE DISTANCE (m) Figure 18. Results of verification 1932-1953 (variable depth of closure) SANDY HOOK CRITICAL ZONE VERIFICATION: K,=0.80 Ko=0.10 Oc=6.0 M MEASURED 1932 MEASURED) US!Si3i si ee CALCULATED 1953 eccccccccees , ! SEAWALL 4s0 400 3a 200 30 0 S00 1000 1s00 2000 2500 3000 LONGSHORE DISTANCE (m) Figure 19. Results of verification 1932-1953 (maximum constant depth of closure) 61 119. Table 6 summarizes results of the three calibrations and verifica- tions. It is concluded that use of the calibrated model with either a vari- able or constant maximum depth of closure will produce essentially equivalent results. Longshore sediment transport rates associated with the maximum depth of closure model will be higher by a factor of about 2. Table 6 Comparison of Calibration and Verification Results Ydiff for Ydiff for Calibration Verification D,(m) Ky Ko (m) (m) Variable 0.41 0.10 597 919 35 0.28 0.15 591 WSe/ 6.00 0.80 0.10 515 1,094 i _———_———————————————————————— 120. Longshore sediment transport rate capacity. The calibrated model was used to compute the longshore sediment transport rate capacity from Shark River Inlet to Sandy Hook. Transport rate capacity is the potential maxi- mum transport rate assuming infinite sediment supply and no interception or interruption by structures. The transport rate capacity was calculated with GENESIS by removing all groins and the seawall from the 1985 shoreline config- uration. The beach could then erode and accrete in response to local breaking wave action only. This situation was run to give an indication of the trans- port rates which created Sandy Hook, when the coast was in its natural state before implementation of shore protection measures. 121. Calculated longshore sediment transport rate capacities are given in Figure 20. Rates were obtained by averaging yearly volumes of transported sediment obtained from 3 years of simulated shoreline change with the 3-year hindcast wave time series. 122. Transport rate capacities with a constant maximum depth of closure are approximately double those for a variable depth of closure. The magni- tudes of these computed transport rate capacities are about one-third those obtained by Caldwell (1966) in his arithmetic budget analysis based on long- term change in shoreline position at Sandy Hook. 123. The computed sediment transport is directed to the north, as expected, but the difference between output and input rates at the ends of the 62 SHORELINE POSITION Y (KM) 0 so) 1.0 1.5 2.0 SS GIs) -N- * | 90 (174) DN ATLANTIC OCEAN HIGHLAND BEACH 74 (146) = SEA BRIGHT \ So - 80 (157) S \ = MONMOUTH BEACH z 94 (183) = 2) f a 15 86 (159) Ww oc [e) = f o Oo 60 (119) z Oo ae O d 54 (109) Wi (138) ASBURY PARK 25 UNITS: | o° m°/ year 47 (93) 30 Figure 20. Calculated transport rate capacities using a vari- able and a constant maximum (in parentheses) depth of closure 63 grid is small. The small difference means that the only slightly more sedi- ment leaves the grid than enters it, producing a relatively low rate of shore- line erosion. This result is considered to be incorrect in view of the his- torical trend of shoreline recession. For example, the rates obtained by Caldwell (see Figure 6) indicate a net transport rate out of the project area of approximately 174,000 cu yd/year (493,000 minus 319,000). 124. The suspected cause of the low net transport rate is the limited representation of the shadowing effect of Long Island in the wave refraction modeling. In the refraction study described in Part III, it was assumed that a single WIS hindcast would be sufficient to represent the wave climate over the entire study reach (approximately 16 miles). However, if wave shadowing is proven to be a significant factor, it would be necessary to perform the WIS hindeast at several locations along the reach and develop a methodology to incorporate these individual hindcasts in the nearshore refraction calcula- tion. This level of effort was beyond the scope of the present study. 125. Based on the successful calibration and verification of shoreline change at Sandy Hook, the wave input and shoreline model in their present forms are believed to have restricted applicability for evaluation of the relative merits of beachfill designs. A comparison of alternative designs was not performed in this study. 64 PART V: STORM-INDUCED BEACH EROSION Introduction 126. An important goal of this study is development of a quantitative means of investigating the potential impact of storm events on the integrity of an existing beach complex. Specifically, alternative beach fill berm con- figurations designed to protect the seawall along the open and approximately straight coastal area between Sea Bright and Ocean Township will be the sub- ject of this chapter. 127. This goal can be realized through the development of a one- dimensional numerical model based on certain simplifying assumptions which are particularly appropriate for the project area. These assumptions are de- scribed below. The analysis involves development and application of a numeri- cal model to estimate shoreline recession caused by storms of foreseeable intensity. 128. Although modeling procedures have recently been developed for numerically simulating the dynamic response of beaches to storm events, this methodology has not yet been extended for the simulation of storm-induced ero- sion of beaches protected by vertical seawalls. The purpose of this section is to present the modified model and methodology developed for accomplishing this task. The final product represents the first application of this new technology in a form which will, in the future, be available for District use. It is noted that scour at the foot of a seawall is not addressed in the model. 129. Following a brief description of the basic theory of the storm- induced beach erosion model, a demonstration of the verification of the model will be made. Although appropriate data for this purpose are sparse, compari- sons of pre- and post-storm surveys with model predictions demonstrate that the model is capable of producing realistic estimates of both erosion volume and berm recession. This comparison is followed by a description of the approach used to incorporate storm surge model information in the beach ero- sion model, which results in a means of comprehensively evaluating various beachfill configurations. Design beach profiles are evaluated by assessing each option with respect to a computed beach recession-recurrence interval curve. Included is an empirical estimate of the natural longshore variability 65 of beach erosion. In this manner, an efficient yet economically feasible beachfill program can be evaluated and realized. Beach Erosion Numerical Model Background 130. A numerical modeling scheme for beach and dune erosion developed at the University of Delaware and the University of Florida was selected for modification and enhancement to include the capability of simulating vertical seawalls. The background and application of this model have been reported by Kriebel (1982; 1984a,b), and Kriebel and Dean (1985a,b). A quantitative com- parison of this model, referred to as the "Kriebel" model, was made to various existing dune and berm erosion models by Birkemeier et al. (1987). 131. The comparison concluded that the Kriebel model was conceptually superior to other existing models, although, with proper calibration and "tuning," each was capable of predicting approximately similar results with respect to erosion volumes. This is not unexpected since all presently avail- able coastal zone erosion models are empirical in their formulation. It was found, however, that the Kriebel model was less susceptible to user manipula- tion and yielded realistic results without having to alter coefficients. This is due, in part, to the following desirable features of the model which are not shared by the other investigated models: a. The model does not require detailed knowledge of the berm or dune geometry or the offshore bathymetry. In The model relates the computed offshore sediment transport rate (which is used to compute dune or berm erosion) to several wave and storm parameters. Although the model is still empirical, this feature is viewed as advantageous. Models which neglect this coupling are rendered site specific in that they must be calibrated for each new application. 10 The model is easy and economical to apply. This feature permits the simulation of multiple storm events required for the computation of the final berm (or dune) recession and erosion volume versus recurrence interval plots. Basic Theory 132. The objective of the beach erosion model is to determine both the volume of material which will be eroded during a specific storm event and the corresponding amount of berm recession which can be expected to occur. These quantities are a function of the existing (pre-storm) berm and subaqueous 66 beach configuration, storm surge characteristics, and time. The prediction is accomplished by relating relevant characteristics of the storm with offshore sediment transport, which can then be related to erosion and recession. A complete treatment of the basic theory and methodology of the model is pre- sented by Kriebel and Dean (1985a,b) and in the documentation of the original model (Kriebel 1984a,b). 133. As previously noted, one of the primary advantages of the model is its ease of application. This is due, in part, to an assumption that the entire area to be modeled can be schematized into three distinct regions. These regions, shown in Figure 21, are as follows: a. A dune area defined by a constant height and constant face slope. The distance to the base of the dune face is specified as a distance from an arbitrary reference point. Io A berm area defined with an initial horizontal width (which can be specified as zero), height, and face slope. This area extends from the base of the dune face to a location equiva- lent to 0.5 ft (0.15 m) below mean sea level (MSL). An offshore area defined according to the equilibrium beach profile concept in which the depth h monotonically increases with distance offshore x according to the following relationship: 10 lay ef xe/3 (6) DUNE FACE SLOPE BERM WIDTH BERM FACE SLOPE DUNE HEIGHT =| HEIGHT OFFSHORE Figure 21. Schematized model area 67 where the coefficient A , called the "shape parameter," determines the steep- ness of the offshore profile. This form of the equilibrium profile was first advocated by Bruun (1954) and has subsequently been shown by others (e.g., Dean 1977, Hughes 1978, Moore 1982) to adequately represent the shape of natural offshore profiles. Since steepness is a function of grain size, the value of A can be related to the grain diameter, although the recommended practice is to select an A-value which produces an optimized representation of the existing (pre-storm) profile. If pre-storm conditions are not known, the grain size diameter relationship shown in Figure 22 (Moore 1982) can be used to estimate an A-value. Comparisons of calculated equilibrium profiles and measured profiles at various project sites are given in Appendix D. 134. One apparent disadvantage of this schematization concept is that certain features, such as offshore bars, troughs, navigational channels, or other abrupt irregularities, cannot be represented. Conversely, since the primary purpose of the model is to approximate subaerial beach erosion, an advantage of the model is that detailed bathymetric information is not re- quired for obtaining reliable estimates. An additional advantage is that the 0.1 1.0 10 100 SEDIMENT DIAMETER (MM) Figure 22. Shape parameter, A , versus sediment diameter (after Moore 1982) 68 geometrical idealizations of the modeled area are not difficult to justify since the majority of undisturbed beaches (i.e., beaches without jetties, detached breakwaters, or inlets in the immediate vicinity) have well-defined features which can be easily idealized. Beaches which are heavily influenced by natural or artificial boundary conditions cannot be modeled by the present or any other one-dimensional erosion model. Complex cases such as these must be modeled by computationally expensive two-and three-dimensional numerical models which more completely represent the physics of the flow regime. These models are not, with available techniques, economically feasible for studies such as this in which multiple simulations of random storm events are re- quired; additionally, the dune and fluid interaction has not yet been well represented in such models. Governing equations 135. The governing equations of the model can now be presented. The most important concept of the model, and one which separates it from most other beach erosion models, is the specification of a sediment transport rela- tionship which states that the cross-shore sediment transport rate in the surf zone, Q, , is a function of the dissipation of wave energy. This is written as Q, = k (D - Day) (7) q where k is an empirical coefficient which was determined by Moore (1982) to have a relatively constant value of 2.2 x 107° mtn (0.001144 ft/lb). The dissipation term, D , is given by dF dz (8) where F represents the energy flux evaluated with linear wave theory. The dissipation D can be reduced to the following simple form: 1/2 dh D = const h an (9) in which "const" is the product of known constant factors. The term Deg is defined as the equilibrium dissipation resulting when the offshore pro- file is in equilibrium according to Equation 6. It has been found that the 69 equilibrium dissipation is independent of both wave characteristics and water depth to the accuracy of available data (Moore 1982). 136. The relationship between cross-shore sediment transport rate and offshore bathymetric change is described by the one-dimensional continuity equation, written in the following form: dx aie (10) dt dh in which the distance x toa known contour line is written as a function of the change, or gradient, in the sediment transport rate. For example, if more sediment enters the system than exits, sediment accumulates on the bottom and the distance from a fixed point on the beach to the same contour line increases. Implementation and Assumptions 137. Equation 10 is numerically solved by an implicit scheme at each time step for the time-dependent surge level to yield a total volume of sediment eroded or deposited in the area bounded approximately by the water line and the breaker line. Recession of the dune and berm is related to the total volume deposited offshore through the assumption that this deposited volume originated from the dune and berm complex. If onshore transport is indicated, the offshore volume of erosion is assumed to accrete uniformly on the berm face. Accretion to the dune face is not allowed. 138. Only cross-shore sediment transport is considered in this one- dimensional approach. This points out an important assumption that alongshore sediment transport is negligible with respect to the cross-shore component during a storm. This is often not a valid assumption, but two-dimensional considerations were beyond the scope of model development. Longshore varia- tions will be introduced through empirical considerations. 139. The input data necessary for simulating the impact of a storm event are as follows: a. The dune area is specified according to a height, distance to edge of the dune face (from some arbitrary reference point landward of the feature), and slope of the dune face. In The berm zone is specified according to a width (if a horizon- tal transition between the dune face and berm face exists; otherwise, zero), height, and berm face slope. A shape parameter, A, (Equation 6) is defined to provide a best fit to the active portion of the offshore profile. 10 70 The breaking wave height must be specified. An estimate such as one inferable from WIS is adequate for this purpose. 1a e. The time history of the storm surge in 0.5-hr increments must be specified. 140. Dune and berm elevations in increments of 0.5 ft were specified in the original model. Material eroded from the dune and berm (or deposited on the berm) is assumed to erode equally from each elevation contour. The internal empirical parameters (such as k in Equation 7 and Q, in Equa- tion 10) have been shown to produce acceptable results without alteration. The above physical descriptors of the storm event and the areas to be modeled represent all of the data required for the numerical simulation of the inter- action processes. 141. To simulate the presence of the seawall in the project area, the Kriebel model was modified by replacing the dune area (shown in Fig- ure 21) with a vertical wall. Material deposited offshore is equally sup- plied by each berm contour line until recession reaches the seawall. At this point recession ceases and the beach in front of the seawall begins to rapidly erode. Since the model cannot describe the scour process at a wall, the cal- culation is terminated when the seawall becomes exposed at MSL. Further ero- sion might result in undermining of the structure. Figure 23 conceptually illustrates shoreline erosion and recession for cases with and without seawalls. 142. Following the completion of modifications to the Kriebel model, testing was performed to verify that the resultant model was capable of simu- lating the erosion patterns and volumes which were documented to have resulted from an actual storm event. This verification precess is described in the following section. Model Calibration and Verification 143. A quantitative assessment of the predictive capability of the Kriebel model was reported by Birkemeier et al. (1987). Although the range of parameters (storm surge level and duration, and dimensions of the dune and berm) was limited by the available data, comparisons of model output to pre- and post-storm survey data were made for 14 separate profiles representing four different East Coast storm events. Table 7 presents a summary of a EROSION WITHOUT SEAWALL PROTECTION EROSION WITH SEAWALL PROTECTION INITIAL PROFILE Figure 23. Conceptual seawall effects the measured and computed volumes of erosion above MSL. 144, The report by Birkemeier et al. (1987) concluded that the Kriebel model is capable of predicting order of magnitude estimates of erosion rates for storm events of known intensity and duration. Limited sensitivity testing was also performed in which the change in computed volumes of erosion result- ing from the alteration of certain key model parameters was investigated. Results showed the model to be stable with respect to these parameters if they were specified within reasonable limits. None of the 14 profiles referred to in Table 7 contained a seawall. In order to verify the seawall simulation modification of the model in the present study, additional sets of data were sought which were appropriate for that purpose. 145. The scarcity of pre- and post-storm survey data for locations seaward of a seawall-backed beach is primarily due to the fact that post-storm surveys rarely concentrate on areas where no structural damage took place. Unless a seawall is undermined or experiences failure, quantitative measure- ments documenting the erosion of the beach directly in front of the seawall are often not made. However, one data set was obtained that was acceptable Te Table 7 Measured and Computed Volume Changes (cu _m/m) Storm Profile Measured Computed Percent of Measured Westhampton, NY 3 52.8 29.5 56 3 February 1972 4 69.4 23.0 33 5 51.3 37.9 63 Westhampton, NY 3 26.5 33.9 128 19 February 1972 4 35.2 19.1 54 5 Silica 21.1 67 Long Beach Is., NJ 14 30.6 38.2 125 19 December 1977 15 26.2 42.5 162 16% 13.8 35.2 255 Ute 34.5 43.4 126 18* 6.0 41.7 695 Duck, NC 186 31.7 33.4 104 13-14 November, 188 41.3 48.1 112 1981 190 37.4 SBi0(/ 143 * Profiles were located in the vicinity of coastal structures for the purposes of verifying the model (CE 1986). A brief description of the subject area and the associated storm event are presented below. 146. Simulation for Revere Beach. The storm event selected for veri- fication of the beach erosion model with seawall modification was a winter storm of 6-7 February 1978. The area of interest is Revere Beach located in Revere, Massachusetts, approximately 10 miles north of Boston (Figure 24). Onshore profiles and bathymetric surveys were taken prior to the storm during November and December of 1977 and again after the storm. Eight separate sur- vey lines were taken which covered the area from the seawall to a depth of 20-30 ft MLW. A storm summary, prepared by Carol R. Johnson & Associates, Inc. (1978), reported a storm surge in Boston Harbor of approximately 4.0 ft superimposed on a spring tide condition. The two-day event was accompanied by maximum wind velocities of 70 knots. Recorded time histories of the total surge level indicated a maximum elevation of nearly 10.5 ft above the mean. Such documentation of a single storm event and resultant bathymetric change is uncommon. 147. Analysis of the Revere Beach data set was made in order to deter- mine its acceptability for verification of the seawall model. This analysis resulted in the following observations: 73 ATLANTIC OCEAN SCALE IN MILES REVERE BEACH REVERE, MASSACHUSETTS Figure 24. Location map for Revere Beach (from CE 1986) 74 1p Io 10 Of the eight profiles surveyed before and after the storm, two were at locations where erosion of the beach in front of the seawall was so severe that protective material was placed in front of the wall during February and March of 1978, before the post-storm profiles were taken (Profiles 2 and 3). Two profiles contained nonvertical barriers (a sloping concrete apron at Profile 4 and a sloping stepped wall at Profile 7) which were inappropriate for the model without further modifi- cation. An additional profile (Profile 8) was in a location so protected by land that accretion of the beach was experienced instead of erosion. Obviously, this profile was also inappro- priate for a one-dimensional analysis. The remaining three profiles (Profiles 1, 5, and 6) were selected for model simulation. The assumption of one-dimensionality necessary for making this type of analysis economically feasible has been previously mentioned. The present project area is the Atlantic Ocean coast of New Jersey from Sea Bright to Ocean Township. A one- dimensional modeling effort is particularly well suited to this area since it is located on an open coast and has no major structural or natural features which dominate the cross- shore flow regime, invalidating the one-dimensional assump- tion. Revere Beach, however, is a protected beach which appears to be a highly two-dimensional system. As can be seen in Figure 24, the beach is located within a small bay (Broad Sound), protected to the northeast by a headland. The Febru- ary storm originated from the northeast quadrant and winds were almost shore parallel during the entire storm event. This is clearly not one-dimensional. This was documented in the storm summary which reported a considerable amount of alongshore transport below NGVD. Use of the Revere Beach data was therefore limited to the area above NGVD where erosion of the dune in front of the seawall was observed. The assumption is therefore made that the offshore two-dimensional effects can be considered to have a negligible effect on erosion of the dune and berm. This implies that, although the predomi- nant transport of sediment seaward of the NGVD shoreline is alongshore, the cross-shore component which governs the one- dimensional model can still be used to compute dune erosion. The conclusion is, therefore, that the Revere Beach data set is acceptable for a limited verification of the model (i.e., examination of dune erosion and not of offshore erosion and deposition patterns). This approach is valid if limitations of the basic model are considered during interpretation of the results. Surge data for the 6-7 February storm event were provided from a tide gage in Boston. The maximum surge level was recorded to be 10.3 ft NGVD. Due to differences in location and degree of sheltering, the surge level at Revere Beach was slightly higher since it was reported that the seawall (elevation of approximately 15 ft NGVD) was overtopped. This difference in surge levels was accounted for in the verification of the model. 15 1a Each of the post-storm profiles was lacking a survey point at the seawall. It is not known whether or not the berm eroded back to the seawall (although photographs seem to indi- cate that erosion did extend to the seawall). Since the total volume of eroded material is somewhat dependent on that mea- surement, a comparison of volumes will not be made. Verifica- tion is made by comparing the computed change in elevation at specific distances from the seawall with the measured changes at those locations as shown by the post-storm survey and reported in the storm report. 148. The beach erosion numerical model with seawall modification was used to simulate the storm event for the three profiles described above. Input data for the berm geometry were taken from the surveys, whereas storm data were extracted from the storm report. Results of those simulations are shown in Table 8 and discussed in paragraph 151. 149. Simulation for Sea Bright. To demonstrate that berm recession in the project area could be adequately predicted by the model, a simulation of beach erosion was performed for the November 1953 storm that made landfall in Table 8 Computed and Measured Elevation Changes at Revere Beach Number ft change, ft change, ft 1 75.1 -1.71 -1.05 187.0 +0.20 -1.05 236.9 -1.51 -0.46 average -1.02 -0.85 5 62.3 -1.12 -1.97 87.6 -1.38 -1.97 WAZ -2.00 -1.38 200.1 -2.00 -0.23 average -1.64 -1.38 6 (Ber -0.69 -0.46 134.8 -2.00 -0.46 22511 -0.49 +0.56 average -1.05 -0.13 76 the vicinity of Sea Bright. Detailed comparisons of the impacted area were not made since the precise amount of recession is not known. It was reported (CE 1954, p 48) that "At elevation 10 ft above mean low water recession aver- aged 98 ft." It was also reported that ". . . selected profiles subsequent to the November 1953 storm indicate recession of from 30 to 85 ft from the summer shore line position" (CE 1954, p D3). It can be inferred that the reported recession was not solely a result of the November storm, but of all storms occurring between the surveys of June - August 1953 and the November post- storm survey. Another major storm event, occurring on 22 October 1953 with an "observed storm tide" of 6 ft MSL (Dendrou, Moore, and Taylor 1981), probably contributed to the erosion documented by the November post-storm survey. The 30-, 85-, and 98-ft recession figures most likely represent an overestimate of the November 1953 storm damage; nevertheless, it is well documented that erosion from this single event was devastating. 150. The storm event was simulated by subjecting the summer 1953 Profile 13 (CE 1954) to the November 1953 storm surge as documented by Pore and Barrientos (1976). The numerical simulation gave a maximum recession of 48.3 ft, within the recession limits reported by CE (1954). Caldwell (1959) presents additional data on the November 1953 storm, which are given in Table 9 and used in the present study to assess longshore variability of beach erosion. Table 9 Measured Changes in Contours Between Surveys Made in the Summer of 1953 and Immediately after the 6-7 November 1953 Storm, Sea Bright, New Jersey (after Caldwell 1959} Contour Elevation Landward Retreat of Contour (ft) (ft above MLW) Average Maximum 0 65 110 5 63 90 10 98 180 15 53 120 151. Results of the two verifications for seawall-backed beaches indi- cate that the model acceptably reproduces measured recession of a berm in front of a seawall which occurs in response to a known storm event. Although Ut the physical situation of Revere Beach violates the one-dimensional assump- tions used in developing the model, reasonable results were obtained. Further verification tests should be performed in the future when additional data become available. For the present purpose, the model has been shown to be adequate for estimating berm recession for the Sea Bright to Ocean Township area. The following section will focus on application of the model to deter- mine the amount of recession in front of the project seawall as a result of a stochastic application of both hurricanes and northeasters of known frequency of occurrence. Storm Simulation 152. The need for an effective beach renourishment program was recog- nized following the occurrence of the November 1953 storm which caused wide- spread damage to the project beach area. Three alternate berm width plans (CE 1984), based on post-storm recommendations, were selected for evaluation: a. A historical maximum width of 100 ft corresponding to the average dune recession measured during the November 1953 storm. The design profile for the 100-ft width is shown in Figure 25. b. A minimum width of 30 ft recommended by the Beach Erosion Board (BEB) in 1959. ec. An intermediate design of 50 ft (also recommended by the BEB) as a realistic compromise of the two extremes. 153. The beach erosion model was selected as a means of systematically evaluating the effectiveness of a coastal system to withstand a variety of storm events. The methodology used to evaluate the proposed berm width alter- natives makes extensive use of the stage-frequency analysis described in Part VI of this report. In that chapter, details on the generation of hurri- canes and northeasters are presented. The generated storms are used to develop stage-frequency curves corresponding to both hurricanes and north- easters. Both types of storm events were analyzed because of basic differ- ences in the characteristics of the two; hurricanes tend to have very high surge levels which are relatively short in duration, whereas northeasters are lower in surge level but substantially longer in duration. Each generated hurricane and northeaster is composed of a randomly selected storm surge of known duration superimposed on a randomly selected tidal series. The 78 (1961 NYNAQ Wous payeotuMumod) YAPTM Wdeq 3J-OO| 243 4OJ eTTJoud uBTseq “Gz eunsTA (14) JONWILISIG 002! oool 008 009 OOv 002 82! ‘ON 3NIT ; 2 ee ONILSIXS Gz Ol- % ° SS MIW aor Ve S3d07S NOIS3G G3ON3WNOO3Y iE a Ee Ol i MW ‘14 Ol+ LHOISH WY38 G3GNaNWOo38y ————=— : ol A i Hi S3INVA HIGIM WY3a i i oO iS] TIVMV3SS ONILSIX3 (MIW L4) NOILWAI14 19 resulting storm event is identified by a total surge level (storm surge plus tide), a recurrence period (in years) associated with the basic storm event, and a numerical identifier uniquely associated with each generated storm. This identifier is used to reconstruct the time history of each generated storm for subsequent input to the beach erosion model. 154. The reconstruction of each randomly selected storm event from the storm identifier results in the development of an event of known return period and surge height, but of variable duration. Erosion of dune and berm areas has been shown (Birkemeier et al. 1987) to be highly dependent on storm dura- tion as well as surge level; two storms of equivalent return period and surge level do not necessarily produce the same amount of erosion and recession. For this reason, multiple storm simulations were made for both hurricanes and northeasters. 155. The procedure for generating hurricane and northeaster storm events and the methodology for using these events to calculate berm recession versus recurrence interval relationships were specifically developed for this project. Fifty-five hurricanes were randomly selected, reconstructed, and generated corresponding to discrete maximum total surge (storm plus tide) elevations ranging from 4.0 ft to 14.8 ft NGVD at 0.2-ft increments. Simi- larly, 24 northeasters were generated for total surges ranging from 5.0 ft to 9.6 ft NGVD at 0.2-ft increments. All storm events were based on stage- frequency curves (for both hurricanes and northeasters) computed for Monmouth Beach, New Jersey (see Part VI). 156. Due to differences in both erosion and recession produced by storms of equal surge level but different duration (and distribution of the surge peak within each storm event), the random storm selection was performed five times in order to create a large comprehensive data base of storm events of known total surge level, duration, and return period. This approach re- sulted in the generation of 275 hurricanes and 120 northeasters which are individually input to the beach erosion model for evaluation of the desired berm configurations. The approach provides a reliable data base for comparing potential storm-associated damage as a function of erosion resulting from storm events with frequencies of occurrence ranging from several years to over 1000 years. 80 Evaluation of Existing Conditions 157. An evaluation of existing conditions was requested by CENAN in order to document the necessity for developing a beach fill protection plan for the project beach complex. An investigation was therefore made of four existing profile locations specified by CENAN to be representative of the Sea Bright to Ocean Township area. Locations of the profiles are shown in Fig- ure 26. Detailed information related to profile locations and shoreline change is contained in Appendix D. 158. Case A. Profile 160, shown in Figure 27, was selected to repre- sent a typical cross-section characterized by an existing beach backed by a protective seawall. Profile 160 is located in Block 15, the Long Branch south block (Figure D1), along which the average rate of shoreline change from 1953- 1985 was +0.39 m/year (Table D18). The average measured grain size for this block is in the range of 0.33-0.37 mm (Table D19). Analysis of the offshore profile indicated that the value of the shape parameter A of Equation 6 was equal to 0.176 m'/3, corresponding to a sediment diameter of 0.40 mm. 159. Case B: Profile 82, shown in Figure 28, was selected to represent the case of a narrow beach in front of the seawall. This profile is located in Block 5 in the Navesink Beach-Normandie Beach area. The average rate of shoreline position change for this area was -0.31 m/yr between 1953 and 1985 (Table D18). Grain size measurements are not available for Navesink Beach, but Normandie Beach average grain size varies between 0.35 and 0.60 mm for the 1953 and 1985 surveys respectively (Table D19). Data from Normandie Beach also show a trend of increasing grain size with distance offshore (Table D19). For the 1985 survey, the mean diameter reaches 1.05 mm at the -30-ft contour. Analysis of the offshore profile data yielded an A-value of 0.283 m!/3 | corresponding to a sediment diameter of approximately 1.40 mm. This large diameter is characteristic of a surf zone subjected to steep waves which remove much of the small-diameter sediment. In addition, the possible absence of an adequate supply of fine sediment to replace that lost offshore as a result of wave action is indicated. Note that a small subaerial beach area is specified as an initial condition; a finite beach initial boundary condition is required by the model to give an initial source of material for offshore sediment deposition. 160. Case C. Profile 140, shown in Figure 29, was selected to 81 NEW YORK SANDY HOOK gr J] SEA BRIGHT NEW JERSEY CSS SHARK RIVER INLET “YOY MANASOUAN INLET ATLANTIC OCEAN » BARNEGAT INLET Figure 26. Location map of selected profiles in the Sea Bright to Ocean Township project area 82 ELEVATION (FT NGVD) ELEVATION (FT NGVD) 30 PROFILE LINE 160 ———— el JUNNSS tf cas) is ®S t-=) i} ran t.s) -20 -500 ) 500 1000 1500 2000 2500 3000 3500 DISTANCE (FT) Figure 27. Profile 160 30 PROFILE LINE 82 ——' 15 JUN 85 ia*) is) rare c.s) -500 Q 500 1000 1500 2000 2500 3000 3500 DISTANCE (FT) Figure 28. Profile 82 83 PROFILE LINE 140 —— 20 JUN 85 ELEVATION (FT NGVD) -500 Y) 500 1020 1500 2000 2500 3000 3500 DISTANCE (FT) Figure 29. Profile 140 represent a beach with a high dune but with no protective seawall. This case was utilized to demonstrate the possibility of unrestricted dune recession. Profile 140 is located in Block 12 (North Long Beach). The average rate of shoreline position change in this area (-1.18 m/yr) exceeds that of all other blocks for the 1953-1985 time period (Table D18), suggesting that beach mate- rial has been available for transport. Average grain size is 0.34 mm for both the 1953 and 1985 surveys (Table D19). A coefficient of A = 0.136 m'/3 was computed from offshore profile data. This value corresponds to a sediment diameter of 0.28 mm. 161. Case D. Profile 186, shown in Figure 30, was selected as an ex- ample of a low-elevation dune unprotected by a seawall. This profile is located in the Deal south block, along which the average rate of shoreline position change (1953-1985) was -O.44 m/yr (Table D18). Average grain sizes of Block 19 for the 1953 and 1985 surveys were in the range of 0.28-0.31 mm (Table D19), indicating the availability of sediment for transport. An A-value of 0.135 m'/3 was computed for this profile, corresponding to a sedi- ment diameter of 0.29 mm. 84 PROFILE LINE 186 —— 30 JUN 85 ELEVATION (FT NGVD) -500 @ 508 1000 1500 2000 2500 3008 3500 DISTANCE (FT) Figure 30. Profile 186 162. All water levels and profile elevations were adjusted to MSL for input to the beach erosion model. The relationships between datums used for this conversion are MSL = (NGVD + 0.57 ft) = (MLW + 2.2 ft) as determined by telephone conversation with the National Ocean Service (NOS) of NOAA in January 1987. 163. Evaluation of the existing-condition cases was accomplished by subjecting the four individual profiles to each of the 275 hurricanes and 120 northeasters in a total of five separate runs. A value was then computed which corresponds to the maximum computed recession of any contour line be- tween the the dune crest and MSL during the entire storm event of known recur- rence interval. Maximum recession does not usually occur at the end of a storm, during which time beach recovery may be in progress. Therefore, the maximum was selected as a reliable indicator of storm-related damage. 164. Recession-recurrence diagrams. The recession-recurrence inter- val plots represent the computed recession resulting from each of the stochas- tically generated storm events. These plots demonstrate the natural variabil- ity in recession which occurs as a result of storms of equal surge. For the 85 present numerical investigation, these differing recession values are attri- buted to the fact that storms of equal surge level result in varying volumes of erosion due to differences in total surge duration. In order to develop a single design curve descriptive of the overall relationship, an upper envelope line is provided for each plot. In each case, a small number of points lie above this line. These points were considered to be atypical in that they were found to result from generated storms of unusually long duration. 165. The resulting straight line recession-recurrence interval rela- tionships for both hurricanes and northeasters were combined to produce a final single curve for each representative profile. This design curve was generated by adding the frequency of occurrence (reciprocal of the recurrence period) corresponding to a given envelope curve recession value for both the hurricanes and northeasters and taking the reciprocal of the sum to produce a return period for the combination. Both the individual storm simulations and the combined storm design curves are presented in the following analysis of existing condition profiles for the Sea Bright area. This procedure is not considered rigorous, but is believed to give a reasonable first-order estimate. 166. Case A-Profile 160. The recession-recurrence interval curves corresponding to hurricanes and northeasters for Profile 160 are shown in Figures 31 and 32. The computed spread in recession values can readily be seen in each plot. As an example, one 400-year storm shown in Figure 31 indi- eates a recession of 67 ft. Analysis of the generated total surge shows that this event has a duration of 33.0 hr whereas the average generated surge dura- tion is on the order of 18-20 hr. For this reason, this point was allowed to fall outside the design envelope. The combined hurricane-northeaster design curve is shown in Figure 33. 167. Case B-Profile 82. The recession-recurrence interval calculations for Profile 82 indicate that all storm events (both hurricanes and north- easters) result in erosion of the entire beach to the seawall. Although re- covery of the beach often follows a storm event, the present calculations are made for maximum recessions only. An implication of this analysis is that even a 10 to 20-year event could severely jeopardize the integrity of the seawall. The use of a smaller shape parameter, indicating a smaller grain size than first used, also resulted in complete erosion of the beach fronting the seawall. Since prediction of scour at the toe of a seawall is beyond the 86 ow 00S 0 on 0°02 (L4) NOISS39]3¥4 WNWIXVW 2) a = a ir) — =z lu oO =z ry a a =) Oo ve a HURRICANES 160 PROFILE NO. CASE A: ile 160 recurrence plot for Prof leane recession Hurri Figure 31 ow 0°05 Oo oo 00 (L4) NOISS3034 WOWIXVW 7) (4 < w > a a ww = z= iv) (5) z w a (o4 > oO w x NORTHEASTERS 60 1 PROFILE NO. CASE A: recurrence plot for Profile 160 Northeaster recession Figure 32. 87 72.0 @0.0 DOGDAURROOORROOROORRRORODORRGDONGOUNRROODOOOOOED FVOTONOUAUUOOQGQNUOURUEOORERUOOUACUAUAALY a we z ° 77) ” ws oO w ae = =) = < < = 24.0 12.0 TOT TTT TUTTI TTT CTT RECURRENCE INTERVAL (YEARS) CASE A: PROFILE NO. 160 HURRICANES AND NORTHEASTERS Figure 33. Maximum recession-recurrence design curve for combined hurricanes and northeasters for Profile 160 capability of this or any known model, no further analysis of Profile 82 was made. 168. Case C-Profile 140. Calculations for Profile 140 illustrate the protection afforded by a relatively high dune. This example profile has a berm crest elevation of approximately 20 ft. Recession-recurrence curves for hurricanes (Figure 34) and northeasters (Figure 35) indicate maximum reces- sions of less than 30 ft for a 500-year storm. The model indicates that re- cession rarely exceeds 30 ft regardless of the intensity of the storm event. These results are consistent with the observed tendency for a dune to afford protection to the beach by eroding and thereby making material available for transport offshore. This results in a decrease in surf zone depth which forces incident waves to break further offshore, dissipating their energy over a wider beach and causing less erosion of the dune. The combined hurricane- northeaster curve for Profile 140 is shown in Figure 36. 169. Case D-Profile 186. Profile 186 was selected as an example of an unprotected beach with a low dune, in contrast to the Profile 140 high-dune case. Maximum recession curves for hurricanes and northeasters are shown in 88 os OO a4 ow orst 0°01 (14) NOISS393¥Y WNWIXVW RECURRENCE INTERVAL (YEARS) HURRICANES 40 1 PROFILE NO. CASE C: recurrence plot for Profile 140 Hurricane recession Figure 34. ow ost oF (L4) NOISSZ0394 WOWIXVW RECURRENCE INTERVAL (YEARS) NORTHEASTERS 40 1 PROFILE NO. CASE C: ile 140 -recurrence plot for Prof Northeaster recession Figure 35. 89 0.0 0.0 U.0 (= we z °o ” ” uw [s) ws (3 = 2 = bad = < < = RECURRENCE INTERVAL (YEARS) DESIGN A: 100-FT BERM WIDTH HURRICANES Figure 40. Hurricane recession-recurrence plot for Design A: 100-ft berm width 93 Oo 0°02 ool a4 inn NOISDSEG Satie n a a wu S z Ww (>) =z ivy) a a =) Oo ws 4 NORTHEASTERS 100-FT BERM WIDTH DESIGN A: ion-recurrence plot for Northeaster recess igure 41. F ign A: 100-ft berm width Des HURRICANES AND NORTHEASTERS RECURRENCE INTERVAL (YEARS) = Ll 2 = = a HUTT TTT TTS, @ AUT TTT : ae ° oO < =z ii | n” om! ol 06 Ce: a 2 o°s! 5 (14) WoIGEREE watwUReen [ase ee en ee a ee) recurrence design curve for combined Maximum recession Figure 42 ft berm width 100- hurricanes and northeasters for Design A g4 173. Design B: 50-ft berm width. The recession-recurrence interval plots for Design B are shown in Figures 43 and 44, and the combined curve is shown in Figure 45. The combined storm analysis indicates that maximum reces- sions of 50 ft can occur with recurrence intervals of about 30 years. This amount of erosion would completely eliminate the flat protective berm region fronting the seawall, continued erosion would uncover the face of the seawall, rendering it increasingly vulnerable to scour. 174. Design C: 30-ft berm width. The recession-recurrence interval relationships and the combined design curve for Design C are shown in Fig- ures 46-48. Results of the combined design curve indicate that maximum recessions to the seawall can occur in as short a time as two years. This severe recession indicates the vulnerability of the 30-ft design to storms of relatively short recurrence intervals. 48.0 60.0 TUMANAANNNLUUL SNUNTONOUUSTUTT PUGCORE BP IBORE B.0 iS ira z ° 7) 7) uu oO iv) a = 2 = ~ < = 24.0 12.0 RECURRENCE INTERVAL (YEARS) DESIGN B: 50-FT BERM WIDTH HURRICANES Figure 43. Hurricane recession-recurrence plot for Design B: 50-ft berm width 95 te fe} Gy n a) a a (e} = a os no Ww a, < rs a n o x= ai .3) x S ° bE [0] io z = a me 50 Re ” O «4 an < oz Sm w u = z aa 1 & > << = (op te = 2 = ow =; = S pal ice) = 5 oi ae fe = = © = = ore z “ iA TITER CTT " z me b , TUTTI TTT ae ee n Se 2 2] be wt ae Sa Page bn a ao oA = a ie 3 ° wo e wo a f= a a 3 a ne i SC a = a0 o% OW 0% Om 0% oe ore = (14) NOISS393¥8 WOWIXVW a (L4) NOISS3934 RTINCaIR ft berm width ign curve for combined 50 B ign recurrence des 96 Maximum recession- hurricanes and northeasters for Des Figure 45 0S om oo 002 (L4) NOISS303¥4 WNWIXVW RECURRENCE INTERVAL (YEARS) HURRICANES 30-FT BERM WIDTH DESIGN C: recurrence plot for e recession- ican ign C Hurri Des Figure 46 30-ft berm width OG 0 ou 0°0e (14) NOISS39034 WOWIXVW RECURRENCE INTERVAL (YEARS) NORTHEASTERS 30-FT BERM WIDTH DESIGN C: recurrence plot for ft berm width Northeaster recession igure 47 F ign C: 30- Des 97 72.0 @0.0 B.0 NUTT 24.0 12.0 | = = = a =a = [—-———] == _ = == — =m uw = ~— =a (a) z =< fa) on => — = n [I no i] Ww [=] oI Oo i= enh i) a i ——} = ———<—} = am i————F =I _ = ~< —— 7) < oa aa = <= = a <= <= el = =a ==] ea | el TUT RECURRENCE INTERVAL (YEARS) DESIGN C: 30-FT BERM WIDTH HURRICANES AND NORTHEASTERS Figure 48. Maximum recession-recurrence design curve for combined hurricanes and northeasters for Design C: 30-ft berm width Variability Factor 175. The recession-recurrence analysis procedure developed for this project treated storm descriptors in a stochastic (random) manner. The beach was idealized as having a known cross-section, and longshore variability of the beach profile was neglected. In nature, the hydraulic eroding forces and the beach profile will individually exhibit longshore variability. These separate variabilities are site-dependent and also change in time. Therefore, it is impossible to quantify, in a deterministic manner, the mesoscale detail of the longshore variation in erosion potential due to storm action. 176. Although an estimate of the longshore variability in storm erosion potential cannot be calculated at present, it is possible to arrive at an estimate based on measured variations. Birkemeier et al. (1987) and Savage and Birkemeier (1987) examined the variation in erosion associated with 588 different profiles at seven East Coast localities for 13 storms. A variabil- ity factor was defined to encompass 75 percent of the observed change. This value can be interpreted as the multiplier of the median required to include 98 volumetric erosion expected to occur on 75 percent of the shoreline. The so- defined variability factor was determined to have the value 2.0. Assuming that the results computed in the present study are median values, multipli- cation by the factor 2.0 would encompass 75 percent of the expected recession of the shoreline. It should be noted that the data set compiled by Savage and Birkemeier does not include events greater than the 100-year storm. 177. There are limited data available for the project site with which to test the variability concept. Caldwell (1959) provides average and maxi- mum values of landward retreat of selected contours above MLW for the New Jersey Storm of 6-7 November 1953 (Table 9). A total of 20 profiles spaced over approximately 40 miles of the north New Jersey coast were averaged in the comparison, omitting measurements which were believed to be influenced by the presence of the seawall. This information is a good source with which to examine longshore variability in recession at the site (subject to consi- derations of storm sequence given in the next paragraph). Data presented in Table 9 are compatible with a variability factor of 2.0 as discussed above. 178. There is an ambiguity associated with the data given in Table 9. As discussed above, at least one other storm is known to have impacted the area in the interval between profile surveys. Caldwell (1959) takes note of this and states "... it is believed that the greater part of the indicated erosion took place during the (6-7 November 1953) storm itself." 179. The methodology and recommendations given here were developed to estimate beach erosion resulting from a single event. A natural shore exper- iences several annual erosive events of various strengths, continual shoreline evolution caused by longshore sediment transport, and accretion under summer swell conditions. Given the long-term trend of the project coast to erode, consideration of multiple erosive events and natural longshore variability indicate that a variability factor on the order of 2.0 should be incorporated in the protective berm design. 180. Incorporation of the variability factor is made by determining the frequency of occurrence of maximum recession equivalent to one-half the design berm width. This approach gives a conservative estimate of the minimum recur- rence interval for complete erosion of the flat portion of the design berm. Initial computed recession values can not be doubled for the variability analysis because recession is limited by the presence of the seawall; con- tinued erosion beyond the design width would result in scour at the face of 99 the wall. Following this approach, 50 ft of recession for the 100-ft design would have a frequency of occurrence of approximately 35 years. Recurrence intervals on the order of years are indicated for complete recession of the flat berm with the 50-ft and 30-ft designs. Because the continued lowering of the beach is nonlinearly related to the computed maximum recession frequency of occurrence curves, further comparisons of the 50- and 30-ft designs cannot be made without additional analysis. The results indicate, however, that both the 30- and 50-ft berm width designs provide far less protection to the seawall face than does the 100-ft design. 181. Consideration of the potential recession experienced by the 30- and 50-ft design profiles leads to the conclusion that both berm widths would provide insufficient shore protection in the project area. Present model results quantitatively substantiate a recommendation for a 100-ft design berm width. 100 PART VI: STAGE FREQUENCY RELATIONSHIPS Introduction 182. The products of this portion of the study are stage-frequency curves which relate the elevation of flood waters to the average waiting time between floods of equal or greater severity. The ordinate of these curves is stage, measured in feet NGVD, and the abscissa is return period expressed in years. 183. Flooding in the study area is caused by the combination of storm- induced water level and astronomical tide. Storm-induced water level has two main components, storm surge and wave-induced water level. Storm surge is composed of the combined effects of storm winds piling water along the shore and low barometric pressure raising the water surface. The wave component is produced by breaking waves; a portion of the momentum of the waves is trans- formed into a rise in water level called wave setup. This project is pri- marily concerned with the combined effects of storm surge and tide. However, an estimate is made of the contribution due to wave setup in areas where this effect is considered important. 184. Two distinct classes of storms affecting the study area are north- easters and hurricanes. Northeasters, named after the predominant direction of the associated winds, are large-scale, low pressure disturbances which usu- ally occur from late September through April. Wind speeds associated with a northeaster are generally less than those of a hurricane. Although gusts can reach hurricane strength in a very severe northeaster, sustained wind speeds are rarely greater than 50 knots. Flood damage caused by a northeaster is a function of the storm's duration and intensity. Storms of longer duration are more likely to destroy both natural and engineered flood protection features. Also, since a northeaster may last several days, the possibility of a storm occurring simultaneously with a spring tide increases, thereby increasing potential flood damage. An average of 2.4 northeasters per year with maximum storm surges greater than 2.5 ft affect the study area (Prater, Hardy, and Butler in preparation). The maximum recorded northeaster storm surge near the study area was 8.5 ft, which occurred during the November 1950 storm. 185. Hurricanes are a rarer occurrence in the study area. By the time hurricanes approach the latitudes of the northern New Jersey coast, they are 101 usually in a stage of rapid decay and are far out to sea on a path that is curving away from the coast. The average waiting time between hurricanes passing within approximately 100 n.m. of the study area was found to be 5.7 years (0.175 hurricanes per year) (Prater, Hardy, and Butler in prepara- tion). Despite their infrequent occurrence, hurricanes have the potential to cause devastating flooding in the study area because of the large storm surge caused by high wind speeds and low pressures. The duration of a hurri- cane is typically shorter than that of a northeaster. 186. Stage-frequency curves were developed using a probability model in conjunction with a numerical storm surge model. The calculation technique can be outlined as follows (Figure 49): a. The probability model selected storms and assigned probabil- ities to each (no tide) to create separate northeaster and hurricane ensembles. b. Surge level time-histories at the boundary of the study area were obtained either from gage data or numerical simulations from the FIMP study. ce. Each storm surge time history in the two ensembles was com- bined with a large number of tide time-histories to create two very large ensembles of synthetic northeaster and hurricane surge plus tide events. PROBABILITY MODEL STORM SURGE MODEL EVENT SELECTION PROBABILITY ASSIGNMENT STAGE-FREQUENCY CURVES MAXIMUM FLOOD LEVELS Figure 49. Flow chart of project technique 102 d. Events were selected from each of the ensembles to be simulated by the storm surge model. e. The storm surge model was calibrated and verified for the study area. f. Each of the selected northeaster and hurricane events was simulated by the storm surge model to produce a time history of surge plus tide water levels throughout the study area. g. At various locations throughout the study area the maximum still water level produced by each simulated event was assigned the probability represented by that event. After all selected events were simulated, the resultant maximum total water levels and corresponding event probabilities were used to create stage-frequency curves. Review of the FIMP Study 187. Since the present study area lies within the geographical boun- daries of the FIMP study, this previous project served as a "parent" study. The present study used techniques from FIMP to refine and extend results into areas which were outside the initial FIMP study area (New Jersey coast) or not modeled in FIMP (Shrewsbury and Navesink River basins). Due to the dependence of the present study on FIMP, the following short review will be provided. 188. The FIMP study investigated the frequency of storm plus tide water levels along the coast and within the bays of southern Long Island, New York. In order to model storm surge it is customary to extend the computational grid beyond the edge of the continental shelf into deep water. Since it is also desirable to have small cell sizes in areas of interest, a large number of grid cells may be necessary to model a study area using a single grid. Conse- quently, in regions with a wide continental shelf, as in FIMP, a two-grid system is often developed. A global grid with coarse resolution extends throughout the study area and past the edge of the continental shelf while a nearshore grid, with a much finer resolution, covers only the immediate study area. A storm is first simulated on the global grid. Then, using water sur- face fluctuation time-histories computed on the global grid as boundary values for the nearshore grid, the storm event simulation is made for the immediate study area. For FIMP, the global grid covered the New York Bight from a point south of Atlantic City, New Jersey, to beyond Cape Cod, Massachusetts, includ- ing New York Harbor and Long Island Sound. The global grid encompassed the present study area but did not resolve the Shrewsbury and Navesink Rivers. 103 The FIMP nearshore grid covered only the coast and bays of southern Long Island. 189. The scarcity of historical water level records for southern Long Island required a synthetic modeling approach to generate the large number of independent events (water levels) needed for construction of stage-frequency curves. For hurricanes, the joint probability method (JPM) (Meyers 1970) was used to create synthetic storms. An individual hurricane can be represented by five parameters: central pressure deficit (DP), forward speed (FS), radius of maximum winds (RM), track angle (TA), and landfall point (LP). Probability was assigned to an individual storm by determining the probability of each parameter value in that storm. If the parameters are independent, the storm probability is the product of the probability of each component parameter. However, in the New York Bight, not all parameters were found to be indepen- dent. There were two "dependency limbs," one involving DP above and below 2 in. Hg, and one involving bypassing storm tracks. A total ensemble of 918 synthetic hurricanes was generated from all possible combinations of selected parameter values within each dependency limb as shown in Table 10. Table 10 Construction of Synthetic Hurricane Ensemble for FIMP Study Parameter Combinations Dependency Limb DP RM FS TA LP Total Landfalling Storm Tracks High Pressure 2 3 3 3 5 270 Low Pressure 4 3 3 3 5 540 Bypassing Storm Tracks 6 3 3 2 1 108 Total 918 —— ———————————————————————————— 190. Northeasters are more difficult to parameterize than are hurri- canes; therefore, historical data were used to establish a northeaster storm ensemble for FIMP. Twenty-seven storms were selected as representative of the 41-year period of 1940 through 1980. Water level data, after the subtraction of predicted tide, were used to develop a partial duration stage-frequency curve of northeaster surge levels at Sandy Hook, New Jersey. Probabilities 104 were assigned to the 27-member storm ensemble according to the portion of the stage-frequency curve that they represented. 191. The interaction of surge and tide is significantly nonlinear in shallow water. Also, the contribution to flooding in back-bay areas from the overtopping and breaching of barrier islands is a highly nonlinear process. For these two reasons, in shallow water, surge and tide cannot be separately modeled and then added together to produce a combined water level. However, in FIMP, the boundary of the nearshore grid was in relatively deep water and far enough away from the barrier islands so that it was unaffected by the above-mentioned nonlinearities. Therefore, it was possible to simulate surge without tide on the global grid, and then linearly add tidal time histories to the surge time histories. By combining multiple tides with each storm, a large number of synthetic surge plus tide events were created. This process created more than 600,000 hurricane and 18,000 northeaster surge plus tide time-histories. 192. With the assumption that a storm has an equal probability of starting at any point during a tidal cycle, stage-frequency curves were con- structed for surge plus tide water levels. This was done on the boundary of the nearshore model as well as at several open-coast locations such as Sandy Hook. 193. It was not feasible to model on the FIMP nearshore grid any but a small portion of the large ensemble of events created by the convolution of global results with tide. A procedure was therefore devised to select events to be modeled. A total of 51 hurricane plus tide and 40 northeaster plus tide events were simulated. The stage-frequency curve at the nearshore boundary was discretized by height with a single event selected at each height interval to represent all events having maximum water levels falling in that interval. At these discrete locations on the global stage-frequency curve, the maximum water level occurring during each selected event was assigned a probability equal to the probability mass of its interval of the stage-frequency curve. After the event was simulated on the nearshore grid, this probability was assigned to the maximum water level caused by the event at various locations throughout the study area. Thus, stage-frequency curves were created for multiple locations throughout the FIMP nearshore grid. 105 Adapting FIMP Ensembles and Results to the Present Study 194. The present study area is located southwest of the FIMP area. Considering the relatively small spatial scale of hurricanes, it was necessary to alter the FIMP hurricane storm ensemble for use in the present study. All 918 original synthetic storms were retained and one additional landfall point accompanied by a track angle was added. This additional storm track was pat- terned after the 1903 hurricane which made landfall near Barnegat Inlet on the south New Jersey coast. After these new LP- and TA-values were combined with the six DP, three FS, and three RM values from FIMP, an additional 54 synthe- tic hurricanes resulted. The additional 54 hurricanes were simulated on the FIMP global grid. The hurricane storm ensemble for the present project is shown in Table 11. Surge time-histories from a numerical gage located at Sandy Hook were used to create the hurricane surge plus tide ensemble for these 972 simulations. Table 11 Parameter Values of Synthetic Hurricanes DP RM FS TA LP X NG No. (in. Hg) (n.m. ) (Knots) (Deg )* (n.m.)** (n mi) ** 1 0.9 20 12 54.5 7.4 13.2 2 een 36 19 34.5 25.4 Vthos 3 eal 50 27 -0.5 45.4 23.0 4 1.4 69.5 64.2 28.0 5 ear 14.5 88.9 32.9 6 2.3 -28.5 94.1 8.2 if -10.0 49.4 * Referenced clockwise from north; direction toward which the storm is traveling. #* The landfall point was defined as the distance in nautical miles from the origin of the numerical grid of this study (40°18' N, 74°06' W) to where the storm landed. 195. Considering the relatively large spatial extent of northeasters, the 27 historical northeasters from FIMP were assumed to be adequate for the present study. The Sandy Hook tidal gage operated by NOS is located at the boundary of the present study area. Gage data were available for all 27 storms in the northeaster ensemble. After subtracting the predicted tide, 106 these data were used to create the synthetic northeaster events in place of the results from the FIMP global simulations. The maximum surge (total water level minus predicted tide) for each historical storm is listed in Table 12. 196. Astronomical tides were predicted at Sandy Hook using tidal con- stituents provided by NOS. A 19-year tidal cycle was generated for each of the two storm types. Hurricane season was defined as May to October and northeaster season as October to April. Tides for hurricanes were predicted at 20-minute intervals, and tides for northeasters were predicted at 60-minute intervals. 197. To develop a data base for boundary conditions to drive the proj- ect model, two separate sets of surge plus tide event ensembles (northeaster and hurricane) were created by convolving each storm time-history from the global FIMP model with a tide time-history. The convolution process entailed the superposition of each surge time-history on a predicted seasonal tide record at Sandy Hook, starting at the beginning of the storm season and con- tinuing through the season advancing the starting position by one hour. The maximum total water level during the surge event is recorded for each hour shift. To reduce the amount of data to be handled, a random selection proce- dure was developed to select 500 tide combinations for each northeaster (13,500 events) and 100 tide combinations for each hurricane (97,200 events). The recorded maximum water levels for each ensemble (keeping hurricane and northeaster events separate) were ranked by magnitude. 198. Just as in FIMP, these two very large event ensembles could not be simulated by the storm surge model. Therefore, events were selected according to maximum water level from the ranked files described in the preceding para- graph. These selections were made to duplicate the separate stage-frequency curves that were created in FIMP for northeaster and hurricane surge plus tide at Sandy Hook (Figure 50). The probability of a selected event is equal to the proportion of the probability mass which it represents. Events were selected in sets of 20; two sets of hurricane and two sets of northeaster events. By simulating the two sets of a storm type separately and creating separate stage-frequency curves, a measure of the variation in the selection process can be determined. 107 Table 12 Historical Northeasters Used to Generate Synthetic Surge Plus Tide Events Maximum Surge No. Date (ft) 1 7-9 Nov 1947 Bre 2 19-21 Dec 1948 5/5) 3 22-24 Mar 1950 2.9 4 23-29 Nov 1950 8.5 5 4-5 Mar 1952 2.3 6 19-23 Nov 1952 355 T 6-7 Nov 1953 68} 8 14-15 Dee 1954 2.9 9 8-11 Jan 1956 3.9 10 8-9 Apr 1957 1.8 11 18-21 Mar 1958 3.4 12 11-13 Mar 1959 2.8 13 18-19 Feb 1960 4.0 14 3-5 Feb 1961 4.2 15 5-8 Mar 1962 4.8 16 4-7 Dee 1962 Sie 17 12-14 Jan 1964 3.9 18 24-25 Feb 1965 2.9 19 22-24 Jan 1966 4.5 20 26-28 Jan 1967 3.4 21 13-15 Jan 1968 3.0 22 9-11 Nov 1968 5.2 23 25-27 Dec 1969 S140 24 25-27 Jan 1971 Ae | 25 17-20 Feb 1972 4.2 26 21-22 Mar 1973 ea aT 8-10 Dec 1973 Sem 108 ELEVATION (FT NGVD) 3. 8. 7. 6. s. 4. 3. e. dq @.. LEGEND sie iy oa eer rn tee at SURGE PLUS TIOE STAGE FREQUENCY _—-—-— — NORTHEASTER COMB>NEO HURRICANE ANDO NORTHEASTER SANOY HOOK NJ Figure 50. Still-water level stage-frequency curves - Sandy Hook, New Jersey Storm Surge Model 199. The WES Implicit Flooding Model (WIFM) was used as the hydrody- namic storm surge model. The numerical and hydrodynamic features of WIFM are discussed in Butler (1978) and the application of WIFM to coastal studies is demonstrated in numerous reports, including Butler (1983). WIFM solves the vertically integrated, dynamic, shallow-water wave equations of fluid motion using an alternating direction, implicit, finite difference algorithm. The model allows subgrid barriers (which can be non-overtoppable, overtoppable, or submerged) to be included in the grid. An important feature of WIFM is the capability for using an exponentially stretched numerical grid, which permits concentration of grid resolution in areas of interest. Also included in the code is the capability to flood or dry individual cells during a simulation. Numerical grid 200. The computational grid for this project contains 1,938 cells, with 34 cells along the vertical axis and 57 cells along the horizontal axis. The grid covers the area from Sandy Hook Coast Guard Station to Long Branch and the tidal regions of the Navesink and Shrewsbury River basins. Cell size 109 ranges from 178 ft x 540 ft in the high-resolution region of the Shrewsbury River, to 2,720 ft x 3,180 ft along the landward boundary of the grid. The nearshore grid used in the surge model is illustrated in Figure 51. 201. Water depths and land elevations were assigned to grid nodes based on information obtained from: a. A topographic map (1954, photorevised in 1981) of Sandy Hook and Long Branch, New Jersey. NOAA Nautical Chart No. 12324 (edition 22, January 1984). ce. Shrewsbury River map (May 1985) supplied by CENAN. Ina Water depths and land elevations used in the calibration are shown in Figure 52. 202. Channel bathymetries were sparsely marked on both the topographic map and nautical chart. The 1985 field survey was only conducted for a few channel sections: the downstream reach of the Shrewsbury River; the Barley Point Reach; and the lower part of the Long Branch Reach. The lack of accu- rate data on channel bathymetry from Highland Beach to Sea Bright made channel Simulation more complicated and difficult than anticipated. Calibration 203. Before conducting water level simulations, the model was cali- brated and verified using data obtained from the study area to ensure the accuracy of model results. Tide gages were installed at the following five locations (Figure 53): Sandy Hook Bay Marina (TGI), Red Bank (TG2), Rumrunner Restaurant (TG3), Mariner's Emporium at Manhasset Creek (TG4), and Locust Point Bridge (TG5). Unfortunately, data from TG5 were lost and only a small portion of the data from TG4 was available due to mechanical malfunction of the gages. 204. A 35-hr period, from 0700 26 May to 1800 27 May, 1985, was chosen for calibration since usable data were available at gages TG1, TG2, and TG3 and wind data indicated low wind speeds. To account for wind effects during the calibration period, field data from TG1 were used to drive the boundary of the storm surge model. Previous simulations showed negligible differences between the boundary and gage data from the Sandy Hook Bay Marina gage (TG1). The model was calibrated and verified by adjusting depths, frictional coeffi- cients (Manning's n), channel sizes, and back-bay volumes. Values of the coded friction array and Manning's n used in the calibration are shown in 110 suoTjeynduod wu0qjs suoysueau UOJ pTus TeoTuoumN “°|G aun3ty k ig__95 $5 9 £9 29 18 0S oy Lpsperipes 98 ve ce opazorre 22.02 ISLE O62 9 9 py € 2 1 4 E : S + 3 f + = too 4OOH AGNVS meeerace ca HONVus ONOT NISSAVN 40 > ONVIHOIH | | nose as | z | NosWwnY D t t rT I Os T Loy | | | | Ly Le |_| Ke il | SD I ss | or | | Ze [ ras | LLL | H3ATIS [ ii ‘an Lt =n 5 eee ee Hitt Za | ae %S [| / HLL Zz Ie MTT + | HT | -HEPINVa Gay al anlinctinnl ana? + + [ | 4 i 000¢2:1 31v9s | 111 “thee Ee, We F. wou TK Anon 8 8 fi) a 1 2 3 Cy, th fh Ch eb 9. «99. «99% 9. 9 (99. 9%. 99, je 2591) 9 Ch bh teh Sb Ch 9, 9%. 99% fh eh OG) 99, 8. 8. B Ch Ch fh CE 99. 99. 99. Eh fy) Cee {Ch 9. 9. 9% Ch <0) Gh £5 99. 99. 98. 99, 99, 99. 99, 99. 99. «59. 9950 1299! 9 995 meS9s 99. 99, 10. Ch Gh Gh oh 9. 99. 10. a tags Noah eg 99. 99. 10. ch oy oy ce 33. . D Ch Gh Ey th 33. Gh Gy GC) oh 33. Oh Eb Eb th 2: 9. «999. (99%, 99, 9. 995) 199) 199. 99) 59. Ch Gh GR Sh 33. Oh Eb €h Cy 99. 9. 9. 9. 99, 9. 4, 99. 99. 99. 99, 9. 4. Ch ih ©) Gh 9. 4. Ch Ob oh bh 39. 10 if Ch yee te 9. 10. 6. in ete 99. 99, 99: 99. 99. 99. «99. 10 9, 99. 99. 995) (OSes: 99, 99, 99. 99. 9, 99. 99. 99, 99. 99, 99. 99. 99, 99, 99 99; (Sores Figure 52. Water depths and land elevations used {n calibration of tides BARRIER (depths given in ft MLW; elevations given in ft NGVD) = BSESsSEEEssssss 3 B8SBS8S5-° BSSSSS SSS a) SESSBERES SYSSERssssss LS = XVINH soses [eoTuaumu pue saves apty jo suot e007 ANISSAVN 39vV9 1VOINaWAN (1) 3ovO3;qdll @ GN3931 "€G aunsTy L= NIWH ve= XVINA L= NIWA 115 Figure 54. The computed gage data were compared with field data for gages TG1, TG2, and TG3 (Figures 55-57). 205. An excellent match in phase and range was obtained between model and recorded tides at Sandy Hook Bay Marina (TG1). This was expected since TG1 gage data were used as boundary conditions. Good results at high tides were obtained at Red Bank (TG2) and Rumrunner Restaurant (TG3). Differences between tide gage data and model results at gages TG2 and TG3 were less than 0.10 ft and 0.15 ft, respectively. The model predicted a later and higher low tide, 7.5 min and 0.1 ft, respectively, at Rumrunner than was indicated in field gage data; however, this discrepancy had little significance in this project. 206. Although a small difference between calculated and measured water level is indicated at hour 39 for Red Bank (Figure 56), the computed peaks follow a pattern similar to that predicted in the NOAA Tide Tables (1985) for the given time period. The discrepancy may be related to error in the gage data and/or a meterological factor which was neglected in the calibration process. Verification 207. The calibrated model was verified for a time interval of 39 hr, from 0400 June 3 to 1900 June 4 1985. The results, presented in Figures 58 through 60, show the model adequately predicted the tide gage data during this period. 208. Two historical storms, the March 1962 northeaster and Hurricane Donna (1960), were simulated on the nearshore grid. Data from the Sandy Hook tide gage were used to supply the boundary condition to drive the storm surge model. Wind speed and direction data were obtained from the Standard Project Hurricane (SPH) windfield model (National Weather Service 1979) for Donna and FIMP for the March 1962 storm. 209. The most accurate method for calibration and verification of the historical storm would involve using historical inlet and channel configura- tions. However, this was beyond the scope of this project, and present-day bathymetry was used instead. 210. Results of both historical simulations are compared with water levels reported in "Flood Mark Determination for Selected Storms" prepared by VEP Associates, Inc. (1985). The comparisons, given in Tables 13 and 14 for the 1962 storm and Donna, respectively, show a good match between the computed 116 55 50 45 40 35 30 25 20 15 10 ORO OOOO OMO MOM OOS 0s OO mol 0m0 OO MO SOS ONO Os Sis SmS 5 mr4manit OOOO Om ORS mSaeSas5i4a4 CODE MANNING’S RANGE OF ELEVATIONS OO MOM ORO M0000 0l0 0510 Om ON OO me OO Om Om ORO ORO On 0) 1 2 3 4 5 6 7 8 9 10 V1 1 OO OL OO OmsO mn OMOMON0 ORO OR OO OOF 007050 OU ORO Os O0 Ol0 00020 Nea = 22) oie Ont Ol Ol 10) 1050/5 10)80 1 1 ONO MOM OMS 0 OOOO On OmOK-Si-3 OM OOO OOM OS Os OO ROOM ROIs 0/10) 10) 010) 10) (0-0) (0) 0) <0) 50) 00m 0F-4 OF 00S 00 OMOn ORO 0s Os 0m:0 1 1 Ops OO Oles Ole Os One Ol SiG a- Gia: 0000000 5 4-4 -4 OR OM OR OR OR ORs OS c- 5-4) Oo OS000 O-0 8 OOS Of Ont One OlOle5 1 2 3 4 0.023 0.025 0.028 0.035 0.040 0.030 -35 to -5 ft 1 -5 to -1 ft 1 -1 to 1 ft 50) 0) 50570810) 50 1 1 1 1 OOM Oke Ol One 0) OOmn0l0 030m Ole Olen Oke5: ONO Si ONO OOO) 00s OOM On 0) Om 0h Om On Ole 5 OM ONS = 4 550 Ol 0 ee Os On Ons OOO cs OOM ORR OlE5 O10) 52-3. 65.55 0-5) Si Ob 022010 OlsOmOm Ons ONsO-kOmn5, 0000000000000 0 1 lat omesitit 5 to 20 ft Constriction On0 0 0N50 ORO OR ORO) 1 1 1 1 1 1 1 1 5 uv 1 1 1 1 1 OS 5)| = 21-4 Ske 3-5 Ol Olen Ons O nO On Ole Om Ule Ot S a5, Op iO 0: On Oh Oka Ole Olu Os On Olt Oe Ol Oe Ol Sa Sa Om, Op Sx=5))=20-25-1 OMS ORs OP OOO) 5205) OS ON Oe 5a Ser5ae5 115 00 0 ij 0 000 1 1 1 1 1 OP OMO} {ONO Ol O SOOM OM 0 ORO ROM ORs OM Om Oh 5a 5 ec5be5, 5S OlNO) 2000) 10) 500) Ol Opn Om Ol OO Ol OR OM O05’ 1 1 1 1 1 On 5 h-5a- 3-4 -5a- 4 535 ae Onl Oe On OO ORs etO Onn Olt 12 13 14 15 16 17 1 OM OMOMOMO MONO ONO 0 0F O00. 0n0e:5. 1 lied 1 1 LSU 1 1 1 ORODONS EON 0) ORO) Og ONO OMOR OE 0b0 © 2)) = 4 ee eee Os Oe Ole Ol Ole Ot Os ON OO ON-5i1-2: -1 -1 ORRO ORO ORO OOOO 0) 181 1 ORO OOM ON IOh OOO OOM OlNO} Ih OBS Fo) O.0-0 040-020: On. 0 Sl 1 1 1 1 1 1 1 1 1 1 1 0-5 -5|-4 0 Of-2 OsSis5) 1 -1 1 1 1 rscrd BPS 8 OO) OO O-O-0F0 3-5 8 1 1 O20! 505 O50) (0) ORO 0000 1 4 =1 1 1 ei 2is= 2) = Sige tO Ol Ole OleOle Ol Ole See 5 0-5. 1 -1 Wea 28-25-2510 OOO) Onn On 00 On-5ic-4. Veale 2S Se oe SOMO OM Ol Sion 5-2, 1 vis bow ie{SeSie52 10 1 1 1 Nece= 2a 2 eel pal 1 Je Ait ec 1 1 ORROR iOS Ole OF Ot Om ODOR Ol 5 in: O}SO0i 0) 100) SOMO Om 10) 10k.5a85: OO OF Ole Os Os ORON One 4-5) OOO OOP On OM OMS 45-4) OOM ORO O ROMO S144 Ue ud 4 Vo A 1 1 1 4 pee A a Cob Ce Pyocb yok j -1 =2 2. SS Bo. (0 by ee} coos} <2 UES BOBS WOW) By 5) br 1 1 vA 1 1 1 1 1 1 =i] ch} of] ~2n=SE-5 yor der) =2) =2' -2 1 1 =i1 1 1 1 0-4-4 -5 -4 OS k~ 454) 0) By} O. 0-5-4 5 4 Oo B= G On SiS 5. 19 20 21 117 ig G5 5 0 Oo @ Se Oma 1 arse) pers =2eseace =i =i 1 1 4 1 1 “1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 of] Fey 4 Oy Oy) Bi ers cr O73! Cd] ehcthn bh Bhs Gnily eb 4] 124) 1 1 22 23 TY eer aoe IS Sy) = ) 0) 1 1 1 l= = ead 1 1 -1 eal Sef Of OO. 0 © et esys or) crs ath rd cr decr) sch] -1 1 1 OR OOF OM Ot ORs Oi Siac4 Oo OO OW G4 Baby bh eB) Se S33) Bre hd] 24 et eae WO 5 O OO 0 Asa it= 24155 ORs OF 0) 0 Om On O00 -1 1 rats 34 BR Be Be sy 5 Reeord crs cr BB Ss Baby Bh 06} PY 073 of? 97 BB eB By Bec Pes od cd 72 (yb) Sh By bh) Bee ced od cyt cr 1 1 1 1 =a 1 1 1 1 1 1 1 1 1 1 1 1 1 U.S 74 ce} 14 1 1 1 1 1 ee Ob 1 1 000 5 4 4}-1}-3 26 27 28 29 30 = 355 Oe ONO; Ol OOM OM ON Oln0 1 1-1-3 -4 1 1 1 1 0) OO O- 0 5-9 9 SS Seeds 000 5 4 -4]-2}-3]-1 55 ee Olea Ofes Ot Ole Os Os Om Ol O00 1 000 5 4 -5]-2]-4 -4 ORO SO MOR OMS i4s Sea5 Be 5 One Ole On Oka Ol On OF Om ORsOM 00 1 5250 OOS OR OM OO RORIOMOMOmI0)2'0 1 G5 OO ORO ORO 0M0M Om0m0n0n0m0) 10 4 1 UY tikes Oe 1 1 OF 10) 10) OOo oS Se -Si=20 21-305) 4) 0 OOO O05 0 Oebiete betes Sao, ha 31 GEG 5 5 eee Oe Oss Ome Oe Oe OO ee Ol Onn Oe OPO Om 000) 10 Sie 4455 ol Ole Of O ee Ot Om Oe Ole! Oe Ont Oba Olt On Om Os OOP ORO es Ol OM, 1 1 4 4 1 32 1 1 (et O 0-0) 0) UWS0- WU 00. SS So Ss 24114) 33 ORO MOR OF OO ION Ole Ol Om Oke Ole Ole Ol OF Ol Ole Of Om Onn Op Ole One OS sOnn Oe en Ol Of Ole Ole Ol On Ol Os Ol Ole Ole Olen One Of Of Olen On 0-1 Oli Oe Onn Oe One Oe On Ons One Ome Onn 34 0 value represents upland cells Notes (-) denotes cells which can be either flooded or dry ive sign Negat Values of the coded friction array and Manning's n Figure 54. the calibration of tides in ts used icien 2 coeff CURVE GAG HOR VER GAGE NAME 41 2e GAGE 19(SANDY HOOK BAY MARINA) RAU: a > o z b ra z ° b < > ve) = es) TIME (HOURS) GAGE 19 (UIFR) US FIELD DATA (SN.15@ SANDY HOOK HARINA 5-26-85) Figure 55. Tide calibration data - Sandy Hook Marina (TG1) CURVE GAG HOR VER GAGE NAME e8 16 6 GAGE 28(RED BANK) scinbsc 1 RAU: HOUR - 088 iS Cer ee Sri Ss 1 wW ELEVATION (FT NGVD) mo Ps ee ee ae aes Ese Sea zSaA 9 13 17 21 es 29 33 37 41 TIME (HOURS) GAGE 28 (UIFM) US FIELD DATA (SN.156 RED BANK S-e6-85) Figure 56. Tide calibration data - Red Bank (TG2) 118 a > oO z be we z fo) be < > w ~ ws CURVE aye nek yer GAGE NAME GAGE Senate REST., SEA BRIGHT) RAW: HOUR TIME (HOURS) GAGE 21 (UIFA) US FIELD DATA (SN.155 SEA BRIGHT S-26-85) Figure 57. Tide calibration data - Rumrunner L--) ELEVATION (FT NGVD) =i Figure 58. Restaurant, Sea Bright (TG3) CURVE GAG HOR VER GAGE NAME er 19 Ailianac ©) GAGE 19(SANDY HOOK BAY MARINA) Shoas 1 RAU: HOUR - 880 | TIME (HOURS) GAGE 19 (UIFM) US FIELD DATA (SN.150 SANDY HOOK MARINA 6-3-85) Tide verification data - Sandy Hook Marina (TG1) 119 a > o z b ve z ° = < > ws =i w CURVE GAG HOR VER GAGE NAME 28 16 6 GAGE e8(RED BANK) 24 20 22 26 TIME (HOURS) GAGE 28 (UIFA) US FIELD DATA (SN.156 RED BANK 6-3-85) Figure 59. Tide verification data - Red Bank (TG2) ELEVATION (FT NGVD) CURVE GAG HOR VER GAGE NAME et el 20 33 GAGE 21 (RUMRUNNER REST., SEA BRIGHT) BSOaaS 1 RAU: HOUR - 880 6 _ HALEN TH AEA TEEN EEE GAEL LEELA FNL LEAEEEET TTS \ y N i ra i N / LLL TELS EN PLEELLLESeA LLL itz 4 Ne H My i INE i H Ht ine 8 12 16 20 24 e8 3e 36 48 44 TIME (HOURS) GAGE 21 (UIFA) US FIELD DATA (SN.1SS SEA BRIGHT 6-3-85) Figure 60. Tide verification data - Rumrunner Restaurant, Sea Bright (TG3) 120 Table 13 Flood Mark Comparison for the March 1962 Northeaster Flood* WIFM Point* Grid Cell Mark Max. Elev. Max. Wave Effects No. Location* (Y,X) (Et) (ft) (Git) 33 Highland - Beetle's (36,26) Uo 8.2 - Drug Store, Corner of Miller St. & Bay Ave. 34 Highland - Bay Ave. (37,24) 8.5 8.2 - in front of Katz's Confectionary Store 35 Highland - Inter- (36,26) 7.3 8.2 - section of Bay Ave. & Miller St. 36 Red Bank - Marine (156) 9.1 7.0 4.5 Park, at Bulkhead 37 Red Bank - Marine (15, 6) 10.6 7.0 4.5 Park, at Bulkhead Oo Little Silver - at (6,14) 9.4 6.6 3.9 residence on Point Rd. 44 Little Silver - (4,15) 5.4 6.8 3.9 Gooseneck Point Rd. * From "Flood Mark Determination for Selected Storm Events," prepared by: VEP Associates, Inc., (1985). Table 14 Flood Mark Comparison for Hurricane Donna, September 1960 Flood* WIFM Point* Grid Cell Mark Max. Water Elev. No. Location* GX) (&t) (ft) 25 Highland - Bay (36,26) Loo 9.1 Ave. & Valley St. by Katz's Confectionary Store 26 Highland - Bay (36,26) 9.0 9.1 Ave. at Katz's Confectionary Store (Continued) * From "Flood Mark Determination for Selected Storm Events," prepared by: VEP Associates, Inc., (1985). 121 Table 14 (Concluded) Flood* WIFM Point* Grid Cell Mark Max. Water Elev. No. Location* (Y,X) (iit) (ft) 28 Rumson - Inter- (17,29) 10.4 6.9 section of Waterman Ave. & Grant Ave. 29 Red Bank - Marine (15, 6) 8.6 Uoit Park, at Bulkhead 30 Red Bank - Marine (155) 6) 9.8 est Park, at Bulkhead and reported water levels at locations 33, 34, and 35 for the 1962 storm and locations 26 and 29 for Hurricane Donna. 211. Model results at Red Bank and Little Silver, in the upper reaches of the Navesink and Shrewsbury basins, respectively, are typically much lower than observed water levels. These areas are susceptible to significant in- creases in water level due to wave setup and wave crest overtopping. As de- scribed above, wave setup is not included in model calculations and is consi- dered insignificant in most of the modeled area. To include these effects, procedures similar to those used in the FIMP study were adopted. These proce- dures reflect the physical principles involved, but are greatly simplified. The maximum wave setup is shown to be 0.15 times the significant wave height. Maximum significant wave heights for the gages in question are approximately 7 and 6 ft, respectively, resulting in wave setups of 1.0 and 0.9 ft. An effec- tive flood elevation, or wave crest elevation, can be expressed as the surge plus tide plus wave setup elevation plus an additional 0.5 times the signifi- cant wave height. This calculation allows a total maximum water level of 4.5 and 3.9 ft above the model surge level at Red Bank and Little Silver, respec- tively. The reader is cautioned to interpret the results with care. These additional levels represent a maximum wave effect occurring at the peak surge plus tide level. However, the difference noted between observed and calcu- lated results is expected since additional water level due to wave effects is not included in the model. 212. Results for the Highland area (near VEP gage points 25 and 26) are good for Hurricane Donna. The 2-ft discrepancy between the flood mark at 122 gage 25 and that at gage 26 cannot be supported. The 9.1-ft model elevation is representative of the entire lower Sandy Hook Bay area. The comparison at Red Bank can be explained by the omission of wave effect impacts. Winds associated with Donna blew straight down the Navesink basin, peaking at 60 knots just prior to the arrival of the peak surge level. The duration of these winds was much shorter than those experienced during the 1962 north- easter, and consequently the additional water level due to wave setup and wave crest overtopping should be significantly less than that shown in Table 13. 213. The location of VEP gage 28 (Rumson) is on the south side of the entrance to the Shrewsbury basin. As shown in tidal simulations, a signifi- cant hydraulie loss occurs across this entrance. Surge plus tide levels in the south Navesink basin are approximately 7.5 ft. With a loss across the entrance and southerly winds blowing, it seems highly unlikely that water levels could reach 10.4 ft as reported by VEP. 214. To make an additional check on the accuracy of using Sandy Hook data to drive the nearshore model and to add confidence to the augmented JPM curve developed for Sandy Hook, a comparison of the Sandy Hook stage-frequency curve with historical data was made. The comparison, presented in Figure 61, confirms that procedures used in the JPM process and boundary condition data are appropriate. Simulation of Storms 215. Two sets of 20 hurricanes each were simulated on a Cyber 205 com- puter using the calibrated and verified storm surge model. Boundary condi- tions for each surge plus tide event were generated from the FIMP data base. The time from the start of hurricane simulation to the time of landfall was computed within WIFM based on the ratio of RM to FS. Faster moving storms require less time for simulation than do slower ones, and hurricanes with large spatial extent require more execution time than smaller storms. The fastest storm was simulated for 18 hr, beginning 12 hr before landfall and ending 6 hr after. For the slowest storm, the simulation was initiated 24 hr before landfall and continued for 10 hr afterward, for a total of 34 hr. For each of the 40 surge hurricane plus tide event time-histories, all maximum still-water levels greater than 6.5 ft NGVD were included in the development of the stage-frequency curves. A time step of 60 sec was used for lower 123 iE eee LERCEREEEEE LCE (GQADN Lj) NOILWA313 124 RETURN PERIOD (YEARS) SURGE PLUS TIDE STAGE FREQUENCY NO ———— MSL 1981 (6.6 FT NGVD) HURRICANE —.—-———— NOR THEASTER : < : : : : : = i p 5 8 8 a Comparison of JPM results with historical data at Sandy Hook, New Jersey Figure 61. intensity storms, whereas a 30-sec step was required for accurate simulation of high-intensity events. 216. Figures 62 through 63 display example wind field patterns and isovelocities for synthetic storm No. 847 for simulation hours 9 and 12. This storm had the following parameters: DP = 2.3 in. Hg, RM = 36.0 n.m., FS = 19 knots, TA = -0.5 deg, and LP = (64.2 n.m., 28.0 n.m.). Similar data were developed on file for each storm. 217. Two sets of 20 selected northeasters were also simulated. Simu- lation time for the storms ranged from 14 to 26 hr. A constant time step of 60 sec was used for all northeaster simulations. Development of Stage-Frequency Curves 218. The method used to develop the stage-frequency curves is similar to the one used by Hardy and Crawford (1986). For each numerical gage, an array of stage intervals each with a width of 0.1 ft was created. The proba- bility of each storm event was added to the interval bracketing the maximum water level computed for that event. The exceedance, or cumulative probabil- ity, for each interval was calculated by adding its probability to the cumula- tive probability of the next higher interval whose stage is one increment greater. 219. Exceedance was calculated separately for each of the two sets of hurricanes and northeasters producing four separate raw stage-exceedance curves. To smooth these raw stage-frequency curves, linear regression was performed on the cumulative probabilities to compute a straight line through the data. Most linear regression lines resulted in minimum correlation fac- tors of 0.98 for the hurricane sets and 0.96 for the northeaster sets. Fig- ure 64 contains a plot of both raw and regressed stage-frequency curves for the numerical gage at the west end of the I-36 bridge (gage 6). 220. Confidence in the results is obtained by analyzing the variability in the stage frequencies generated by each set of hurricanes and northeasters. Results of this analysis for gages at the entrance to the back-bay river sys- tem (gage 6) and in the upper reach of the Navesink basin (gage 12 near Red Bank) are displayed in Figures 65 through 68. By averaging the two sets of regressed curves for both hurricanes and northeasters, a single, more accurate regressed curve for each gage is obtained. These stage-frequency curves for 125 (A) WINDOFIELD wo o¢ 2 [=] tf z Oo oO <= oOo z M i=] ~ ISOVELS ) ( Wind field and isovelocities, Figure 62. 847, hour 9 design storm No 126 yn cS oO z uM Ls] w (A) WINDOFIELO \ ’ 1 v i] i UJ , Mee wwe oe” 1@ KNOT CONTOURS ISOVELS (B) Wind field and isovelocities, Figure 63 847, hour 12 design storm No 127 Figure STAGE (FT NGVD) 8 ? 6 5 4 3 e 1 ) 10 cs 5@ 100 200 RETURN PERIOD (YEARS) STAGE FREQUENCY STILL WATER LEVEL REGRESSION CHECK GAGE 6(U END OF I-36 BRIDGE) LEGEND SELECTED HURRICANES, SET A Rea _ REGRESSED 64. Example of raw and regressed stage-frequency curves STAGE (FT NGVD) acon TES — ser A STAGE FREQUENCY eornmwtr Un DN e eo wu wn Se 10 RETURN PERIOD (YEARS) 200 STILL WATER LEVEL COMPARISON BETWEEN SETA AND SETB GAGE 6(U END OF I-36 BRIDGE) Figure 65. Variability in stage frequency for the two hurricane data sets at gage 6 128 ee or uN W a (=) > ° =z b we w o < (> n” pO PrP MUS Haw w el ° un RETURN PERIGD (YEARS) SELECTED NORTHEASTERS SETA ———— SET B Figure 66. STAGE FREQUENCY STILL WATER LEVEL COMPARISON BETWEEN SETA AND SETB GAGE 6(W END OF I-36 BRIDGE) Variability in stage frequency for the two northeaster data sets at gage 6 eee STAGE (FT NGVD) 9 8 7 6 S 4 3 2 A () = [J ———— SET B Figure 67. LEGEND SELECTED HURRECANES ET A es Se 100 200 RETURN PERIOD (YEARS) STAGE FREQUENCY STILL WATER LEVEL COMPARISON BETWEEN SETA AND SETB GAGE 12¢(E FRONTAGE & EXT. OF HACEN Variability in stage frequency for the two hurricane data sets at gage 12 129 a > o z re we ws °o < B n RETURN PERIOD (YEARS) LEGEND SELECTED NOPTHEASTERS SET A ———— SET B STAGE FREQUENCY STILL WATER LEVEL COMPARISON BETWEEN SETA AND SETB GAGE 12(E FRONTAGE & EXT. OF HACEN Figure 68. Variability in stage frequency for the two northeaster data sets at gage 12 hurricanes and northeasters are presented in Appendix G for each of the 20 numerical gages. Locations of all gages are shown in Figure 53. 221. The required final product is a single stage-frequency curve for the combined hurricane- and northeaster-induced water level. This combined frequency curve was obtained by adding the exceedance probabilities of hurri- canes and northeasters at each stage increment. Combined regressed frequency curves for gages 6 and 12 are shown in Figures 69 and 70; curves for all 20 numerical gages are presented in Appendix G. 222. The combined stage-frequency curves discussed above do not in- clude any additional still-water level increase due to wave setup or wave crest overtopping. Most channels within the river system are narrow and do not provide adequate fetch lengths for these processes to occur. However, the upper reaches of the Navesink and Shrewsbury basins are susceptible to such wave impacts. Standard methods given in the SPM (1984) were applied to esti- mate the maximum wave setup and wave crest elevation that can be expected in these areas. Results for these estimates are shown in Figures 71 and 72 for numerical gages 15 and 18. 130 a > ro) z b uw w oO ro) z b we lJ oO < b o pOrMWUsUND NAL es Se 108 200 RETURN PERIOD (YEARS) COMBINED HURRICANE AnD NORTHEASTER STAGE FREQUENCY ———— TOTAL UATER LEVEL + uAvE SETUI TOTAL UATER LEVEL + SETUP + CREST STILL WATER LEVEL GAGE 1S5(MAPLE AVE. & E. FRONTAGE) Figure 71. Estimation of maximum wave effects in the upper Navesink basin (gage 15) FT NGVD) w o = re ve ce ia q - *) 7 = ' - oa + i = a = q . ss oF ae 7 1 7 i ; n : hi i i 5 a 2 7 7 i = "7 ey i ‘ = x 2 7 ve ? ae cs . 7 Hy) or Yat 1A ier pon Tie ; a alte M Lit sft ical i ex Ciahngeh pas lately i i : «