Loading...
HomeMy WebLinkAboutSUP 06-10X2A; AGUA HEDIONDA OUTER LAGOON MAINTENANCE; BEACH EQUALIBRIUM ANALYSIS OF MIDDLE AND SOUTH BEACH DISPOSAL OPTIONS FOR DREDGED SANDS FROM AGUA HEDIONDA LAGOON; 2017-07-07Beach Equilibrium Analysis of Middle and South Beach Disposal Options for Dredged Sands from Agua Hedionda Lagoon, Carlsbad, CA Submitted by: Scott A. Jenkins, Ph.D. Dr Scott A. Jenkins Consulting 14765 Kalapana St Poway, CA 92064 Submitted to: Sheila Henika, P.E. MBA-TM NRG Cabrillo Power Operations Inc. Encina Power Station 4600 Carlsbad Blvd. Carlsbad, CA 92008 Draft: 7 July 2017 Cabl-02 Cab 1-01 sand remaining@ t = + 90 days Middle+ South Beach= 173,053 cubic yards North+ Middle+ South Beach= 214,994 cubic yards Executive Summary: A detailed set of beach profile surveys at Middle and South Beach in Carlsbad CA were provided by Cabrillo Power I LLC, delineating beach surfaces before and after the 2104/2015 dredging of Agua Hedionda Lagoon, (AHL), which placed 229,693 yds3 between the south inlet jetty to Agua Hedionda Lagoon and the Encina Power Station (EPS) generating building. The surveys were accurately performed by Noble Engineers using differential GPS and known historic benchmarks. Three-dimensional CAD models were lofted from the measured points along the three (3) Middle Beach survey range lines (Cab 1-04 -Cab 1- 06) and three (3) South Beach survey range lines (Cab 1-01 -Cab 1-03) to delineate the beach surfaces immediately before beach dredge disposal (based on the 22 December 2014 profile measurements) and immediately after dredge operations were completed (based on the 17 April 2017 profile measurements). When these two surfaces were lofted together in a common reference frame, it was determined that 178,584 cubic yards of beach fill have been retained after placing 229,693 cubic yards on Middle and South Beach between 1 January 2015 and the post- dredge survey on 17 April 2015. This calculates to an average sand loss rate of 473 yds3/day and projects a sand retention time of 485 days. This is significantly longer than the retention time at the North Beach disposal site where retention time projected to only 33 days. To understand the reasons for the contrasting sand retention characteristics of North Beach vs. :Middle and South Beach, a baseline beach evolution study was conducted using the Coastal Evolution Model (CEM) to hindcast the fate of beach fill placed on the three receiver beaches. The CEM was developed at the Scripps Institution of Oceanography with a $1,000,000 grant from the Kavli Foundation, (see http://repositories.cd1ib.org/sio/techreport/58/ ), and is based on latest thermodynamic beach equilibrium equations published in the Journal of Geophysical Research. Inputs to the CEM baseline study were based on measured shoaling wave data, grain size data for the dredged sands, and daily beach fill volumes were derived from the dredge monitoring reports to the Regional Water Quality Control Board (Cabrillo, 2015) and from Cabrillo dredging data bases. Between 1998 and 2015 there have been eight (8) different events when Agua Hedionda dredged sands have been disposed concurrently on all three receiver beaches (North, Middle and South Beach). Historic dredged sand volumes placed on North Beach ranged from 62,030 yds3 to 161,525 yds3, while beach fill placed on Middle and South Beach ranged from 163,996 yds3 to 281,195 yds3 . The CEM beach evolution simulations of these events determined that the minimum sand loss rate occurs when beach fill volumes on the three receiver beaches are equal to the critical mass, which was found to be Vc,u == 79,471 yds3 for North Beach, (see companion North Beach report, Jenkins, 2017), and Vc,it == 200,890 yds3 for Middle and South Beach. The critical mass is the theoretical maximum carrying capacity of a beach fill site for supporting a beach profile in equilibrium. The carrying capacity of a beach is limited by the width of the wave-cut platform in the bedrock on which beach sands have accumulated over geologic time scales. The wave-cut platform at North Beach is only 550 ft. to 600 ft. in width and 600 ft. to 650 ft. at Middle and South Beach. Many of the beaches throughout north San Diego County are perched on narrow wave-cut platforms. The platforms are narrow because they were carved by wave action into erosion resistant bedrock formations during the present high-stand in sea level, and these narrow wave-cut platforms physically cannot hold large quantities of beach sand; and often become fully denuded during periods of high-energy winter waves. Another contributing factoi to the limited cairying capacity of the three receiver beaches is that they are exposed to a prevailing negative divergence of drift caused by the way the bathymetry surrounding the Carlsbad Submarine Canyon produces variable wave shoaling along the length of these beaches. The presence of the Carlsbad Submarine Canyon creates a bright spot in the shoaling wave pattern that diminishes in intensity with increasing distances toward the north. For example, wave heights are locally higher at the inlet jetties than further to the North around Maple Avenue. The prevailing littoral drift transports beach sand southward throughout the entire Oceanside Littoral Cell; but the alongshore imbalance in shoaling wave height causes higher southerly longshore transport rates of sand at the southern end of each of the receiver beaches than at the northern ends. Consequently more sand exits each receiver beach at its southern end due to longshore transport, than enters at the northern ends from sand sources further updrift. This inequality in sand transport rates between the north and south ends of the receiver beaches is referred to as divergence of drift, and when the sand transport rates are higher at the down-drift end of the receiver beaches, it becomes a constant loss system referred to as negative divergence of drift. So, when beach fill volumes exceed the critical mass of the receiver beaches, the excess sand cannot be supported in equilibrium on its narrow wave-cut platform and is quickly lost to the negative divergence of drift. However, this effect is somewhat muted at Middle and South Beach because the AHL inlet jetties and the EPS discharge jetties produce a groin field which impedes the longshore transport at both the updrift and downdrift ends of the Middle/South Beach complex, and provides extra storage capacity for sand on the wave cut platform. Consequently retention times for beach fill on Middle and South Beach is longer than at North Beach. Historically, the CEM baseline study finds that when a standard 1: 10 (rise over run) beach fill template on North Beach is filled to critical mass, the theoretical minimum sand loss rate to negative divergence of drift is 1,495 yds3/day, and the sand retention time is 53 days (see the companion North Beach report, Jenkins, 2107). By contrast, the standard 1: 10 beach fill template at Middle and South Beach historically achieved minimum sand loss rates of 431 yds3/day, and the sand retention times of 466 days. But, when beach fill sand volumes at Middle and South Beach were increased by 41 % over critical mass ( as occurred during the 2000/2001 dredge event when 281,195 yds3 were placed on Middle and South Beach), the retention time is only increased by 7 % from T0 = 466 days to T0 = 497 days. In contrast, over-filling the North Beach receiver site produces an even worse return on beach fill investment. During the 2002/2003 dredge event, 161,525 yds3 were placed on North Beach, (103% increase over critical mass), but the retention time increased by only 26 % from T0 = 53 days to T0 = 67 days, while the sand loss rate increased by 61 % to 2,411 yds3/day. This is an increase in sand loss rates at North Beach of916 yds3/day. Unfortunately, such increases in sand loss rates at North Beach correlate with proportional increases of sand influx rates into Agua Hedionda Lagoon. The 2010/2011 survey data show that AHL sand loss rates also increase when the fill volumes are less than the critical mass. Sand influx rates in 2010/2011 were 519 yds3 /day when only 163,996 cubic yards were placed on Middle and South Beach (36,894 yds3 below critical mass requirements). Bear in mind that the critical mass is the minimum volume of sand required to establish an equilibrium beach profile on a wave-cut platform; and a beach is in its most stable state with an equilibrium profile. But with a prevailing negative divergence of drift along Middle and South Beach, equilibrium cannot be achieved due when there is insufficient sand volume, and consequently sand loss rates increase with a destabilized, non-equilibrium profile. Following CEM beach evolution analysis of the Middle and South Beach historic baseline, attention was given to finding a more effective beach fill template that could increase sand retention using beach fill from Agua Hedionda Lagoon dredging. Beach fill has typically been placed on Carlsbad beaches using a standard beach fill template with a flat backshore platform and a 1: 10 (rise over run) seaward facing beach slope extending down to O ft. MLL W. This convention dates back to the Regional Beach Sand Project, (AMEC, 2002). However, stable beach profiles in Nature have a much more gradual, curving profile with slopes that range between 1 :50 to 3: 100. Formulations of equilibrium beach profiles are found in the U.S. Army Corps of Engineers Shore Protection Manual and later the Coastal Engineering Manual; and the latest most advanced formulation is known as the elliptic cycloid. The elliptic cycloid formulation can account for continuous variations in the equilibrium beach profile due to variability in wave height, period and direction when occurring in combination with variations in beach sediment grain size and beach sand volume. Therefore, a new beach fill template has been proposed here for Middle and South Beach referred to as the cycloid-dune template (see Figures ES 1-6). The shape of the template is based on the extremal elliptic cycloid which is the equilibrium profile for the highest wave in the period of record. But the extremal elliptic cycloid extends below the MLL W tide line and earth moving equipment which spread out the beach fill cannot work below MLL W. So, the template truncates the extremal elliptic cycloid at MLL W and places the residual volume of critical mass (totaling 49,680 yds3) in a back-beach dune that stretches 3,680 ft. from the south inlet jetty to Agua Hedionda Lagoon to the north end of the EPS generating building. While an elliptic cycloid is an equilibrium beach surface, it does not produce a state of zero sand loss in the presence of a negative divergence of drift, which is the persistent littoral drift state along Middle and South Beach. When the divergence of drift is negative, the equilibrium cycloidal beach profile will progressively shift landward as it loses sand to negative divergence of drift, eventually intersecting the basement surface of the critical mass envelope. Once this happens, then the cycloidal shape of the profile is disrupted, and the equilibrium state of the profile is lost. The concept behind the cycloid-dune template is that, as the cycloid begins to approach an intersection with the basement surface of the critical mass envelope, (under the erosional effects of continued negative divergence of drift), it also intersects the base of the dune and receives additional sediment cover as the dune erodes and spreads out downslope across the still intact cycloidal surface. Thus; the dune acts as a restoring mechanism that re-supplies the cycloid with sand lost to negative divergence of drift. The construction method envisioned for the cycloid-dune template begins with building the back-beach dune portion first, starting at the south inlet jetty and adding sections to the dredge pipeline until the build-out of the dune extends beyond the South Groin abeam of the north end of the EPS generating building. Building the dune first creates a "safe" reservoir of sand before the template can be fully constructed, and sand from this reservoir is only released to the lower eroded basement surface during periods of the highest tides and waves. After the buildout of the dune to the southern end of South Beach, the cycloid portion of the template is laid out beginning from the toe of the dune and spreading the material down slope to MLLW, and working back across Middle Beach to the south inlet jetty, removing pipeline sections as the cycloids are completed CEM beach evolution simulations of the Middle and South Beach cycloid-dune templates were run for future conditions with the South Groin removed, and show significant improvements in sand loss rate and retention time relative to the historic baseline. Again, the most efficient use of Agua Hedionda dredged sands occurs when the cycloid-dune template is filled to no more than critical mass (200,890 yds3 for Middle and South Beach), which reduces average sand loss rates on Middle and South Beach to an absoiute minimum of 306 yds3/day, while extending retention time to 656 days. This is a 35% improvement in sand retention time over historical dredge disposal practices at Middle and South Beach. If the cycloid-dune template is filled to more than critical mass by adding more sand to the back-beach dune, then Middle and South Beach retention times will increase beyond 656 days. If the reserve sand volume in the dune on Middle and South Beach were increased by a factor of 2.6 to 129,985 yds3 (producing a cycloid dune equivalent to the historic maximum placement volume of V0 = 281,195 yds3) then retention time could be extended to a maximum of 693 days. But, again, this is not a good return on doubling the investment in reserve beach fill placed in the back-beach dune because retention time is only increased by 5 weeks while the sand loss rate on Middle and South Beach would increase by 33% to 406 yds3/day, (an additional 100 yds3/day of sand loss). This inefficiency occurs because the enlarged dune encroaches further seaward into the middle bar-berm portion of the profile that is subject to more frequent wave attack, and the groin field formed by the inlet and discharge jetties at Middle and South Beach is already filled to carrying capacity at the critical mass of vcril = 200,890 yds3• On the other hand, under-filling the cycloid-dune template, (by building a reduced dune), leads to accelerated sand loss rates and reduced retention times. If the Middle and South Beach cycloid dune templates were filled with the historic minimum beach fill of V0 = 163,996 yds3, (by under-building the back-beach dune with only 12,786 yds3) then sand loss rates would increase to 375 yds3/day and retention times would be reduced by to T0 = 437 days, a 33% reduction in sand retention time relative to the ideal build using beach fill equal to critical mass. The prevailing negative divergence of drift across Middle and South Beach causes the initial cycloid profile in the lower portion of the template to shift landward, and once intersection with the basement surface of the critical mass envelope occurs, there are insufficient sand reserves in the reduced dune to resupply the cycloid in the presence of continued negative divergence of drift. Once the reserve sand supply in the dune is exhausted, the cycloidal shape of the profile is disrupted, and the equilibrium state of the profile is lost. Even so, if the cycloid-dune template on Middle and South Beach were filled to a volume equivalent to the 2104/2015 disposal event ( V0 = 229,693 yds3) by using a dune containing only 78,483 yds3, then sand retention times are still significantly better than what was achieved using the standard 1: 10 (rise over run) template. With this over-built dune in combination with the cycloid, retention times following the 2014/2015 dredge cycle could have been T0 = 676 days with sand loss rates reduced to 340 yds3/day, an improvement of 39% over what was achieved using standard Middle and South Beach disposal practices. 25 24 23 22 21 20 19 18 17 16 15 i14 ..J 13 ..J 12 ~ 11 !E..w c:: 9 ,g 8 ! 7 iii 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 ------------------------------ Ii, I j ., \ :,· \, \ ,\ '1 u l \ \ .r \ \ l\.ll; ~ '- South Beach Profiles, Cab 1-01 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune \ ..... --.... -.... ..... ...... --....... '• --.... r--. ---~ -s,.,. ... ~--....,_=--0:0?.1«1 ~"Vij ,, ,. I ~ ~-' ' 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 Distance offshore from benchmark (ft.) Figure ES-1: Proposed beach fill template for South Beach range Cab 1-01 , based on the extremal equilibrium profile truncated at 0 ft. MLL W with a back beach dune to hold-and- release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1994408.5 Easting (ft): 6228847.4 25 24 23 22 21 20 19 18 17 16 15 i14 ...J 13 ~ 12 . 11 E-10 C 9 ,Q 8 ~ 7 ~ 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 ------------------------------- l l. I,, \ "· ,-- -Ii I ,\ I ,, 'I 'I ''\ ~ I 1,\ --r-! "' \ . ,1: South Beach Profiles, Cab 1-02 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune .... ..... --:--r-., 1 ~----' --..... I\,: -----"' ---..... ---,. ' -.:it " ~- 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 Distance offshore from benchmark (ft.) Figure ES-2: Proposed beach fill template for South Beach range Cab 1-02, based on the extremal equilibrium profile truncated at 0 ft. MLL W with a back beach dune to hold-and- release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1995102.9 Easting (ft): 6228562.5 25 24 23 22 21 20 19 18 17 16 -15 3: 14 .J 13 ~ 12 . 11 S-10 ,:: 9 .Q B i 7 w 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 . ---. -----. . . . ---------------- .I /\ \ \ j', .. I \ "' ... ,~ ... \ 11,,' . \.. ..... ..... . ' I'.. K. ,---- South Beach Profiles, Cab 1-03 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune ........ .... ' ,. ..... ' ~ .... ~f"-,.. ~ ~ .... ~ ..... -r--. ... I .... ----~-" ~ .... ...... ··•,,v -- ~I'-1 - o ~ ~ ~ oo 100 1~ 1~ 1~ 100 ~ ~ m ~•~~~~~a Distance offshore from benchmark (ft.) Figure ES-3: Proposed beach fill template for South Beach range Cab 1-03, based on the extremal equilibrium profile truncated at O ft. MLL W with a back beach dune to hold-and- release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1995576.9 Easting (ft): 6228365 25 24 23 22 21 20 19 18 17 16 15 ~14 ..J 13 ..J 12 ~ 11 S-10 C: 9 .Q 8 m 7 iii 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 ------I -(" -, -...: ~ --------------------- ~ ;:;, I ,, \ I'. '\. " ' ' ' 1\1. ,. "" \. ~ r,,. ""I'-.. "'"=111 ~ Middle Beach Profiles, Cab 1-04 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune ...... ' .... ....... ~N. ........ ' ..... _ ' ...... .... ..... I -----... ----.___ I ~---1r '" r--.. i'' -~ ' o ~ ~ oo oo 100 1~ 1~ 100 100 ~ m ~ ~ ~ ~ ~ ~ ~ ~ ~ Distance offshore from benchmark (ft.) Figure ES-4: Proposed beach fill template for Middle Beach range Cab 1-04, based on the extremal equilibrium profile truncated at 0 ft. MLL W with a back beach dune to hold-and- release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1996164.9 Easting (ft): 6228090.5 25 24 23 22 21 20 19 18 17 16 15 i14 ...1 13 ~ 12 . 11 :!::.10 C: 9 .Q B I , jjj 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 ---. ----------. . ---------. ---- ...... ..... -'" \. ' ,.1 I\ I \ \ " ' j ' I ' I ..... I'\. ,_ ~r--... Middle Beach Profiles, Cab 1-05 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune \. ~ \ '\. ' '-' ' ' ' ' ..... 'r(. ~ .... ~I"""-.... ---...... --'~-.......... ---- 0 ~ ~ 00 00 100 1W 1~ 100 100 ~ ~ ~ ~ ~ D ~ ~ B ~ a Distance offshore from benchmark (ft.) Figure ES-5: Proposed beach fill template for Middle Beach range Cab 1-05, based on the extremal equilibrium profile truncated at O ft. MLL W with a back beach dune to hold-and- release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1996778.5 Easting (ft): 6227826.9 25 24 23 22 21 20 19 18 17 16 15 i14 ...J 13 ~ 12 . 11 !S 10 C 9 ,Q 8 iii a, 7 w 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 ----------'""" -------------------- -r'II:- ""'st ~ M 'I ":,, '',;' Middle Beach Profiles, Cab 1-06 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune A U\ \ I~ I • J \ j J ~ i]tr· ,, ' ii-~:'~ • '\ _ :!,1r~ .. 1. . ,\. " ,j ' \., ·~ '-"I ~ ··'·'"'- " ~ ... " ....... "'-,.__ ...... '-...... -------I ..... -----1 .... I....._,"" ..... _ ~ ' ·- 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 Distance offshore from benchmark (ft.) Figure ES-6: Proposed beach fill template for Middle Beach range Cab I-06, based on the extremal equilibrium profile truncated at 0 ft. MLL W with a back beach dune to hold-and- release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1997015.5 Easting (ft): 6227720.2 • Beach Equilibrium Analysis of Middle and South Beach Disposal Options for Dredged Sands from Agua Hedionda Lagoon, Carlsbad, CA by: Scott A. Jenkins, Ph.D. 1) Beach Profile Surveys and Dredge Disposal: Six beach profile survey range lines were monitored on South Beach and Middle Beach before and after the 2014/15 Agua Hedionda Lagoon dredging event. These surveys are plotted in Figures 1-6, and labeled from south to north as: Cab I-01, Cab 1-02, and Cab 1-03 on South Beach; and Cab I-04, Cab I-05 and Cab I-06 on Middle Beach. The captions in Figures 1-6 also give the California planar coordinates of the bench mark for each range line. 12 11 10 9 8 7 6 5 4 ;: 3 ::l ::; 2 ~ 1 C 0 0 -1 .::l l1l > -2 Q) [j -3 -4 -5 -6 -7 -8 -9 -10 -11 100 200 South Beach Profiles, Cab 1-01 ----17 April 2015 ----22 December 2014 ---6 November2014 ---5 September 2013 300 400 500 600 Distance offshore from benchmark (ft.) Coordinates of Bench Mark: Northing (ft): 1994408.5 Easting (ft): 6228847.4 700 800 Figure 1: Measured beach profiles at survey range Cab I-01 on South Beach, before and after the most recent Agua Hedionda Lagoon dredging, which was begun on 31 December 2014 and completed on 15 April 2015. See Figure 7 for bench mark locations. 14 13 12 11 10 9 8 7 6 -5 ~ ~ 4 ~ 3 ~ 2 C 1 0 0 :.:: <ti -1 > Q) w -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 0 100 Coordinates of Bench Mark: Northing (ft): 1995102.9 Easting (ft): 6228562.5 South Beach Profiles, Cab 1-02 ----17April2015 ----22 December 2014 ----6 November 2014 ----5 September 2013 200 300 400 500 Distance offshore from benchmark (ft.) 600 700 Figure 2: Measured beach profiles at survey range Cab 1-02 on South Beach, before and after the most recent Agua Hedionda Lagoon dredging, which was begun on 31 December 2014 and completed on 15 April 2015. See Figure 7 for bench mark locations. 17 16 15 14 13 12 11 10 9 8 ~ 7 ..I 6 ..I 5 ~ 4 .:a:! -3 C 2 8 1 n, > 0 Q) [jj -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 0 100 Coordinates of Bench Mark: Northing (ft): 1995576.9 Easting (ft): 6228365 South Beach Profiles, Cab 1-03 ----17 April 2015 ----22 December 2014 ----6 November 2014 ----5September2013 200 300 400 500 Distance offshore from benchmark (ft.) 600 700 Figure 3: Measured beach profiles at survey range Cab 1-03 on South Beach, before and after the most recent Agua Hedionda Lagoon dredging, which was begun on 31 December 2014 and completed on 15 April 2015. See Figure 7 for bench mark locations. 17 16 15 14 13 12 11 10 9 8 ~ 7 ...J 6 ...J 5 ~ 4 ~ 3 C: 2 0 .::. 1 ro > 0 Q) w -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 0 100 Coordinates of Bench Mark: Northing (ft): 1996164.9 Easting (ft): 6228090.5 Middle Beach Profiles, Cab 1-04 ----17April2015 ----22 December 2014 ----6 November 2014 ----5 September 2013 200 300 400 500 Distance offshore from benchmark (ft.) 600 700 Figure 4: Measured beach profiles at survey range Cab 1-04 on Middle Beach, before and after the most recent Agua Hedionda Lagoon dredging, which was begun on 31 December 2014 and completed on 15 April 2015. See Figure 7 for bench mark locations. 17 16 15 14 13 12 11 10 9 8 ~ 7 ...I 6 ...I 5 ~ 4 E, 3 C: 2 0 ~ 1 (1) > 0 Q) ui -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 0 100 Coordinates of Bench Mark: Northing (ft): 1996778.5 Easting (ft): 6227826.9 Middle Beach Profiles, Cab 1-05 ----17 April 2015 ----22 December 2014 ----6 November 2014 ----5 September 2013 200 300 400 500 Distance offshore from benchmark (ft.) 600 700 Figure 5: Measured beach profiles at survey range Cab 1-05 on Middle Beach, before and after the most recent Agua Hedionda Lagoon dredging, which was begun on 31 December 2014 and completed on 15 April 2015. See Figure 7 for bench mark locations. 17 16 15 14 13 12 11 10 9 8 -7 ~ 6 ...J ...J 5 :'E 4 ~ 3 C: 2 0 :; 1 <ll > 0 Q) [j -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 0 100 Coordinates of Bench Mark: Northing (ft): 1997015.5 Easting (ft): 6227720.2 Middle Beach Profiles, Cab 1-05 ----17 April 2015 ----22 December 2014 ----6 November 2014 ----5 September 2013 200 300 400 500 Distance offshore from benchmark (ft.) 600 700 Figure 6: Measured beach profiles at survey range Cab 1-06 on Middle Beach, before and after the most recent Agua Hedionda Lagoon dredging, which was begun on 31 December 2014 and completed on 15 April 2015. See Figure 7 for bench mark locations. Each range line was surveyed four times, three times prior to the most recent Agua Hedionda beach disposal event, (5 September 2013, 6 November 2014 and 22 December 2014); and once immediately upon completion of beach disposal on 17 April 2015. The dredge logs indicate that no dredging or beach disposal was begun until 31 December 2014. Since the 22 December 2014 beach surveys were performed just 9 days prior to onset of lagoon dredging, these surveys captured the receiver beaches close their most denuded condition. In fact there had been no placement of new sands on these beaches since April 2011; and inspection of the profile measurements in Figures 1-6 indicates little profile change between the 6 November 2014 and 22 December 2014 surveys because much of the beach surface was exposed hard-bottom substrate (bedrock, cobbles and basal conglomerate, see Figure 9). Middle Beach was the first to receive beach fill from the 2014/15 dredge cycle beginning on 31 December 2014, and received a total of 156,056 cubic yards; followed by South Beach that began receiving 73,637 cubic yards of beach fill beginning on 21 February 2015. Unlike previous dredge events, North Beach was the last to receive beach fill, and 64,968 cubic yards were placed there between 23 March 2015 and 15 April 2015, (cf. Appendix-A). Therefore, the 17 April 2015 beach profile surveys represent the receiver beaches in their most built-out state, because these surveys were performed only two days after the beach disposal of dredged sands from Agua Hedionda was completed. Three-dimensional CAD models were lofted from the measured points along the 1 o beach survey range lines (Cab 1-01 -Cab 1-10) to delineate the beach surfaces immediately before dredge disposal (based on the 22 December 2014 profile measurements) and immediately after dredge disposal (based on the 17 April 2017 profile measurements). When these two surfaces were lofted together in a common reference frame (as determined by the bench marks of each survey range line), places where the 17 April 2015 surface intersected the 22 December 2014 surface identified areas along the receiver beach where sand was not retained. Conversely, places where these two beach surfaces remained separated identified areas where dredged sands were being retained to at least some degree. This is shown in Figure 7 where the fully built-out beach surface (post-dredge disposal, 17 April 2015) is lofted in brown, and the severely eroded beach surface (pre dredge disposal, 22 December 2014) is lofted in silver. It is clear from the large areas of silver in this composite CAD model, that beach fill has been poorly retained on North Beach and the northern end of Middle Beach despite the fact that this was the last beach to receive beach fill. The eroded silver areas between Cab 1-06 and Cab 1-10 in Figure 7 are especially prominent in the seaward half of the bar-berm section of the beach profile, but also in the upper portions of the profile on the berm itself. The mass properties tool of the SolidWorks 3-d CAD software was invoked to calculate the rlifference in beach volmne between the surfaces defined by the 22 December 2014 and the 17 April 2015 surveys, and determined that 13,780 cubic yards of beach fill have been retained after placing 64,968 cubic yards on North Beach between 23 March 2015 and 15 April 2015. This means that only 21 % of the beach fill had been retained over a 3 ½ week period that ended just two days after completion of pumping beach fill to North Beach! Clearly sand was being lost at a high rate (1,969 yds3/day, or about 1/3 to 1/2 daily pumping rates) during placement of sand in the beach fill template. A further concern, is that the beach fill placed on North Beach is up- drift of the lagoon inlet in the prevailing southerly littoral drift, particularly during the winter season when the large waves that erode the beach are from the northwest. Consequently, when beach fill is not retained it is immediately ingested by Agua Hedionda Lagoon, where it provides no useful function in protecting the shoreline against erosion and wave overtopping, and where it restricts the tidal prism of the lagoon and degrades the lagoon water quality. On the other hand sand retention was much better on Middle Beach and South Beach. After placing 229,693 cubic yards on these beaches between 1 January 2015 and 23 March 2015, the SolidWorks mass properties tool calculates that 178,584 cubic yards of beach fill have been retained at the time of the post-dredge survey on 17 April 2015. Therefore, the sand loss during this time was only 51,109 cubic yards, or a loss rate of only 473 yds3/day, about I/4th the rate of loss of sand on North Beach. Cab 1-01, ... -----..------11-,0 2000 1000 Northing, ft D Figure 7: Three-dimensional composite CAD model of two overlaid beach surfaces on North, Middle and South Beach, 1) immediately before dredge disposal (as delineated in silver from the 22 December 2014 profile measurements) and 2) immediately after dredge disposal (as delineated in brown by the 17 April 2015 profile measurements). CAD model shown with 10 to 1 vertical exaggeration. 2) Sand Retention Issues: The poor sand retention characteristics of the receiver beach of back-passing are due to several factors including: the timing of placement of back-passing sands, non- equilibrium distribution of those sands, unfavorable geomorphology, and placing more sand than the geomorphology can support in equilibrium. Beginning with timing, the largest fraction of sand that is lost from the beach fill placed on the receiver beaches occurs in the first few months after placement. Because the least tern nesting season restricts Agua Hedionda Lagoon dredging to the winter season, placement of the beach fill typically occurs in the midst of the onslaught of the largest winter waves. However, during the period when sands from the 2014/15 dredge event were being placed on the receiver beaches, (1 January 2015 -15 April 2015) the winter waves were not unusually intense, with the highest waves reach Hmax = 1.94 m while average significant wave heights were <H> = 1.1 m, with an average wave period of <T> = 13.8 sec and a typical northwesterly wave direction averaging< <X > = 279°, (cf Figure 8b). But even these rather ordinary winter waves can exert significant erosional effects due to the way waves refract and shoal at along the receiver beaches, and particularly at North Beach. The presence of the Carlsbad Submarine Canyon immediately south of Agua Hedionda Lagoon creates a bright spot in the shoaling wave pattern immediately north of the inlet jetties, where wave heights are locally higher than further to the north around Maple Ave. This alongshore imbalance in wave energy leads to a negative divergence of dr(ft in the longshore transport rates, which in turn causes higher southerly longshore transport rates of sand exiting North Beach at the inlet jetties than enters North Beach at Maple Ave, (see Section 5 for more detail). These same refraction effects from the bathymetry along the north rim of the Carlsbad Canyon also propagate along Middle and South Beach, although their effect on divergence of drift is weaker due to the obstruction provided by the inlet and discharge jetties, (the so-called groin.field effect). So we turn our attention to how the dredged beach fill has been placed and what quantities. Engineered beach fill has typically been placed on Carlsbad beaches with a 1: 10 (rise over run) slope. This convention dates back to the Regional Beach Sand Project, (AMEC, 2002). However, stable beach profiles in Nature have a much more gradual, curving profile with slopes that range between 1 :50 or 3: 100, (Inman et al., 1993). The theory on equilibrium beach profiles began with Dean (1977) and Bowen (1980) who developed formulations for an equilibrium profile having the form, h=Ax'", where h is the local water depth, x is the horizontal distance offshore, ,n = 2/3, and A is an empirical factor. Later Dean ( 1991) developed analytic approximations for the empirical factor, A, and that formulation was incorporated into the U.S. Army Corps of Engineers Shore Protection Manual and later the Coastal Engineering Manual. However, recently Jenkins and Inman, (2006), proved that the Dean (1977 and 1991) solutions are not unique, and represent only one of a family of equilibrium beach profiles known as an elliptic cycloids. The elliptic cycloid formulation can account for continuous variations in the equilibrium beach profile due to variability in wave height, period and direction when occurring in combination with variations in beach sediment grain size and beach sand volume. Equilibrium beach profiles obey the maximum entropy production formulation of the second law of thermodynamics, and are the most efficient shape for a beach profile because it adjusts itself to dissipate all of the available wave energy. When the waves encounter an inefficient non-equilibrium beach shape, such a steeply sloping beach fill 4 3 E E C> ·a; :i:2 Q) > t'O ~1 25 20 ~ (/) '815 ·c: if. G> 10 > ~ 5 CDIP Station #043-Camp Pendelton Nearshore, 33.2198 N, -117.4394 W 2.3-year mean= 0.97 m ~ Hm•x = 3.8 m <H> =0.95m \ 2.3-year mean = 13.8 sec <T> = 13.8 sec ~ 0 __.___.___.___.___.___.___.___.___.___.___.___.___.___.__.__.__.__.__.__._ ............................ __.____.__...__....__ c: 280 0 .:= (.) ~ iS 240 ! ~ 200 2.3-year mean = 276° 1/1/13 1/1/14 1/1/15 Figure 8: Shoaled significant wave heights, periods and directions at Carlsbad State Beach based on back refraction of wave monitoring data from CDIP Station 043 at Camp Pendleton during the beach survey period. Data shown in black occurred during the 2014/15 Agua Hedionda Lagoon dredging event. templates, then the wave energy is not fully dissipated and the excess wave energy begins eroding and moving that beach fill around until an equilibrium profile is finally achieved. While this is occurring, the beach fill can be highly mobile, particularly in large winter waves; and at Carlsbad, the net southward flowing longshore currents (particularly in winter) will rapidly transport the eroded beach fill from North Beach toward the south and the inlet to Agua Hedionda Lagoon; although some of the mobile North Beach sands are transported onto Middle Beach along the by-passing bar at the lagoon inlet during ebbing tide. There are also geomorphic factors that contribute to poor retention of Agua Hedionda dredge sands at the North Beach disposal site, while limiting retention times at Middle and South Beach. A sandy beach cannot be supported in equilibrium against wave forces without a wave-cut platform in the bed rock to provide a foundation. Wave cut platforms are notches that have been eroded in the bed rock during protracted still-stands in sea level. Once formed, sediment collects in these notches forming a beach which is subsequently molded into an equilibrium shape by wave action. Figure 9a shows an annotated seismic reflection profile measured by USGS across the continental shelf off Carlsbad CA on range 223X of the 1991 Kolpack surveys, (Kolpack, 1991). It shows a series of wave-cut platforms that were formed at present and earlier still-stands of sea level, and subsequently covered with sediment. The most striking feature in Figure 9a is how much more pronounced the paleo wave-cut platforms are than the modern platform; and how very thin the sediment cover is over the modem wave-cut platform, as compared with the thickness of Holocene sediment over the paleo platforms. Although the paleo platforms have been subjected to longer periods of sedimentation, the geometric constraints imposed by small wave-cut platform prevent them from retaining thick layers of sediment cover. Many of the beaches throughout north San Diego County are perched on narrow wave-cut platforms. The platforms are narrow because they were carved by wave action from erosion resistant Del Mar formation during the present high-stand in sea level, and these narrow wave-cut platforms physically cannot hold large quantities of beach sand; and often become fully denuded during periods of high-energy winter waves, as shown in Figure 9b. Sub-bottom surveys by Elwany, et al., (1999) discovered narrow wave cut platforms and exposed hard bottom substrate in the surfzone and nearshore at North Beach and South Beach (Figure 10) while 1884 railroad surveys reveal beach cobble ridges before the influence of Mankind at Agua Hedionda, ( e.g. HWY 101 and the deep water lagoon). When beach cobble ridges and hard-bottom features are found this close to shore, it indicates that these beaches (particularly, North Beach and South Beach) are not geomorphically well suited to retain large volumes of sand. In the particular case of North Beach, attempts to back-pass and place more sand there than its carrying capacity will simply result in low retention time and increased sand influx of into Agua Hedionda Lagoon. The remainder of this report focuses on determination of the carrying capacity at North Beach and the optimal distribution of that carrying capacity within a beach fill template in order to maximize sand retention time of dredged sands. 3) Critical Mass and Middle/South Beach Equilibrium Profiles: The criticai mass is the minimum voiume of sediment cover required to maintain equilibrium beach profiles and represents the nominal carrying capacity of a particular beach. When a long term collection of beach profiles are plotted together over a broad range of wave heights, a well-defined envelope of variability becomes apparent, (Figure 1 la). This envelope of profile variability is referred to as the critical mass envelope, and the volume of sand within critical mass envelope, Ve , increases with increasing wave height and period but decreases with increasing beach grain size, as shown in Figure 11 b. The critical mass envelope is always limited by the breadth of the wave cut platform, a) MSL 0 present 20 wave-cut 40 shelf platform sediment 60 paleo country rock 80 wave-cut platform vertical exaggeration 23x 7 6 5 4 3 2 1 0 Distance, km Figure 9: Wave-cut platforms in north San Diego County: a) Annotation ofUSGS Geopulse sub-bottom seismic profile along range line 223-X in the inner shelf off Carlsbad showing present and ancient wave cut platforms (after Kolpack, 1991); b) exposed wave cut platform in Solana Beach during the 1983 El-Nino winter. E ..c: -a. Q) 0 362000--- 361000- 360000 358000-- 357000 356000 355000 354000-- 353000-- 3520oo---; Pacific Ocean D 10-30% hard substrate D 30-60"-'> hard substrate D 60-100% hard substrate 1660000 1661000 1662000 ' 1663000 I 1664000 . -,., ·-----...... ... C A I. 1665000 1666000 1667000 l ' _..-.-·· ··-· --· ---·---::------~,~ ---; Figure 10: Nearshore survey (left) showing exposed rocky reefs outcrops and other hard bottom substrate; and (right) 1884 railroad survey map showing beach cobble ridge, both indicating minimal sediment cover on the beaches around Agua Hedionda Lagoon. a) -----------MSL --- __ .,.. ... - V.E. 33x E 2500 --.., E_ 2000 ":;:.O l 1500 ~ ~ ~ 1000 (.) 0 Q) E :::, 500 1600 1200 800 Distance, m b} --D, = 12011m, 0, = 80 µm (Rockport, TX) --01 = 200 µm, 0, = 100 µm (Torrey Pines, CA) --D1 = 200 11m, 0, = 200 11m (Scripps Beacll, CA) --o, = 400 µm, 0, = 150 µm (Duck, NC) 400 0 ~ 0 i,c;..._.,___ ........ _..J...._.....,__---1. _ __,L....-_,L..__...L..._...J...._.....J 3 E " UJ' f 2 C: ti :2 I- m rn Cll 1 ~ ~ ~ (.) 0 c) 1600 2 3 4 5 rms Incident Wave Height, H,., , m --D1 = 120 llffl, 0, = 80 µm (Rockport, TX) --D1=20011m, 0, = 100 µm (TOIT8y Pines, CA) --o, = 200 µm, D, = 200 µm (Salpps Beach, CA) --D1 = 400 µm, D, = 150 µm (Duck, NC) 1200 800 400 0 Distance, m 10 5 0 E £ 5 5r 0 10 15 Figure 11: Features of the critical mass of sand: a) critical mass envelope for waves ranging from 1 m to 5m in height; b) volume of critical mass as a function of wave height and sediment grain size; c) variation in the thickness of the critical mass as a function of distance offshore. which forms a hard-bottom boundary condition on the critical mass envelope. The best way to calculate the critical mass is to find the volume between the wave cut platform ( or its layer of basal conglomerate) and the elliptic cycloid equilibrium profile that corresponds to the native beach grain size in combination with the wave height and period of the extreme event wave in the period ofrecord. The volume integral between the surfaces of the wave cut platform and the extremal event elliptic cycloid then give the critical mass volume. At Middle and South Beach, the sub-bottom reflection data is too spotty to resolve the complete surface of the wave-cut platform along the 3,680 ft. length of the Middle/South Beach disposal site (between south inlet jetty and the north end of the Encina generating building). Therefore we will use the surface given by the 22 December 2014 beach surveys as a surrogate bottom of the critical mass envelope. This is a reasonable approximation because there has been no placement of new sands on Middle and South Beach since the 2010/11 dredge event, which only placed 163,996 cubic yards on Middle and South Beach in April 2011 (cf. Appendix-A). Inspection of the profile measurements in Figures 1-6 indicates little profile change between the 6 November 2014 and 22 December 2014 surveys because much of the beach surface had partially exposed hard-bottom substrate (bedrock, cobbles and basal conglomerate). The extremal elliptic cycloid equilibrium profile is a curve that is traced by a point on the circumference of a rolling ellipse, see Figure 12b. It is calculated from Jenkins and Inman (2006) using the following: h = 7r& X ( 1-COS 0) + Z I 2J?l 0-sin 0 (1) Note this has the same basic formulation of the original Dean (1977 and 1991) solutions in the U.S. Army Corps of Engineers Coastal Engineering Manual Here Z1 is the elevation of the berm crest (cf. Figure 12a) given by Hunt's Formula [Hunt, 1959; Guza and Thornton, 1985; Raubenheimer and Guza, 1996]: (2) In equation (2), r is the runup factor taken herein as r = 0. 76, and H b is the breaking wave height. The cycloid in (1) is based on the elliptic integral of the second kind that has an analytic approximation, I !2l = ~ (2 -e2 ) / 2 , where e is the eccentricity of the ellipse given by e = ~ 1-b2 I a 2 , with, semi-major and semi-minor axes are a, b, (cf. Figure 12b). The wave parameter,€, in equation (1) is given by: ( ) l/2 4/S ( J2/5 Hb O' H00 S=O' -"" ----y g -2"s gy (3) here a-= 2n/period is radian frequency, H 00 is incident wave height, g is the acceleration of gravity, and yis the wave breaking criteria taken as y= 0.8. The rolling angle of the ellipse is: [ ( 1-l \a] 0 = arccos 1 -2l ~ j _A y hc_ (4) a. -5 0 E ~ 5 £ ~ 10 0 15 20 Et :::1: Shorerlse Xc2 Xe b. type-a cycloid X C. \ \ \ 8=7t t ' ' ' .... a) profiles: eccentricity and shear stress linearity ----e= 0.845: n= 3 ----e=0.798;n=2 ----e=0.707:n=1 ----e=0.447;n=O 't ----e = 0 ; n = -0.33 ; brachistochrone solution 700 600 500 400 300 200 Cross-Shore Distance x2 ,m Bar-Berm -1 X1 ~A X1 h 0 2 4 ..I "' 6 ~ E - 8 ~""- 10 :g_ Cl) 0 12 14 100 0 Figure 12. Equilibrium beach profile theory: a) nomenclature, b) mathematical basis for an elliptic cycloid, c) Typical range of elliptic cycloids on a 700 m wide wave-cut platform. where A is the shoaling factor relating breaker height to incident wave height, A= Hool Hb, ( .2 / ) 115 which for shoaling Airy waves, becomes A= 2215 H~5 a/ g y . The closure depth, he in equation (4) is grain size and wave period dependent and is given by: (5) where k = a I~ ghc is the shallow water form of the wave number, Ke and 1// ~ 2.0 are non- dimensional empirical parameters, set at Ke= 2.0 and 1// ~ 0.33; D50 is the median grain size; and Do is a reference grain size taken as D0 = 100 µ m. Equation (5) is transcendental and is solved numerically within the CEM. Calculation of the extremal elliptic cycloid equilibrium profile at Middle and South Beach requires knowledge of the characteristic median grain size, D50 , of the dredged sediments to be placed there. Recent sediment grain size analyses by Merkel, (2008) based on three sampling locations on the flood tide bar in the West Basin of Agua Hedionda Lagoon (Samples L1 -L3) were compared against native sediments on the three receiver beaches (RB 1-RB3). These grain size distribution are plotted in Figure 13. Note Middle and South Beach is represented by samples RB 1. Grain sizes at the lagoon sample sites and beach sites were similar with median grain sizes of 0.32 millimeters (320 microns) on the food tide bar in the West Basin of Agua Hedionda, while residual sediments that still remained on Middle and South Beach prior to disposal of material from the 2008/09 dredging averaged 0.374 millimeters (374 microns). To determine the highest waves to reach to effect Middle and South Beach disposal, the waves measured at ½ hour sampling intervals at CDIP Station 043 were back refracted into deep water from the monitoring location off Camp Pendleton, and then forward refracted into Middle and South Beach. An example of this procedure is shown in Figure 14 for a wave occurring 8 January 2002. This effort produced a continuous wave record throughout the historic period when Middle and South Beach disposal of Agua Hedionda dredged sands was practiced, (1998- 2015), see Figure 15 .. The highest energy wave (extremal) event occurred in January 2007, when a Gulf of Alaska storm brought 4.8 m high waves approaching Carlsbad at 276 ° with a 15 second periods. This extreme event wave was used to calculate the extremal elliptic cycloids on Middle and South Beach. To calculate the critical mass of Middle and South Beach, we combine the extremal waves with theD50 grain size values from Figure 11 to solve equations (1)-(5) for the extremal elliptic cycloid profile. These extremal cycloid profiles are plotted on the Middle and South Beach Range survey range lines (Cab-1-07 -Cab 1-10) in Figures 16-21. These profiles represent the beach shape that can be sustained in an equilibrium state during the most severe wave events of the 1998-2015 Middle and South Beach disposal period. These profiles form the top of the critical mass envelope, while the most eroded profile (from the 22 December 2014 c,11~ui::n1c., \ f-n h~uP Ar-r-11...,..~rl 1n that ciamP t'\Pt-1Arl Af rPf'nrrl rlPtin,=.c thP hnttnm nf r-ritir~ 1 m!:l~~ IJWJ. YVJiJJ \.V J.J.""""' VVVUl..&.VY J.J.J. l,J.J.U-1, ..,1,,.4,J.J.J.V yvJ.J.Vllo.A v.a. ........ V\J.1.-...................... ..., ............. ...,.....,. ................ '-'& _ _...._ .. ._ __ ......... -.... .... envelope. When lofted in the 3-d CAD SolidWorks software, the SolidWorks volume tool calculates the critical mass envelope to hold of 280,361 cubic yards along the entire 3,680 ft. of • Dredge Area L 1 • Dredge Area L2 • Dredge Area L3 • • C:.:• • Reciever Beach RB 1 • • Reciever Beach RB2 • Reciever Beach RB3 • •• • • • •• • • • • • •• • • • • •• • • • 10 1 Grain Size (mm) 0.1 100 90 80 & 70 ! ~ CD 60 ~ .c en i 50.;.; .c en 'i 40 i i i 30 'S E 20 10 0 0.01 ::::, u Figure 13: Grain size distributions form Agua Hedionda Lagoon (Samples L1 -L3) and from the receiver beaches (RB1-RB3). Note Middle and South Beach is represented by samples RBI, (from Merkel, 2008). 33.200 33.175 33.150 33.125 33.100 33.075 117.450 117.425 117.400 0 117.375 Longitude 2 3 B.Jena \Asta Lagoon 117.350 117.325 4 5 Incident Wave Height at CDIP-043 = 1 .66 m, Period = 18 sec, Direction = 263 deg 117.300 Figure 14: Regional wave shoaling during 8 January 2002 from back-refraction of wave monitoring data at CDIP Station# 043 in 20 m local water depth offshore of Cam Pendleton. Carlsbad Waves Derived from CDIP Station #043-Camp Pendelton, 33.2198° N, -117.4394° W 5 4 25 20 ~ °8 15 ·c: Q) D.. Q) 10 > ~ 5 360 340 Q> 320 ~ 300 c 280 0 ts 260 ~ o 240 ~ 220 (1' 3: 200 180 1998 Hmax = 4.8 m ---..,.,.:l- long term mean = 0.95 m 2000 2002 2004 2006 2008 2010 1998 2000 2002 2004 2006 2008 2010 long-term mean = 272° 2012 2014 2016 2012 2014 2016 160 --....---.---.----.---.----r-..-----r-..----.--r--.---,..---.---,--,--~--.---, 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Figure 15: Shoaled significant wave heights, periods and directions at Carlsbad State Beach based on back refraction of wave monitoring data from CDIP Station 043 at Camp Pendleton for the period of record of Middle and South Beach disposal, 1998-2016. 12 11 10 9 8 7 6 5 4 ~ 3 ...J ...J 2 :::ii s, 1 C: 0 0 -1 .:; «l > -2 (I) w -3 -4 -5 -6 -7 -8 -9 -10 -11 100 200 South Beach Profiles. Cab-01 ----22 December 2014 ----Extremal Cycloid (H"'"' = 4.8 m) Critical Mass, Ve 300 400 500 600 Distance offshore from benchmark (ft.) 700 800 Figure 16: Critical mass envelope at range line Cab 1-01 on South Beach based on the extremal elliptic cycloid solution using a 4.8 m high design wave height with 15 second wave period. 12 11 10 9 8 7 6 5 4 :i: 3 ..J ..J 2 :;: ~ 1 C: 0 0 -1 :,= ro > -2 Q) w -3 -4 -5 -6 -7 -8 -9 -10 -11 0 100 South Beach Profiles, Cab-02 ----22 December 2014 ----Extremal Cycloid (H,..,., = 4.B m) Critical Mass, V c 200 300 400 500 Distance offshore from benchmark (ft.) 600 700 Figure 17: Critical mass envelope at range line Cab 1-02 on South Beach based on the extremal elliptic cycloid solution using a 4.8 m high design wave height with 15 second wave period. 17 16 15 14 13 12 11 10 9 8 s:: 7 ..J 6 ..J 5 ~ 4 = -3 C: 2 0 ~ 1 ~ 0 Q) [i -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 0 100 South Beach Profiles, Cab-03 ----22 December 2014 ----Extremal Cycloid (Hm., = 4.8 m) Critical Mass, V c 200 300 400 500 Distance offshore from benchmark (ft.) 600 700 Figure i8: Critical mass envelope at range line Cab I-03 on South Beach based on the extremal elliptic cycloid solution using a 4.8 m high design wave height with 15 second wave period. 17 16 Middle Beach Profiles, Cab-04 15 22 December 2014 14 Extremal Cycloid (H.,,,, = 4.8 m) 13 12 11 10 9 8 -7 ~ 6 ..J ..J 5 ::::iE = 4 Critical Mass, Ve -3 C: 2 0 ~ 1 n, > 0 G.l jjj -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 0 100 200 300 400 500 600 700 Distance offshore from benchmark (ft.) Figure 19: Critical mass envelope at range line Cab 1-04 on Middle Beach based on the extremal elliptic cycloid solution using a 4.8 m high design wave height with 15 second wave period. 17 16 15 14 13 12 11 10 9 8 ~ 7 6 ::I 5 ~ 4 ~ 3 C: 2 0 .:; 1 co > 0 Q) iii -1 -2 .3 -4 -5 -6 .7 -8 -9 -10 -11 0 100 Middle Beach Profiles, Cab-05 ----22 December 2014 ----Extremal Cycloid (H"'"' = 4.8 m) Critical Mass, Ve 200 300 400 500 Distance offshore from benchmark (ft.) 600 700 Figure 20: Critical mass envelope at range line Cab 1-05 on Middle Beach based on the extremal elliptic cycloid solution using a 4.8 m high design wave height with 15 second wave period. 17 16 15 14 13 12 11 10 9 8 ~ 7 ...J 6 ...J 5 ~ 4 ~ 3 C: 2 0 .:. 1 n, > 0 Q) w -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 0 100 Middle Beach Profiles, Cab-06 ----22 December 2014 ----Extremal Cycloid (H.,., = 4.8 m) Critical Mass, Ve 200 300 400 500 Distance offshore from benchmark (ft.) 600 700 Figure 21: Critical mass envelope at range line Cab I-06 on Middle Beach based on the extremal elliptic cycloid solution using a 4.8 m high design wave height with 15 second wave period. the Middle and South Beach disposal site. This volume represents the optimal carrying capacity of the Middle and South Beach disposal site. Lesser amounts of beach fill will not be able to sustain an equilibrium profile during the highest energy wave events; and without an equilibrium profile, the beach will not dissipate all the incident wave energy, and the excess wave energy will erode the beach. If Middle and South Beach is over-nourished with more than the critical mass of sand, then two processes will intervene: a) the excess sand will spill off the wave cut platform (which is 600 ft. to 650 ft. wide) and be re-suspended over the rocky outcrops and hard bottom substrate at Middle and South Beach (cf. Figure 10); and/orb) the excess sand will be swept away by the net longshore transport (littoral drift) which flows from north to south throughout the Oceanside Littoral Cell ( cf Figure 22). 111roo' ,,,,.-----, ' ' ,, \ '-, i \ km I I \ -\ -· .. ' ' \ \ I ,,-...) I ------USA ________ _ iio1co 33•30' • WOO' 32'45' --- Figure 22: Oceanside Littoral Cell showing net north-to-south littoral drift, (from Inman and Brush, 1970) 4) Cycloid Beach-Fill Template Design: Ideally the optimal beach fill templates for Middle and South Beach would duplicate the critical mass envelope in Figures 16-21, as these would prescribe an adequate amount of sand to support an equilibrium profile in the presence of extreme event waves without exceeding the carrying capacity of the otherwise limited wave cut platform that exists there. However, Figures 16-21 indicate that the critical mass envelope extends well below mean lower low water (MLL W) to depths ranging from -4 ft. MLL W to -8 ft. MLL W. With present beach fill construction methods, it is not possible to build a template that extends below the waterline. Beach fill is pumped to Middle and South Beach via a hydraulic dredge pipeline and initially deposited as a slurry. After the slurry dewaters, the sands are spread out across the beach using conventional earth-moving equipment, which cannot effectively operate in anything deeper than ankle deep water. Therefore, we must pose a beach-fill template that adapts to this construction constraint. We begin by examining the percentage of time that a dry beach is available for construction operations at the lower end of the beach profile during in the months of September to mid-April, (the months during which dredge disposal is permitted in order to avoid impacting the least tern nesting season). Figure 23 plots the relationship between ocean water level and percent time a given elevation remains dry, (referred to as the hydroperiodfunction), based on ocean water levels measured at the nearby Scripps Pier tide gage (NOAA# 9410230). Figure 23 indicates that the beach fill construction operations can proceed down to elevations as low as 0 ft. MLLW at least 7% of the time, or during about 50 hours in a given month. These times are clustered during the spring tides that occur twice each month. If the beach fill template is filled from the top down (ie, spreading sand at the highest elevations of the template first, and then proceeding downslope towards the lowest end), then 50 hours should be adequate to allow filling the lowest portion of a template that terminates at 0 ft. MLL W. A significant fraction of the critical mass envelope in Figures 16-21 lies below the 0 ft. MLL W water level, and if the extremal cycloid profile is terminated at that elevation in the beach fill template design, then additional sand must be added elsewhere to the template in order to achieve the critical mass volume along the entire 3,680 ft. reach of the Middle and South Beach disposal site. The additional sand is provided by combining a back-beach dune with the elliptic cycloid that has been truncated at Oft. MLLW. The back-beach dune placement strategy was first implemented in Carpinteria by Bailard and Jenkins (1980 and 1983) and later during the replacement of seawalls at Mission Beach Sea (Jenkins 2014) and Del Mar (cf Figure 24a.). These previous implementations of the back-beach dune strategy involved very popular beach sites, yet no adverse recreational incidences were encountered. The back beach dune is a conservation/storage mechanism that prevents rapid sand loss from over-builds of the intertidal portion of the beach profile, yet still allows the fill site to receive its full allocation of critical mass, and provides gradual re-nourishment as the dune erodes during brief periods of spring tides and/or high waves. The dune proposed for the Middle and South Beach disposal site is shown in Figure 24b, and is roughly 9 ft high and 55 ft. wide, with a reserve storage capacity of 13.5 cubic yards per running ft. of beach. When built along the 3,680 ft. length of the Middle and South Beach disposal site, this dune will provide 43,200 yds3 sand perched in the upper portion of the truncated equilibrium cycloid profile. The beach fill templates that result from this strategy are shown in Figures 25-30 for Middle and South survey range lines Cab I-01 -Cab I-06. When lofted over the entire 3,680 ft of the Middle and South Beach disposal site using the SolidWorks 3-d CAD software, ( cf. Figure 31 ), we calculate that these templates will provide 200,593 cubic yards of disposal volume, which compares almost exactly with the required critical mass of 200,890 cubic yards, (the optimal carrying capacity of the Middle and South Beach disposal sites.) ~ ....J ....J ~ ~ C: .Q -ro > ~ w ~ Q) ....J .... 2 ~ C: ro Q) () 0 8 7 6 5 4 3 2 1 0 -1 -2 -3 0 La Jolla, Scripps Pier Tide Gage NOAA # 9410230 EHW = +7.47 ft. MHHWz: +5.13 ft. • +2.11 ft. I Beach Dry @ ' MLW 17% of the time ---------r -------------- l Beach Dry@ iMLLW 7% of the time --r ---------r -------------- 1 1 % Lowest Water Level r : ~ =-1.2ft.MLLvV ~, I ~I if..1 1_1 1 ~I ~ I ~ I I EL:W = -3.06 ft MLLW 10 20 30 40 50 60 70 Percent Time Dry 80 90 100 Figure 23: Hydroperiod function of ocean water levels during the months of September-April, based on the Scripps Pier tide gage (NOAA # 9410230; based on the 1983-2001 tidal epoch, (from Jenkins and Taylor 2015; and Jenkins and Wasyl, 2011) a) back-beach dune b) 10 -----.----.-----.----.--,--,.--,--,---,---,--,---.--....-~-.-~~~~~~~~~---.----.-----.----.----.---, 9 -------1---1-----'f--l--l-+--L -~ 7 -----l-----l-----1-----1-----1-----I-----I-- E 5 -----1-----1-----1-----1-----1-----1-----1-~· C> "ii> 5 --+--+--+--+---+----1---lf---ll: I ~ 4 -------1---1----1-c ::J 3 --+--+--+--+--+----1--.1 0 1 --t---+--+-~ 0 ~""'*=....:::...,..;.;::.,;"""-"'i=:;:;~;....-...=,,......~::,;:;;..:;;;._;;:.;..~;;:;,=~ ...... ._.;.:."'+"""""''-'-w--+'-i=.:.::.,...--.-::a.;,,.""'-r"-9_,.----t---l 0 10 20 30 40 50 Dune Width (ft.) Figure 24: The back-beach dune beach fill placement strategy, a) as implemented in Del Mar during seawall replacement; and b) as proposed at Middle and South beach in combination with a truncated elliptic cycloid beach profile. 60 25 24 23 22 21 20 19 18 17 16 15 i14 ...J 13 ...J 12 ~ 11 ~10 C: 9 .Q 8 I ~ 5 4 3 2 1 0 -1 -2 -3 -4 -5 ------------------------------ J, l" 'J l\11 \,, ' :\ \ '1 I' \ ) \ \ ' X_-'-''llkl',-'- .... r,i..,._ South Beach Profiles, Cab 1-01 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune \ 1, -......... -.... ....... -..... I'-,. --NI:. ---......... I ---..... ....... ~ ... ' I ,1 ' I ' ' 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 Distance offshore from benchmark (ft.) Figure 25: Proposed beach fill template for South Beach range Cab 1-01, based on the extremal equilibrium profile truncated at 0 ft. MLL W with a back beach dune to hold-and-release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1994408.5 Easting (ft): 6228847.4 25 24 23 22 21 20 19 18 17 16 15 i14 ..J 13 ..J 12 ~ 11 ~10 C: 9 .2 8 ~ 7 w 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 -----------j -I -Ir --'( -~r -,, ------------- . ,, I \ I ',\ I I b\ .,~ '' ·, l \ --. -,~ \ ...... South Beach Profiles, Cab 1-02 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune ~-'I) :--., f .......... ,1· ~ ... ..... I\. I . ----._. I 1-------........ ~-_,,. ~ ~---' o ~ ~ oo oo 100 1~ 1~ 100 100 ~mm~~~ m ~ ~ ~ ~ Distance offshore from benchmark (ft.) Figure 26: Proposed beach fill template for South Beach range Cab I-02, based on the extremal equilibrium profile truncated at O ft. MLL W with a back beach dune to hold-and-release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1995102.9 Easting (ft): 6228562.5 25 24 23 22 21 20 19 18 17 16 15 i14 ..J 13 ..J 12 ~ 11 ~'10 C: 9 ,Q 8 i 7 iii 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 ------------------------------- I\ I \ \ I I j /·. \ \ ... " 'l 1 \ ' \. 'I "Ii I'.. "'Nil ' South Beach Profiles, Cab 1-03 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune .,...., .... 1 " ...... ..... .... .... -.... ....... ,....._ r-.... .· ........... " -....... ~--........ -----'"' ,.,,. 111 o m ~ ~ oo 100 1m 1~ 1~ 100 ~ ~ ~ B ~ ~ ~ ~ ~ ~ ~ Distance offshore from benchmark (ft.) Figure 27: Proposed beach fill template for South Beach range Cab 1-03, based on the extremal equilibrium profile truncated at O ft. MLL W with a back beach dune to hold-and-release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1995576.9 Easting (ft): 6228365 25 24 23 22 21 20 19 18 17 16 15 i14 ...1 13 ...1 12 ~ 11 ~10 C: 9 .Q 8 ~ 7 ~ 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 ----------= ------. ----. ---. . --- " ·, .I I ' I ]. .. -~ J: -- I 1:l .\ ·\.. J " -Oil/'l I\. I ~ ' '-\. 'll , " ... Middle Beach Profiles, Cab 1-04 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune '"-. ' ".", I\. ..... ·, ....... ... ,... .~ .... I'... ) ,......._ ' I T ..... -r-,;... 1i:1.~1:1 ~ ..... r--... I ......... ---=-, .... ..J.'-: ........... -............ '-r---,-! ' -~ I o ~ ~ w ~ 100 1~ 1~ 1w 1~ ~ m ~ ~ ~ D m ~ ~ ~ ~ Distance offshore from benchmark (ft.) Figure 28: Proposed beach fill template for Middle Beach range Cab 1-04, based on the extremal equilibrium profile truncated at 0 ft. MLL W with a back beach dune to hold-and-release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1996164.9 Easting (ft): 6228090.5 25 24 23 22 21 20 19 18 17 16 15 i14 ...1 13 :jj 12 . 11 !!:,. 10 C 9 .Q 8 ! 7 iii 6 5 4 3 2 1 0 -1 -2 -3 -4 -5 -------------------,_ ------------- -r-,..,.: ....... ' I\. " - A /J,\ ,, \ \ ' j I '-I '\ ~ -r,..., X Middle Beach Profiles, Cab 1-05 22 December 2014 (most eroded historic profile) Cycloid Beach Fill Template with Dune '-" h \ " \. " ' '-'~,,-V -.. r-.... '-I',. ...... "'oi;; --...... .... ---..... --.... ---..... -- 0 ~ ~ M 00 100 1~ 1~ 1M 100 D m ~ ~ ~ ~ m ~ ~ ~ ~ Distance offshore from benchmark (fl.) Figure 29: Proposed beach fill template for Middle Beach range Cab 1-05, based on the extremal equilibrium profile truncated at O ft. MLL W with a back beach dune to hold-and-release residual critical mass as the profile adjusts to changing vvave climate. Coordinates of Bench Mark: Northing (ft): 1996778.5 Easting (ft): 6227826.9 25 24 23 22 21 20 19 18 17 16 15 i14 ...1 13 ...I 12 ~ 11 ~10 C: 9 .Q 8 1ii 7 > 4) 6 iii 5 4 3 2 1 0 -1 -2 -3 -4 -5 -- Middle Beach Profiles, Cab 1-06 ----22 December 2014 (most eroded historic profile) ----Cycloid Beach Fill Template with Dune -A ---+--t--+--t-+-+--t--t-+-+~, ... \------------------------t----1 -----+--+--+--+-+--t---t--t-+-+---HOhl '--------------------------.... ---... -t--+--t-+-1--t--t-+-+-ttm . i ________________________ .... -----j ' ..... . . . ----------. -- ''k-+-t-it--t-+-+-t--+--t--+--+--t--t-+-+-+--+--t--+--+--t--t-+-+-1 ....... -1--1---1---1---..i-;"''<!l!," ' '1-1--t--+--t--+-+-t--t---t·-+--+--t--+-+-t--t-+-+--t--+--+-+-----1 '~~: : :li~~ \ ---+--+--+-+--1--1--'<I'''' . ----1--t--+-+-\"'" ' "'·--t--1--t-+-+--1---t-+--+-+--+--t--+-+--1---t-+-+-t--t--t--+--i --+---+--+--t--+-+-f--1'\,.._ ,I ' '.__1---t-+-+--+--+--+-+-t-11---t-+--+--+--+-+-+--t--t-+--, \,; I( • ' :•' • '\. ----+--t--+-+-+--1---1..,._'"r .. t.; '."'~,-~i I! ~:o..-------+----------,----..., ---+--t--+----+--1--1-+-t--t--l'~<-, '" '--,..... _ _,__,__,_ __ 1--t-+-+-t----+--t--+--1--t--+-----+---< " ' ' j'.'I.,. -4-+-+-+-+--1----f-+-ll--l-+--I',;:_--...., I , l ~~~ ... -+-+--+--t--+----+-t-1-+--+-+--+--t--+-+--I----II--I----, i:, "I[;· ..... ___ _,_ ___ ,_, ___ _,_ ___ ,__,_ '"""'-'-+-+-+-11-t--+-+--t---t--+-+-1-f-l--+---, ~"""'--ij~ I '""""' -4-+-+-+-+-+-+-+-11-+-+-+-+-+-+--f-+-ll-""i-..:._,,,j ~ I I' ....... ..._:-+-+--+--t--+-+-+--1--1-+---,.-+-+--i -4-+-+-+-+-+-+-+-11-+--+-+-+-+-+-+-+--1-+--+;::,;;,,ir '1_;'1 ii ......,cc....,-i__..._-.,.-+_--1-+-_+-f---__:--1---t4 -_ -+~--+:-_:-_--t;-~:-1--, -4-+-+-+-+-+-+-+-11-+--+-+-+-+-+-+-+-'l--l--+-+-P ... ~-~ . -~ ":' .. _ ........ ___ __,_-t---1 ---- o ~ ~ ~ ~ 100 1~ 1~ 1~ 1~ ~ ~ ~ ~ ~ ~ m ~ • ~ ~ Distance offshore from benchmark (ft.) Figure 30: Proposed beach fill template for Middle Beach range Cab 1-06, based on the extremal equilibrium profile truncated at O ft. MLL W with a back beach dune to hold-and-release residual critical mass as the profile adjusts to changing wave climate. Coordinates of Bench Mark: Northing (ft): 1997015.5 Easting (ft): 6227720.2 Cabl-01 Cabl-01 -----------a4,o 2000 1000 Northing. ft Figure 31: Three-dimensional CAD model of the proposed composite cycloid/dune beach fill template on North, Middle and South Beach, where the silver surfaces indicate the beach surface on which the template is built (from the 22 December 2014 profile measurements) and brown represents the beach surface resulting from the proposed beach fill template. 5) Beach and Shoreline Evolution Analysis of Middle and South Beach Disposal Options Sand movement and retention on the receiver beaches (North Beach, Middle Beach, and South Beach) is not only determined by wave-induced erosion and littoral drift, but is also influenced by the flow of sea water in and out of Agua Hedionda Lagoon. Therefore two distinctly different hydrodynamic models are required to address the study objectives listed in Section 1, namely: 1) a tidal transport model used to compute tidally induced transport of littoral sediments in and out of the inlet; and 2) a littoral transport model used to determine the equilibrium states of the receiver beaches, as well as the rates of delivery of littoral sediments to and from those receiver beaches and across the lagoon inlet and discharge channels. These models are coupled in a generalized Coastal Evolution Model (CEM) whose architecture is shown in Figure 32 with computer code detailed in Jenkins and Wasyl, 2005, which is available on-line from the Digital Library of the University of California at (http://repositories.cdlib.org/sio/techreport/58D. In the work described herein, these models are driven by wave measurements and Encina Power Station (EPS) flow rate measurements and calibrated with beach profile measurements and dredge disposal volumes preceding and following the most recent Agua Hedionda Lagoon dredging, December 2014 -April 2015. The tidal transport model used in this study is known as TIDE _FEM originally published in Inman & Jenkins, 1996. TIDE_FEM was built from some well-studied and proven computational methods and numerical architecture that have done well in predicting shallow water tidal propagation in Massachusetts Bay [Connor & Wang, 1974] and estuaries in Rhode Island, [Wang, 1975 ], and have been reviewed in basic text books [Weiyan, 1992] and symposia on the subject, e.g., Gallagher (1981). The TIDE_FEM model was peer reviewed by the Science Advisory Panel (SAP) of the California Coastal Commission, and used in three certified environmental impact reports (EIR's), including: the San Dieguito Wetland Restoration Project, (SCE, 2000, 2005); the Interstate 5 (I-5) North Coast Corridor Project for the California Department of Transportation (Jenkins and Wasyl, 2011), and the Otay River Estuary Restoration Project for Poseidon Water LLC (Jenkins, 2014; Jenkins et al., 2015). A discussion of the physics of TIDE_FEM along with its significant algorithms can be found in Appendix-B. The littoral transport model is formulated from the longshore transport relations of Komar & Inman (1970) and Longuet-Higgins (1970), and from the cross-shore transport and equilibrium beach profile relations of Bowen (1980), Inman et al. (1993) and Jenkins and Inman (2006). The littoral transport model has two computational domains, afar-field computational cell (Figure 14) and a near.field computational cell (Figure 34) that is nested inside the farfield and contains the details of the lagoon inlet and discharge channel and adjacent North Beach, Middle Beach and South Beach and the South Beach Groin. (The South Beach Groin is retained in the nearfield cell during the calibration runs, and removed to reflect future conditions for model runs testing optimal beach fill templates). The far-field cell provides mass balance for sediment entering and exiting the nearfield cell while the nearfield cell computes the interaction between the lagoon tidal and discharge transport and the littoral transport. The nearfield cell also computes the beach profile changes on North, Middle and South Beach due to sand volume changes occurring from dredge disposal, and subsequent erosion occurring in between dredging events. The Coastal Evolution Model (CEM) is a process-based numerical model. It consists of a Littoral Cell Model (LCM) and a Bedrock Cutting Model (BCM), (Figure 32), both coupled and operating in varying time and space domains determined by sea level and the coastal boundaries of the littoral cell at that particular sea level and time. Over the time scales of this study, the LCM is the relevant module. At any given sea level and time, it accounts for transport of mobile sediment along the coast by waves and currents. The BCM accounts for the cutting of bedrock once the sediment cover is denuded by wave erosion. However bedrock cutting, and notching of the bedrock to form a wave cut platform is a process that occurs over decadal to millennial time scales. ~ ~ ~ ~ "' :::E ~ "C ~ ~ a !l =-+ l g' i ~ l .. c, . fj ~ ~ ~ C ai C -g :! :e ·e Cl C ~ E a "' ... ! e j Forcing Functions i ,-----, ~ .-----------I Wet/Dry 1---~ P,.tlpltaUon Climate r---'--------l Cycles ,----;---, Watershed Erosion &Transport t -" ,;; l 1• 8oundary_Condltlons .......................................... , ! Control Cells Farfteld Grid ' ! i Accretion / N111tlng Regional Erosion Wave Bathymetry & Topography [ ,'. llt1oralTran1port ..... :J .................................... i ............................................. . ' . l Sediment , ____ ___,~----: .. ----1 Refraction/ ; Budgat .....___,'--....., Dlfftactlon i Sediment · Covar •-----~ r ....................................................................................... 1 .............. · ••••••••••••••••• -.... . • ...................................... , ••• ,1 ••••••••••••••••• 1.,, •••••••................................... GIS Accumulation of Profiles Seal.evel Change 1mag11ry or coa·stal Morphology Figure 32: Architecture of the Coastal Evolution Model (CEM); from Jenkins and Wasyl, (2005). In the LCM, the coastline of the far-field computational cell is divided into a series of coupled control cells, as illustrated schematically in Figure 33b. Each control cell is a small coastal unit of uniform geometry where a balance is obtained between shoreline change and the inputs and outputs of mass and momentum. The model sequentially integrates over the control cells in a down-drift direction so that the shoreline response of each cell is dependent on the exchanges of mass and momentum between cells, giving continuity of coastal form in the down- drift direction. While the overall computational domain of the far-field cell remains constant throughout time, the beach profile within the individual control cells can change shape or shift on or offshore, as shown in Figure 33c, in response to changes in wave height or longshore transport rates, or due to the introduction of new sediment from dredge disposal as shown in Figure 33a. These changes are computed from time-stepped solutions to the sediment continuity equation (otherwise known as the sediment budget) applied to the boundary conditions of the coupled control cell mesh round Agua Hedionda Lagoon, as in Figure 34. The sediment continuity equation is written (Jenkins, et al, 2007): dv' o ( 8v) av -=-&--U1 -+J(t)-R(t) dt ay ay ay (5) Where v' is the beach volume per unit length of shoreline (m3/m), 6 is the mass diffusivity, U, is the longshore current, J(t) is the flux of new sediment into the control cell from dredge disposal and R(t) is the flux of sediment lost from the control cell due to the tidal influx of sediment into Agua Hedionda Lagoon. The first term in ( 1) is the surf diffusion while the second is divergence of drift. For any given control cell that does not enclose the lagoon inlet, equation (5) may be discretized in terms of the rate of change of beach volume, V, in time, t, given by: dV -= qin -qout +J(f)-R(f) dt and V = I [qin -qout + J(t)-R(t)] dt (6) Over any given period of time, T = nAf , comprised of n number of time steps of interval !!.t , the volume change of the beach can be computed by discretizing equations (6) according to: n ~v = L [qin -qout +J(n)-R(t)]nM (7) I Referring to the control cell schematic in Figure 33c, sediment is supplied to the control cell by dredged beach fill, J(t), or by the influx littoral drift from up-coast sources, qin =q LI, (where q L is the longshore transport rate on the updrift side of the control cell). Sediment is lost from the control cell due to the action of wave erosion and expelled from the control cell by exiting a) Accretion / Erosion Wave J accretion (sand delta) t 1 b) Coupled Control Cells c) Profile Changes / closure depth qin-0-qout I T h 1 accretion --•--- erosion ~ Figure 33: Computational approach for modeling shoreline change after Jenkins, et. al., (2007). N 0 San Luis Rey River SCALE 2000· 6000' 10000· ---o· 4000· sooo· ELEVATION REFERENCE TO MLLW Oceanside Buena Vista Lagoon Carlsbad Agua Hedlonda Lagoon South Beach Groin Batlqultos Lagoon Figure 34: Coupled control cell mesh of the farfield computational cell used to model the impacts of dredge disposal and removal of the South Beach Groin. littoral drift, q0u1 =qL2 , or by becoming ingested by the lagoon's tidal inlet, R(t). Here fluxes into the control cell (J(t) and qL,) are positive and fluxes out of the control cell (qL2 and q,ide) are negative. The beach sand volume change, dV/dt, is related to the change in shoreline position, dX/dt, according to: dV = dX .z./ dt dt (8) where Here, h is the height of the shoreline flux surface equal to the sum of the closure depth below mean sea level, he, and the height of the berm crest, Z1, above mean sea level (from Hunt's Formula.); and l is the length of the shoreline flux surface (see Figure 33c). Hence, beaches and the local shoreline position remain stable if a mass balance is maintained such that the flux terms on the right-hand side of equation (2) sum to zero; otherwise the shoreline within each control cell will move during any time step increment as: When dredge disposal produces a large episodic increase in J, an accretionary bulge in the shoreline (like a river delta) is initially formed (cf. t1 in Figure 33a). Over time the accretionary bulge will widen and reduce in amplitude under the influence of surf diffusion and advect down-coast with the longshore current, forming an accretion-erosion wave (cf. t2 & t3 in Figure 33a). The local sediment volume varies in response to the net change of the volume fluxes, between any given controi celi and its neighbors, referred to as divergence of drift= q;n -q0"1 , see Figure 33c. The mass balance of the control cell responds to a non-zero divergence of drift with a compensating shift, Ax , in the position of the equilibrium profile whose shape is calculated from equations (1)-(4) after Jenkins and Inman, (2006). This is equivalent to a net change in the beach entropy of the equilibrium state. The divergence of drift is given by the continuity equation of volume flux, requiring that dqldt on the left hand side of equation (5), is the net resultant of advective and diffusive fluxes of sediment plus the influx of new sediment, J, from dredge disposal, per the right hand side of equation (1 ). In response to the rate of change of volume flux through the control cell, the equilibrium profile will shift in time according to equation (9). If the divergence of drift is positive because more sand fluxes into the control cell due to longshore transport than leaves the cell, ( q;n -q0"1 = qL1 -q L2 > 0 ), then the equilibrium beach profile in that cell will shift seaward. Conversely, if the divergence of drift is negative because less sand fluxes into the control cell than is expelled from the cell by longshore transport, ( q;n -q011, = qL1 -q L2 < 0 ), then the equilibrium beach profile in that cell will shift landward, as diagramed schematically in Figure 33c. If a negative divergence of drift causes the equilibrium profile to shift sufficiently landward that it intersects the basement surface of the critical mass envelope, then the cycloidal shape of the profile is disrupted, and the equilibrium state of the profile is lost. The formulation for the longshore transport rate q L is taken from the work of Komar and Inman (1970) according to: (10) where q L is the local potential longshore transport rate; Cn is the phase velocity of the waves; S yx =Esinab cos ab is the radiation stress component; ab is the breaker angle relative to the shoreline normal; E=l/8pgH; is the wave energy density; p is the density of water; g is the acceleration of gravity; Hh is the breaking wave height; and, K is the transport efficiency equal to: K=2.2F:, (11) (12) Here c,b is the reflection coefficient which is calculated from the mean bottom slope, fJ (which is known either from the measured profiles or from the elliptic cycloids); and, a-is the radian frequency = 21T:!T, where T is the wave period. These equations relate longshore transport rate to the longshore flux of energy at the break point which is proportional to the square of the breaking wave height and breaker angle. By this formulation, the CEM computer code calculates a local longshore transport rate for at the up-drift and down-drift sides of each side of each control cells of the mesh in Figure 34. Longshore transport around the inlet jetties to Agua Hedionda Lagoon causes the sediment budget of the North Beach disposal site to be coupled to the sediment budget of the Middle Beach disposal site, since the prevailing southward littoral drift (cf. Figure 22) results in: (12) where the longshore transport entering Middle Beach at its northern end is the net of longshore transport exiting North Beach and the tidal influx into the lagoon, R(t). To calculate the tidal influx rates we run a tidal hydraulics model known as TIDE_FEM, (Jenkins and Inman, 1998). TIDE_FEM was built from some well-studied and proven computational methods and numerical architecture that have been successful in predicting shallow water tidal propagation in Massachusetts Bay [Connor & Wang, 1974] and estuaries in Rhode Island, [Wang, 1975 ], and have been reviewed in basic text books [Weiyan, 1992] and symposia on the subject, e.g., Gallagher (1981). A discussion of the physics ofTIDE_FEM is given in Jenkins and Wasyl (2003 & 2005). In its most recent version, the TIDE _FEM modeling system has been integrated into the Navy's Coastal Water Clarity Model and the Littoral Remote Sensing Simulator (LRSS) (see Hammond, et al., 1995). The TIDE FEM code has been validated in mid-to-inner shelf waters (see Hammond, et al., 1995; Schoonmaker, et al., 1994). A detailed description of the architecture and codes of the TIDE_FEM/ is given in Appendix-B. Calibrations for determining the appropriate Manning factors and eddy viscosities were performed by running the TIDE_FEM model on the Figure 35 bathymetry file and comparing TIDE_FEM simulations of inlet channel velocities against measurements by Elwany et al. (2005) during a complete spring-neap cycle of 13 -30 June 2005, as shown in Figure 36. Plant flow rates during this lagoon monitoring period were input to TIDE _FEM according to daily recordings by Cabrillo Power 1 LLC. Iterative selection of Manning factor n0 = 0.03011 and an eddy viscosity of & = 6.929 ft2/sec gave calculations of inlet channel velocities that reproduced the measured values to within 2% over the 18 day spring-neap monitoring cycle. Figures 3 7-42 map the tidal velocities in Agua Hedionda Lagoon from hydrodynamic simulations of the spring, neap and mean tides during the spring-neap cycle calibration period using the TIDFEM model. The flood and ebb current maximums and minimums in the inlet channel are found to lead the high and low ocean water levels by as much as 13.7 hours during the spring tides on 21 June 2005. Maximum flood tide currents on this day were 5.16 ft/sec, while maximum ebb tide currents were -2.87 ft/sec; the flood tide dominance due to the scavenging effect of the power plant intake rate on the available lagoon water volume which was operating at 501 mgd. Throughout the 18 day monitoring period, average flood tide currents in the inlet channel were 1.91 ft/sec while average ebb tide currents were -0.91 ft/sec while the power plant averaged an intake flow rate of 430.97 mgd. The amplitudes and degree of non-linearity in the inlet current time series simulated by the model closely duplicate that observed in the measured currents. The maximum error in simulating the ebb tide currents was found to be & L = +0.1 ft/sec. The maximum flood tide error in the modeled currents relative to observations was found to be & H = -0.05 ft/sec. Using the calibrated Manning factor n0 = 0.03011 and an eddy viscosity of & = 6.929 ft2/sec from these 2005 calibration simulations, the ocean water level forcing for the disposal period 1 January 2015 to 17 April 2015 was input into the TIDE _FEM model in order to calculate the tidal influx term, R(t), for the following beach evolution simulations in Sections 5.1 and 5.2 below. 5.1 Calibration of the Coastal Evolution Model for Middle and South Beach: The Coastal Evolution Modei (CEM) for Middle and South Beach was calibrated using the measured pre-and post-dredging beach profiles for Middle and South Beach during the 2014-2015 dredging event, in conjunction with daily beach fill placement volumes as reported in the monitoring report to the Regional Water Quality Control Board, San Diego Region, (NRG, 2015). Wave forcing for the CEM was based on shoaling wave data from Figure 8, while beach fill grain size was based on Figure 13. Daily beach fill volumes were assumed to be laid out over the 22 December 2014 profiles from Figures 1-6 in a standard beach fill template with a flat backshore platform and a 1: 10 (rise over run) seaward facing beach slope extending down to 0 ft. MLLW. Throughout the Middle and South Beach disposal period, that began on 1 January 2015 and ended on 15 April 2005, daily beach fill increments ranged from 0 yds3/day to 5,480 yds3/day and were successively added to the Middle and South Beach control cells in Figure 34, while the wave forcing continued to rearrange those fill volume increments according to flux balance relations in equations (5)-(9). Free parameters in the CEM, including the mass diffusivity in equation (5) and the longshore transport efficiency in equation ( 10) were adjusted through successive iterative simulations until the change in beach sand volume between 1 January 2015 and 17 April 2015 (when the post dredging beach surveys were done) matched the volumetric changes of the measured profiles in Figures 1-6. These volumetric changes were computed by the SolidWorks 3-d CAD model in Figure 43. ---+7.7 contour ---+4.0 contour ----3.0 contour -..e.o contour ----13.0contour ----18.0contour ----23.0 contour ---+7.1 contour ---+2.0 contour ----4.0 contour ----9.0 contout ----14.0 C:0111Dur ----19.0 contour ----24.0 contour ---~.4 contour ---0.0 contour ----5.0 contour ----10.0 contour ----15.0 contour ----20.0 contour ----25.0 contour ---+5.8 co moor ----1.0 contour ----6.0 contour ----11.0contour ----16.0 contour ----21.0 contour 5.6 contour ----2.0 contour ----7.0 contour ----12.0 contour ----17.0 cootour ----22.0 contour Figure 35: Pre-dredging bathymetric survey of Agua Hedionda Lagoon prior to the 2006/2007 dredging of Agua Hedionda Lagoon C > C, z -= C 0 ;:: • > • iii 6 4 2 0 -2 I Ocean Wate1r Level I ----Simulated lrdet Velocity I + + + Measured Inlet Velocity 1 average ebb current = -0. 91 ft/sec maximum ebb current= -2.87 ft/sec average flood current = 1. 91 ft/sec maximum flood current= 5.16 ft/sec 6 4 2 0 -2 -4 ____ ........, ...... _____ _..,.......,. __ ........,,_ _____ ._.,.......,. _________________ 4 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 June,2005 u ., i: i .2 • > -.. .5 Figure 36: TIDE_FEM model calibration using spring--neap cycle velocity measurements from Elwany, e al., 2005. Ocean water levels at station SO indicated in blue; hydrodynamic simulation of inlet channel velocity shown in green and compared against ADCP inlet velocity measurements shown as black crosses. Figure 37: Hydrodynamic simulation of tidal flow into Agua Hedionda Lagoon during spring flood tides. Figure 38: Hydrodynamic simulation of tidal flow out of Agua Hedionda Lagoon during spring ebb tides. Figure 39: Hydrodynamic simulation of tidal flow into Agua Hedionda Lagoon during mean flood tides. Figure 40: Hydrodynamic simulation of tidal flow into Agua Hedionda Lagoon during mean ebb tides Figure 41: Hydrodynamic simulation of tidal flow into Agua Hedionda Lagoon during neap flood tides. Figure 42 : Hydrodynamic simulation of tidal flow out of Agua Hedionda Lagoon during neap ebb tides. Erosion exposes pre-disposal beach surface per 22 Dec. 2014 survey Sand remaining 17 April 201 S on Middle and South Beach = 178,584 cubic yards Figure 43: SolidWorks 3-d CAD model of a composite surface overlay of the modeled post- disposal Middle and South Beach surface (brown) and the pre-disposal beach surface (silver). The SolidWorks mass properties tool calculates a volume of 178,584 yds3, representing the volume of sand retained after placing 229,693 yds3 over a 80 day period of beach nourishment. In Figure 43, the CEM modeled beach surface on 17 April 2015 (represented in brown) shows numerous patches where erosion has occurred and the pre-disposal beach surface (represented in silver) is exposed. The mass properties tool in SolidWorks calculates that the volume in the region between the two surfaces (representing the volume of beach fill retained) is 178,584 yds3, which matches the retention calculated directly from the pre-and post dredging beach profile measurements in Figures 1-6. The time stepped CEM wave-driven flux calculations which led to this calibration result are plotted in Figure 44. Initially, as sand was being placed on Middle Beach during the first 40 days, waves were generally small, short period and mostly approaching from the west/southwest with very little north swell (cf Figure 8b). Consequently, the calibration simulation begins with weak northward flowing longshore transport, Figure 44a, causing some sands near Batiquitos Lagoon to be transported northward and arrive at South Beach. During this early period, more sand enters the South Beach control cells, than leaves Middle Beach at the south inlet jetty, and the difference between the cumulative net transport at the south inlet jetty vs South Beach (Figure 44 b) initially creates a positive divergence drift (Figure 44c ). Consequently, Middle and South Beach gains sand volume during the first 40 days from wave driven transport, in addition to the gains from Middle Beach dredge disposal. But then, circa 19 February 2015, several west/northwest storm swells arrived with waves reaching 1.9 m in height prior to breaking ( cf. Figure 8b ). Coinciding with these larger swells, the longshore transport reverses direction toward the south and increases in rate, (Figure 44a) resulting in cumulative net transport toward the south at the south groin that exceeds cumulative net transport into Middle Beach at the south inlet jetty (Figure 44 b). The divergence of drift turns negative and remains that way throughout the remainder of the disposal activities, resulting in a loss of 51,109 cubic yards of sand from Middle and South Beach by 17 April 2015 due to wave-driven transport (Figure 44 c ). Negative divergence of drift is prevalent along Middle and South Beach because of the way the prevailing west/northwest swell and wind waves are refracted around the Carlsbad submarine canyon, creating higher shoaled wave heights at the southern end of South Beach than found further north nPHr thP inlPt to Ag1rn TlPnion,h T Hgoon, UThPrP. rPfrHr.tion pffp_rt« ofthP. f:HrkhHn ~11hmHrine Canyon are weaker, (cf. Figure 14). When the net sand loss to divergence of drift is superimposed on the incremental sequence of beach fill being placed on Middle and South Beach, we get a look at how the sand retention on Middle and South Beach varies throughout the disposal period (Figure 45). It appears from the black line in Figure 44 that sand volume with the standard 1: 10 (rise over run) beach fill template increases throughout most of the Middle and South Beach disposal period, before falling off around 2 April 2015, about 10 days after South Beach disposal was completed on 23 March 2015. Up until 2 April 2015, the placement of beach fill on Middle and South Beach had exceeded or at least kept pace with the sand loss rate to negative divergence of drift during the building north swell period. By the time the post dredging beach surveys were performed on 17 April 2015 (108 days after beach fill placement began), about 78% of the total volume of sand placed on Middle and South Beach remained, and the average rate of loss of beach fill was < dV I dt >= 473 yds3/day. Based on this average loss rate, the retention time,T0 , for fill placed on Middle and South Beach during the 2014/2015 dredging event was T -Vo o -<dV/dt> 229 693 = 485 days 473 (13) _g? C'O 0:: 0.02 ~-8. ~ 0.01 ~ en cu-.... (") ~ en ~ -g_ 0.00 o-.c en 0) C _3 -0.01 -= '5 20.00 '5-(") a.> en o -c 0.00 C >. ~o .... (/) g? -c -20.00 i5 ~ a.> en i g -40.00 C'O ~ '3- § -60.00 (..) 0.0 0.0 ---Inlet Jetty, q.., = Qu South Beach, Cab 1-01, q0.,, = qL2 20.0 40.0 60.0 80.0 ---Inlet Jetty. 'Equn41 --South Beach, Cab 1--01, Eqt2 n.r 20.0 40.0 60.0 21 February 2015 Begin South Beach Disposal " 73,637 yds3 : At' 100.0 ~ --------------- 1 January 2015 Begin Middle Beach Disposal= 156,056 yds3 ~ 'E(qL 1 -qL2) nt,,t t Middle and South Beach sand loss ___ = 51,109 ydsJ 0.0 20.0 40.0 60.0 80.0 100.0 Julian Day, 2015 Figure 44: Wave-driven fluxes during the Middle & South Beach CEM calibration: a) longshore transport rate, b) cumulative net transport; and, c) cumulative divergence of drift. 230.00 220.00 210.00 ;;;-200.00 ~ 190.00 >-180.00 0 170.00 1/l 160.00 -g 150.00 m 140.00 5 130.00 £ 120.00 ~ 110.00 > 100.00 <l 90.00 a> 80.00 E 70.00 :J 60.00 50.00 -c 40.00 ffi 30.00 00 20.00 g --Middle & South Beach Disposal, J,,/n) --Cumulative Divergence of Drift, I:(q,_, -q,2) nt.t --Beach Volume Retained, V(n) 21 February 2015 Begin South Beach Disposal = 78,011 yds3 1 January 2015 Begin Middle Beach Disposal = 151,682 yds3 / 22 March 2015. Middle & South Beach Fill Placed = 229,693 yds3 I 17 April 2015, Retention = 178,584 yds3 17 April 2015 Middle & South Beach Sand Loss = 51,109 yds3 .5 10.00 Jl'=~-,:c-~"':"".=-=-::-::--:-:-:::-:--=--:~..,,, Cl> 0.00 g> -10.00 ~ -20.00 (.) -30.00 -40.00 -50.00 ----------------------1 -60 .00 ---=l-rrn"TTTmrrrrrHTTTTTrrrTir-rrT"TTTTTT"rrHr-rnTTTTTT"rrn.....-.-rrrTTT"rrH"TTTTTT"rrTir-rn"TTT-rrrTTT"....-r,"TTTTTT,.,.,..........,.,..,.,.T> 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 110.0 Julian Day, 2015 Figure 45: CEM simulation of temporal variation in beach sand retention on Middle and South Beach (black line) as a result of the net between the incremental sequence of beach fiii piacement (green line) and the wave-driven divergence of drift (red line). The retention time from equation (13) is less than a typical bi-annual dredging cycle for Agua Hedionda Lagoon, suggesting that Middle and South Beach will be significantly eroded by the time the next increment of dredged beach fill is placed there. This result was a consequence of several factors: 1) placement of a fill on a highly eroded Middle and South Beach surface, that had probably been eroded to the basement of the critical mass envelope, (since it had been 3 years since previous nourishment was provided by lagoon dredging disposal); and 2) the fill placed in 2015 exceeded the critical mass carrying capacity and was laid down in a non- equilibrium profile shape. In the following section, we evaluate potential improvements in beach fill retention time using the composite cycloid-dune beach fill template and the appropriate amounts of beach fill volume. 5.2 Performance Simulations of the Proposed Cycloid Beach Fill Template: In this section, the calibrated Coastal Evolution Model (CEM) is run in long-term simulations of the fate of beach fill placed on Middle and South Beach throughout the entire period of record (1998-2015) that Agua Hedionda dredged sands have been disposed as beach fill on Middle and South Beach. Wave forcing for these simulations was based on shoaling wave data from Figure 15 for the period 1998-2016, while beach fill grain size was based on Figure 15. Daily beach fill volumes were derived from the dredge monitoring reports to the Regional Water Quality Control Board (NRG, 2015) and from an NRG dredging data base contained in an Excel spreadsheet< Dredge History.xis>. The spreadsheet contains volumetric time histories for 8 separate Middle and South Beach disposal events that are summarized in Appendix-A. The CEM was first run for these 8 events assuming the beach fill volumes were distributed using the standard 1: 10 (rise over run) template laid out over an eroded basement surface that was specified by the 22 December 2014 profiles from Figures 1-6. These runs were used to establish a Middle and South Beach baseline performance standard for average sand loss rate,< dV I dt >, and retention time, T0 • This baseline was then compared against a sensitivity analysis of the proposed composite cycloid-dune beach fill template to determine possible advantages of this new template design and establish the optimal beach fill volume for Middle and South Beach, given the limitations of: 1) its small wave-cut platform, and 2) intrinsically high longshore transport rates with negative divergence of drift in the presence of the prevailing west/northwest swell and wind waves. 5.2.1: Historic Sand Retention Baseline: The baseline CEM simulation results are plotted in Figure 46, showing average rate of loss of beach fill< dV I dt > (red) and the retention time, T0 , (blue) as a function of the total beach fill volume, V0 • Sand loss rates are scaled against the left hand axis in Figure 46, while retention time is plotted relative to the right hand axis. The solution points for the 8 historic Middle and South Beach dredge disposal events that occurred between 1998 and 2015 are plotted with star symbols that were subsequently connected by cubic spline fitting functions plotted as the red and blue colored lines. The cubic spline best fit curves in Figure 46 show a minimum in the sand loss rate for Middle and South Beach fill volumes near the critical mass Vc,;i = 200,890 yds3, which is the theoretical maximum carrying capacity of Middle and South Beach for supporting a beach profile in equilibrium. The theoretical minimum sand loss rate using the standard 1: 10 (rise over run) beach fill template is < dV I dt >= 431 yds3/day. When beach fill volumes are less than critical mass, V0 < Vcrit, there is insufficient sand to establish an equilibrium profile. A beach is in its most stable state with an equilibrium profile, Historic Middle & South Beach Fill Retention ----Historic Loss Rate, yds'/day >, <ti 1.00 -0 ----Historic Retention Time, days * * • Disposal Events, 1988-2015 520 ... Q) 510 a. M 0.90 500 (/) -0 490 >, -480 0 0.80 (/) 470 -0 C: 460 <ti 0.70 (/) :::::, 0 £ 0.60 450 ~ <ti 440 -0 a~n , e, ---,... . I\ .... 420 Q) -0 5; 0,50 ,::, V "O 0,40 410 .5 I- 400 c 0 390 ~ C: <ti (J) 0 0,30 minimum loss rate= 431 yds3/day 380 Q) 370 & 360 Q) +-' <ti 350 a:: 0.20 340 (/) (/) 0 330 ...J 0.10 Q) 320 OJ <ti 310 ... 4l 0.00 300 > <t 160.0 180.0 200.0 220.0 240.0 260.0 280.0 300.0 Total Disposal Volume, V0 ( thousands of yds3) Figure 46: CEM simulation of average rate of loss of beach fill, < dV I dt > (red, left hand axis), and the retention time, T0 , (blue, right hand axis), as a function of the total beach fill volume, V0 , for 8 historic Middle and South Beach dredge disposal events , 1998-2015. CEM solution points are plotted with star symbols that were connected by cubic spline fitting functions plotted as the colored lines. Simulations based on standard 1: 10 (rise over run) beach fill template assumption. but Middle and South Beach has a prevailing negative divergence of drift so that when equilibrium cannot be achieved due to insufficient sand volume, sand loss rates increase. For example, the measured sand loss rates during the 2010/2011 dredging event (where V0 = 163,996 yds3) were 519 yds3/day, or 20% higher than the theoretical minimum loss rate of 431 yds3/day (where theoretical minimum loss rate occurs when V0 ~ Vc,.u ). Consequently the retention time in 2011 was only T0 = 316 days; but would have increased to T0 = 466 days if the beach fill volume placed on Middle and South Beach were increased from V0 = 163,996 yds3 to Va = Ve,.;, = 200,890 yds3• Figure 46 also shows that retention time continues to increase as the disposal volume placed on Middle and South Beach exceeds the critical mass volume, but the sand loss rate also increase, because the excess san volume cannot be retained on the limited wave-cut platform which is only 600 ft. to 650. ft. in width. For beach fill volumes above critical mass the improvement in retention time is minor, because the excess sand cannot be supported in equilibrium on the limited wave-cut platform, which is only 550 ft. to 600 ft. wide. The excess sand is quickly lost to the negative divergence of drift, ( cf. Figures 44 b&c ), after which an equilibrium profile can be established, once V0 ~ Vc,.u . The initially high loss rate of the excess sand contributes to a higher average sand loss rates than what would have otherwise occurred if the beach fill volume were limited to the critical mass volume. Historically, what we find that when beach fill sand volume is increased by 41 % over critical mass using the standard 1: 10 beach fill template (as occurred during the 2000/2001 dredge event when Va = 281,195 yds3 on Middle and South Beach), the retention time retention time is only increased by 7 % from T0 = 466 days to T0 = 497 days. This is not a good return on the investment in beach fill for Middle and South Beach because the sand loss rate increases by 31 % to < dV I dt >= 566 yds3/day, or an increase in sand loss of 135 yds3/day over what would have otherwise occurred if the beach fill volume were limited to critical mass. 5.2.2: Sand Retention with the Cycloid-Dune Template. The cycloid dune templates proposed in Figures 25-30 are tested in long-term CEM simulations using the same model calibration parameters and the same wave forcing and grain size inputs used in the historic baseline simulations. We also assume that the proposed cycloid-dune template is laid out over an eroded basement surface specified by the 22 December 2014 profiles from Figures 1-6, the same assumption as was made for the baseline simulations. Because these cycloid-dune simulations are hypothetical, we do not limit ourselves to a mere handful of options for beach fill volumes, as was the constraint with the historic baseline simulations in Section 5 .2.1. Instead, a sensitivity analysis is performed on beach sand loss rates and retention times using beach fill volumes that range from V0 = 151,000 yds3 to V0 = 281,195 yds3• The dune portion of the proposed templates in Figures 25-30 holds Vv = 49,680 yds3 which were designed for a fill volume of V0 = 200,593 yds3, approximately the same as the critical mass, Vcru = 200,890 yds3. For the modeling scenarios that involved less than critical mass, the dune portion of the template was proportionally reduced in volume until reaching the absolute minim fill volume scenario of V0 = 150,913 yds3 , for which no dune component remains in the template. For the modeling scenarios where Va > Vcrit , the dune portion of the template was proportionally increased in width until reaching the scenario of V0 = 281,195 yds3, which is the historic maximum fill volume placed on Middle and South Beach. Because we do not have historical guidance on the temporal beach fill placement for these hypothetical scenarios, we have made several additional assumptions. First, we assume that the beach fill is laid down at a constant daily rate of 4,500 yds3 per day on Middle and South Beach, which was the average daily placement rate in the dredge monitoring reports to the Regional Water Quality Control Board (NRG, 2015) and in the NRG Excel spreadsheet(< Dredge History.xis>). Secondly, we assume a future condition with the South Groin removed (cf. Figure 34 for location). Thirdly, we pose a construction scenario whereby the back-beach dune portion of the composite cycloid-dune template is built first, starting at the south inlet jetty and adding sections to the dredge pipeline until the build-out of the dune reaches abeam of the EPS generating building. Building the dune first creates a "safe" reservoir of sand before the template can be fully constructed, and sand from this reservoir is only released to the lower eroded basement surface during periods of the highest tides and waves. After the buildout of the dune to the EPS generating building, the cycloid portion of the template is laid out beginning from the toe of the dune and spreading the material to down slope to MLL W, and working back towards the south inlet jetty, removing sections pipelines as the cycloids are completed. CEM beach evolution simulations were performed on 18 different disposal scenarios involving cycloid-dune beach fill placement on North Beach ranging from V0 = 151,000 yds3 to V0 = 281,195 yds3• Each cycloid-dune scenario was repeated 8 times using start dates and wave forcing corresponding to the 8 historic Middle and South Beach dredge disposal events listed in Appendix A. Selecting these specific start dates in the wave record in Figure 15 eliminates the randomness effects of the historic occurrence of extreme waves, and allows for direct comparisons with the results of the historic baseline in Section 5.2.1. Sand loss rates and retention times for the 8 separate outcomes of each scenario were ensemble averaged to produce the points plotted as crosses in Figure 47, and these solution points were then fitted to cubic splines (plotted in red for sand loss rates relative to the left hand axis; and plotted in blue for retention time relative to the right hand axis). Sand loss rates were calculated from lofting the beach surfaces in SolidWorks 3-d CAD at a given time, t = t;, during the CEM beach evolution simulations. The mass properties tool in SolidWorks was used to calculate the volume change between the beach surface at t = t; and the pre-disposal basement surface (represented by the 22 December 2014 surveys). An example of this procedure is illustrated by Figures 48 and 49 for the disposal scenario with V0 = Vcru = 280,345 yds3 of cycloid-dune beach fill, at t = +90 days post-dredge disposal. This scenario includes 79,455 yds3 on North Beach and 200,593 yds3 on Middle and South Beach with the South groin removed. In Figure 49, we zoom in on a cutplot at 90 days into the CEM beach evolution simulation, showing how the dune has been eroded and re-distributed by extreme wave runup at high tides, leaving a total residual volume on North Beach, South Beach and Middle Beach of V, = 214,994 yds3 , while the residual volume on Middle and South Beach is V,. = 173,053 yds3 of sand remaining between the basement surface (silver) and the new beach surface (brown). The new beach surface at t;= 90 days conforms closely to an elliptic cycloid profile. Given this residual sand volume the sand loss rate that has occurred on Middle and South Beach between the start of the CEM simulation at t = t O and t = f; = 90 days is given by: >, i 0.50 .... a> a. C") 1/) "C >, 0 0.40 1/) "C C: (1) (/) ::::, 0 .s:: ::.o.3o i "C Si: "C V ..; 0.20 C: (1) Cl) -0 * a::0.10 (/) (/) 0 ...J a> Cl ~ 10.00 160.0 Middle & South Beach Fill Retention Cycloid-Dune Template ----Loss Rate, yds'/day ----Retention Time, days ~ minimum loss rate = 306 yds3/day 180.0 200.0 220.0 240.0 260.0 Total Disposal Volume, V0 ( thousands of yds3) 280.0 300.0 700 690 660 670 660 650 640 630 620 610 Ill 600 >, ca 590 "C 580 1-0 570 560 ~- 550 ·-540 .... 530 5 520 ~ 510 Q) -500 Q) 490 0:: 480 470 460 450 440 430 420 410 400 Figure 47: Sensitivity analysis of average rate of loss of beach fill < dV I dt > (red, left hand axis) and the retention time, T0 , (blue, right hand axis) as a function of the total beach fill volume, V0 , using the proposed cycloid-dune templates (cf. Figures 25-30 ). Each solution point (crosses) is an ensemble average of 8 modeled outcomes coinciding with the timing of historic Middle and South Beach dredge disposal events, relative to the wave forcing in Figure 15. ~ a, t § ~ U1 100 'I" g~ j~ Iii• Ill ____________ ..,...,o 2000 111111 Northing, ft Figure 48: Three-dimensional CAD model of two overlaid surfaces on North, Middle and South Beach. The brown layer is the beach surfaces after waves re-work 280,345 yds3 of cycloid-dune beach fill, at t = +90 days post-dredge disposal; this includes 79,455 yds3 on North Beach and 200,593 yds3 on Middle and South Beach. The silver layer is the bottom of the critical mass envelope or basement surface (from the 22 December 2014 piofilc measurements). CAD model shown with 10 to 1 vertical exaggeration. 0 sand remaining @ t = + 90 days Middle+ South Beach= 173,053 cubic yards North+ Middle+ South Beach= 214,994cubicyards Figure 49: Cross-section of 3-d CAD model of two overlaid surfaces on North, Middle and South Beach after waves re-work cycloid- dune beach fill, (brown) at t = +90 days post-dredge disposal; and 2) the bottom of the critical mass envelope (as delineated in silver from the 22 December 2014 profile measurements). CAD model shown with 10 to 1 vertical exaggeration. 200,593-173,053 = 306 yds3/day 90 This gives an estimate of retention time (at t;= 90 days) of (14) T, Vcrit 200,593 X 90 _ 656 da S 0 (dV I dt)i=t; (200,593-173,053) y (15) Four such SolidWorks loftings of the CEM beach surfaces were done at four different times, t; =t; ,t2 ,t3 ,t4 , during each of the 18 scenario simulations, and the sand loss rates and retention times calculated at each of those times were ensemble averaged to give the points and cubic spline curves plotted in Figure 47. While an elliptic cycloid is an equilibrium beach surface, it does not produce a state of zero sand loss in the presence of a negative divergence of drift, which is the persistent littoral drift state along Middle and South Beach. Recall from Section 5.0 that when the divergence of drift is negative (because less sand fluxes into Middle Beach at the south inlet jetty than is expelled from South Beach abeam the EPS generating building, cf. Figure 44c ), then the equilibrium cycloidal beach profile will shift landward, eventually intersecting the basement surface of the critical mass envelope, (ie, the 22 December 2014 profile). Once this happens, then the cycloidal shape of the profile is disrupted, and the equilibrium state of the profile is lost. The concept behind the cycloid-dune template is that, as the cycloid begins approach an intersection with the basement surface of the critical mass envelope, (under the erosional effects of continued negative divergence of drift), it also intersects the base of the dune and receives i:iciclitioni:t I c::P.climP.nt c.ovP.r i:tc:: thP. cl11nP. P.roriP.'< i:tncl "PrPi:trl'< 011t ciown'<lopP i:tc.roc::.'< thP c::till inti:td cycloidal surface. So, the dune acts as a restoring mechanism that re-supplies the cycloid with sand lost to negative divergence of drift. Comparing the results in Figure 4 7 against the historic baseline in Figure 46, the general trends are similar, but the sand loss rates are greatly diminished and retention times significantly increased by using the proposed cycloid-dune template (in Figures 25-30). Again, the most efficient use of Agua Hedionda dredged sands occurs when the cycloid-dune template is filled to no more than critical mass ( V0 ~ Vcrit = 200,890 yds3), ~hich reduces average sand loss rates on Middle and South Beach to an absolute minimum of< dV I dt >= 306 yds3/day, while extending retention time to T0 = 656 days, sufficient to retain a sandy beach over a typical bi-annual dredge cycle. This is a 35% improvement in sand retention time over historical dredge disposal practices at Middle and South Beach, which could result in a reduction of sand influx rates into Agua Hedionda Lagoon by a similar factor during the months following Middle and South Beach disposal. If the cycloid-dune template is filled to more than critical mass ( V0 > Vc,it) by adding more sand to the back-beach dune, then Middle and South Beach retention time will increase beyond 656 days and reach as much as T0 = 693 days if the reserve sand volume in the dune were more than doubled to VD= 129,985 yds3 (achieving the historic maximum placement volume of V0 = 281,195 yds3). But, again, this is not a good return on doubling the investment in reserve beach fill for Middle and South Beach because retention time is only increased by an extra 5 weeks while the sand loss rate increases by 33 % to < dV I dt >= 406 yds3/day, (an additional 100 yds3 /day of sand loss). This inefficiency occurs because the enlarged dune encroaches further seaward into the middle bar-berm portion of the profile that is subject to more frequent wave attack, and the groin field formed by the inlet and discharge jetties at Middle and South Beach is already filled to carrying capacity at the critical mass of Ve,;,= 200,890 yds3• On the other hand, under-filling the cycloid-dune template, (by building a reduced dune), leads to accelerated sand loss rates and reduced retention times. If the Middle and South Beach cycloid dune templates were filled with the historic minimum beach fill of V0 = 163,996 yds3, (by under-building the back-beach dune with only 12,786 yds3) then sand loss rates would increase to 375 yds3/day and retention times would be reduced by to T0 = 437 days, a 33% reduction in sand retention time relative to the ideal build using beach fill equal to critical mass. The prevailing negative divergence of drift across Middle and South Beach causes the initial cycloid profile in the lower portion of the template to shift landward, and once intersection with the basement surface of the critical mass envelope occurs, there are insufficient sand reserves in the reduced dune to resupply the cycloid in the presence of continued negative divergence of drift. Once the reserve sand supply in the dune is exhausted, the cycloidal shape of the profile is disrupted, and the equilibrium state of the profile is lost. Even so, if the cycloid-dune template on Middle and South Beach were filled to a volume equivalent to the 2104/2015 disposal event (V0 = 229,693 yds3) by using a dune containing only 78,483 yds3, then sand retention times are still significantly better than what was achieved using the standard 1: 10 (rise over run) template. With this over-built dune in combination with the cycloid, retention times following the 2014/2015 dredge cycle could have been T0 = 676 days with sand loss rates reduced to 340 yds3/day, an improvement of 39% over what was achieved using standard Middle and South Beach disposal practices. 6) Summary and Conclusions: A detailed set of beach profile surveys at Middle and South Beach in Carlsbad CA were provided by Cabrillo Power I LLC, delineating beach surfaces before and after the 2104/2015 dredging of Agua Hedionda Lagoon, (AHL), which placed 229,693 yds3 between the south inlet jetty to Agua Hedionda Lagoon and the Encina Power Station (EPS) generating building. The surveys were accurately performed by Noble Engineers using differential GPS and known historic benchmarks. Three-dimensional CAD models were lofted from the measured points along the three (3) Middle Beach survey range lines (Cab 1-04 -Cab 1-06) and three (3) South Beach survey range lines (Cab 1-01 -Cab 1-03) to delineate the beach surfaces immediately before beach dredge disposal (based on the 22 December 2014 profile measurements) and immediately after dredge operations were completed (based on the 17 April 201 7 profile measurements). When these two surfaces were lofted together in a common reference frame, it was determined that 178,584 cubic yards of beach fill have been retained after placing 229,693 cubic yards on Middle and South Beach between 1 January 2015 and the post-dredge survey on 17 April 2015. This calculates to an average sand loss rate of 473 yds3/day and projects a sand retention time of 485 days. This is significantly longer than the retention time at the North Beach disposal site where retention time projected to only 33 days. To understand the reasons for the contrasting sand retention characteristics of North Beach vs. Middle and South Beach, a baseline beach evolution study was conducted using the Coastal Evolution Model (CEM) to hindcast the fate of beach fill placed on the three receiver beaches. The CEM was developed at the Scripps Institution of Oceanography with a $1,000,000 grant from the Kavli Foundation, (see bttp://repositories.cdlib,org/sio/techreport/58/ ), and is based on latest thermodynamic beach equilibrium equations published in the Journal of Geophysical Research. Inputs to the CEM baseline study were based on measured shoaling wave data, grain size data for the dredged sands, and daily beach fill volumes were derived from the dredge monitoring reports to the Regional Water Quality Control Board (Cabrillo, 2015) and from Cabrillo dredging data bases. Between 1998 and 2015 there have been eight (8) different events when Agua Hedionda dredged sands have been disposed concurrently on all three receiver beaches (North, Middle and South Beach). Historic dredged sand volumes placed on North Beach ranged from 62,030 yds3 to 161,525 yds3, while beach fill placed on Middle and South Beach ranged from 163,996 yds3 to 281,195 yds3 • The CEM beach evolution simulations of these events determined that the minimum sand loss rate occurs when beach fill volumes on the three receiver beaches are equal to the critical mass, which was found to be Vc,u = 79,471 yds3 for North Beach, (see companion North Beach report, Jenkins, 2017), and Vc,it = 200,890 yds3 for Middle and South Beach. The critical mass is the theoretical maximum carrying capacity of a beach fill site for supporting a beach profile in equilibrium. The carrying capacity of a beach is limited by the width of the wave-cut platform in the bedrock on which beach sands have accumulated over geologic time scales. The wave-cut platform at North Beach is only 550 ft. to 600 ft. in width and 600 ft. to 650 ft. at Middle and South Beach. Many of the beaches throughout north San Diego County are perched on narrow wave-cut platforms. The platforms are narrow because they were carved by wave action into erosion resistant bedrock formations during the present high-stand in sea level, and these narrow wave-cut platforms physically cannot hold large quantities of beach sand; and often become fully denuded during periods of high-energy winter waves. Another contributing factor to the limited carrying capacity of the three receiver beaches is that they are exposed to a prevailing negative divergence of drift caused by the way the bathymetry surrounding the Carlsbad Submarine Canyon produces variable wave shoaling along the length of these beaches. The presence of the Carlsbad Submarine Canyon creates a bright spot in the shoaling wave pattern that diminishes in intensity with increasing distances toward the north. For example, wave heights are locally higher at the inlet jetties than further to the North around Maple Avenue. The prevailing littoral drift transports beach sand southward throughout the entire Oceanside Littoral Cell; but the alongshore imbalance in shoaling wave height causes higher southerly longshore transport rates of sand at the southern end of each of the receiver beaches than at the northern ends. Consequently more sand exits each receiver beach at its southern end due to longshore transport, than enters at the northern ends from sand sources further updrift. This inequality in sand transport rates between the north and south ends of the receiver beaches is referred to as divergence of drift, and when the sand transport rates are higher at the down-drift end of the receiver beaches, it becomes a constant loss system referred to as negative divergence of drift. So, when beach fill volumes exceed the critical mass of the receiver beaches, the excess sand cannot be supported in equilibrium on its narrow wave-cut platform and is quickly lost to the negative divergence of drift. However, this effect is somewhat muted at Middle and South Beach because the AHL inlet jetties and the EPS discharge jetties produce a groin field which impedes the longshore transport at both the updrift and downdrift ends of the Middle/South Beach complex, and provides extra storage capacity for sand on the wave cut platform. Consequently retention times for beach fill on Middle and South Beach is longer than at North Beach. Historically, the CEM baseline study finds that when a standard 1: 10 (rise over run) beach fill template on North Beach is filled to critical mass, the theoretical minimum sand loss rate to negative divergence of drift is 1,495 yds3/day, and the sand retention time is 53 days (see the companion North Beach report, Jenkins, 2107). By contrast, the standard 1: 10 beach fill template at Middle and South Beach historically achieved minimum sand loss rates of 431 yds3/day, and the sand retention times of 466 days. But, when beach fill sand volumes at Middle and South Beach were increased by 41 % over critical mass ( as occurred during the 2000/2001 dredge event when 281,195 yds3 were placed on Middle and South Beach), the retention time is only increased by 7 % from T0 = 466 days to T0 = 497 days. In contrast, over-filling the North Beach receiver site produces an even worse return on beach fill investment. During the 2002/2003 dredge event, 161,525 yds3 were placed on North Beach, (103% increase over critical mass), but the retention time increased by only 26 % from T0 = 53 days to T0 = 67 days, while the sand loss rate increased by 61 % to 2,411 yds3/day. This is an increase in sand loss rates at North Beach of 916 yds3/day. Unfortunately, such increases in sand loss rates at North Beach correlate with proportional increases of sand influx rates into Agua Hedionda Lagoon. The 2010/2011 survey data show that AHL sand loss rates also increase when the fill volumes are less than the critical mass. Sand influx rates in 2010/2011 were 519 yds3/day when only 163,996 cubic yards were placed on Middle and South Beach (36,894 yds3 below critical mass requirements). Bear in mind that the critical mass is the minimum volume of sand required to establish an equilibrium beach profile on a wave-cut platform; and a beach is in its most stable state with an equilibrium profile. But with a prevailing negative divergence of drift along Middle and South Beach, equilibrium cannot be achieved due when there is insufficient sand volume, and consequently sand loss rates increase with a destabilized, non-equilibrium profile. Following CEM beach evolution analysis of the Middle and South Beach historic baseline, attention was given to finding a more effective beach fill template that could increase sand retention using beach fill from Agua Hedionda Lagoon dredging. Beach fill has typically been placed on Carlsbad beaches using a standard beach fill template with a flat backshore platform and a 1: 10 (rise over run) seaward facing beach slope extending down to Oft. MLL W. This convention dates back to the Regional Beach Sand Project, (AMEC, 2002). However, stable beach profiles in Nature have a much more gradual, curving profile with slopes that range between 1 :50 to 3:100. Formulations of equilibrium beach profiles are found in the U.S. Army Corps of Engineers Shore Protection Manual and later the Coastal Engineering Manual; and the latest most advanced formulation is known as the elliptic cycloid. The elliptic cycloid formulation can account for continuous variations in the equilibrium beach profile due to variability in wave height, period and direction when occurring in combination with variations in beach sediment grain size and beach sand volume. Therefore, a new beach fill template has been proposed here for Middle and South Beach referred to as the cycloid-dune template (see Figures 25-30). The shape of the template is based on the extremal elliptic cycloid which is the equilibrium profile for the highest wave in the period of record. But the extremal elliptic cycloid extends below the MLL W tide line and earth moving equipment which spread out the beach fill cannot work below MLL W. So, the template truncates the extremal elliptic cycloid at MLL W and places the residual volume of critical mass (totaling 49,680 yds3) in a back-beach dune that stretches 3,680 ft. from the south inlet jetty to Agua Hedionda Lagoon to the north end of the EPS generating building. While an elliptic cycloid is an equilibrium beach surface, it does not produce a state of zero sand loss in the presence of a negative divergence of drift, which is the persistent littoral drift state along Middle and South Beach. When the divergence of drift is negative, the equilibrium cycloidal beach profile will progressively shift landward as it loses sand to negative divergence of drift, eventually intersecting the basement surface of the critical mass envelope. Once this happens, then the cycloidal shape of the profile is disrupted, and the equilibrium state of the profile is lost. The concept behind the cycloid-dune template is that, as the cycloid begins to approach an intersection with the basement surface of the critical mass envelope, (under the erosional effects of continued negative divergence of drift), it also intersects the base of the dune and receives additional sediment cover as the dune erodes and spreads out downslope across the still intact cycloidal surface. Thus, the dune acts as a restoring mechanism that re-supplies the cycloid with sand lost to negative divergence of drift. The construction method envisioned for the cycloid-dune template begins with building the back-beach dune portion first, starting at the south inlet jetty and adding sections to the dredge pipeline until the build-out of the dune extends beyond the South Groin abeam of the north end of the EPS generating building. Building the dune first creates a "safe" reservoir of sand before the template can be fully constructed, and sand from this reservoir is only released to the lower eroded basement surface during periods of the highest tides and waves. After the buildout of the dune to the southern end of South Beach, the cycloid portion of the template is laid out beginning from the toe of the dune and spreading the material down slope to MLL W, and working back across Middle Beach to the south inlet jetty, removing pipeline sections as the cycloids are completed CEM beach evolution simulations of the Middle and South Beach cycloid-dune templates were run for future conditions with the South Groin removed, and show significant improvements in sand loss rate and retention time relative to the historic baseline. Again, the most efficient use of Agua Hedionda dredged sands occurs when the cycloid-dune template is filled to no more than critical mass (200,890 yds3 for Middle and South Beach), which reduces average sand loss rates on Middle and South Beach to an absolute minimum of 306 yds3/day, while extending retention time to 656 days. This is a 35% improvement in sand retention time over historical dredge disposal practices at Middle and South Beach. If the cycloid-dune template is filled to more than critical mass by adding more sand to the back-beach dune, then Middle and South Beach retention times will increase beyond 656 days. If the reserve sand volume in the dune on Middle and South Beach were increased by a factor of 2.6 to 129,985 yds3 (producing a cycloid dune equivalent to the historic maximum placement volume of V0 = 281,195 yds3) then retention time could be extended to a maximum of 693 days. But, again, this is not a good return on doubling the investment in reserve beach fill placed in the back-beach dune because retention time is only increased by 5 weeks while the sand loss rate on Middle and South Beach would increase by 33% to 406 yds3/day, (an additional 100 yds3/day of sand loss). This inefficiency occurs because the enlarged dune encroaches further seaward into the middle bar-berm portion of the profile that is subject to more frequent wave attack, and the groin field formed by the inlet and discharge jetties at Middle and South Beach is already filled to carrying capacity at the critical mass of Ve,;,= 200,890 yds3• On the other hand, under-filling the cycloid-dune template, (by building a reduced dune), leads to accelerated sand loss rates and reduced retention times. If the Middle and South Beach cycloid dune templates were filled with the historic minimum beach fill of V0 = 163,996 yds3, (by under-building the back-beach dune with only 12,786 yds3) then sand loss rates would increase to 375 yds3/day and retention times would be reduced by to T0 = 437 days, a 33% reduction in sand retention time relative to the ideal build using beach fill equal to critical mass. The prevailing negative divergence of drift across Middle and South Beach causes the initial cycloid profile in the lower portion of the template to shift landward, and once intersection with the basement surface of the critical mass envelope occurs, there are insufficient sand reserves in the reduced dune to resupply the cycloid in the presence of continued negative divergence of drift. Once the reserve sand supply in the dune is exhausted, the cycloidal shape of the profile is disrupted, and the equilibrium state of the profile is lost. Even so, if the cycloid-dune template on Middle and South Beach were filled to a volume equivalent to the 2104/2015 disposal event (V0 = 229,693 yds3) by using a dune containing only 78,483 yds3, then sand retention times are still significantly better than what was achieved using the standard 1: 10 (rise over run) template. With this over-built dune in combination with the cycloid, retention times following the 2014/2015 dredge cycle could have been T0 = 67 6 days with sand loss rates reduced to 340 yds3/day, an improvement of 39% over what was achieved using standard Middle and South Beach disposal practices. 7) References: AMEC, 2002, "Regional Beach Sand Project Post-construction Monitoring Report for Intertidal, Shallow Subtidal and Kelp Forest Resources", submitted to SANDAG, http://www.sandag.org. Cabrillo, 2015, '"'Order 96-32: First Quarter 2105, Second Quarter 2015 and Final Monitoring Report for Agua Hedionda Lagoon Dredging", submitted to California Regional Water Quality Control Board, 30 pp. CDIP, 1984-1988, "Coastal data information program, monthly reports," U.S. Army Corps of Engineers, California Department of Boating and Waterways, Monthly Summary Reports #97-#150. CDIP, 1976-1995, "Coastal Data Information Program, Monthly Reports," U.S. Army Corps of Engineers, California Department of Boating and Waterways, SIO Reference Series, 76- 20 through 95-20. CDIP, 1993-1994, "Monthly Summary Report," SIO Reference Series (93-27) through (94-19). CDIP, 2016, "Coastal Data Information Program" http://cdip.ucsd.edu/ Elwany, M. H. S., A. L. Lindquist, R. E. Flick, W. C. O'Reilly, J. Reitzel and W. A. Boyd, 1999, "Study of Sediment Transport Conditions in the Vicinity of Agua Hedionda Lagoon," submitted to California Coastal Commission, San Diego Gas & Electric, City of Carlsbad. Elwany, M. H. S., R. E. Flick, M. White, and K. Goodell, 2005, "Agua Hedionda Lagoon Hydrodynamic Studies," prepared for Tenera Environmental, 39 pp.+ appens. Ellis, J.D., 1954, "Dredging Final Report, Agua Hedionda Slough Encina Power Station," San Diego Gas and Electric Co., 44pp. Inman, D. L. and B. Brush, 1970, "The coastal challenge" Science, vol38, no. 5 pp36-45. Inman, D. L. & S. A. Jenkins, 1985, "Erosion and accretion waves from Oceanside Harbor," p. 591-593, in Oceans '85: Ocean Engineering and the Environment, IEEE and Marine Technology Society, v. 1,674 pp. Inman, D. L. and Masters, P. M., 1991, "Coastal sediment transport concepts and mechanisms," Chapter 5 (43 pp.) in State of the Coast Report, San Diego Region, Coast of California Storm and Tidal waves Study, V. S. Army Corps of Engineers, Los Angeles District Chapters 1-10, Appen. A-I, 2 v. Inman, D. L., M. H. S. Elwany and S. A. Jenkins, 1993, "Shorerise and bar-berm profiles on ocean beaches," Jour. Geophys. Res., v. 98, n. Cl 0, p. 18,181-199. Inman, D. L., S. A. Jenkins, and M. H. S. Elwany, 1996, "Wave climate cycles and coastal engineering practice," Coastal Eng., 1996, Proc. 25th Int. Conf,(Orlando), Amer. Soc. Civil Eng., Vol. 1, Ch. 25, p. 314-327. Inman, D. L. & S. A. Jenkins, 1997, "Changing wave climate and littoral drift along the California coast," p. 538-549 in 0. T. Magoon et al., eds., California and the World Ocean '97, ASCE, Reston, VA, 1756 pp Inman, D. L. & S. A. Jenkins, 1999, "Climate change and the episodicity of sediment flux of small California rivers," Jour. Geology, v. 107, p. 251-270. Inman, D. L. & S. A. Jenkins, 2004, "Scour and burial of objects in shallow water," p. 1020-1026 in M. Schwartz, ed., Encyclopedia of Coastal Science, Kluwer Academic Publishers, Dordrecht, Netherlands. Jenkins, S. A. and D. W. Skelly, 1988, "An Evaluation of the Coastal Data Base Pertaining to Seawater Diversion at Encina Power Plant Carlsbad, CA," submitted to San Diego Gas and Eiectric, Co., 56 pp. Jenkins, S. A., D. W. Skelly, and J. Wasyl, 1989, "Dispersion and Momentum Flux Study of the Cooling Water Outfall at Agua Hedionda," submitted to San Diego Gas and Electric, Co., 36 pp.+ appens. Jenkins, S. A. and J. Wasyl, 1993, "Numerical Modeling of Tidal Hydraulics and Inlet Closures at Agua Hedionda Lagoon," submitted to San Diego Gas and Electric, Co., 91 pp. Jenkins, S. A. and J. Wasyl, 1994, "Numerical Modeling of Tidal Hydraulics and Inlet Closures at Agua Hedionda Lagoon Part II: Risk Analysis," submitted to San Diego Gas and Electric, Co., 46 pp.+ appens. Jenkins, S. A. and J. Wasyl, 1995, "Optimization of Choke Point Channels at Agua Hedionda Lagoon using Stratford Turbulent Pressure Recovery," submitted to San Diego Gas and Electric, Co., 59 pp. Jenkins, S. A. and J. Wasyl, 1997, "Analysis of inlet closure risks at Agua Hedionda Lagoon, CA and potential remedial measures, Part II," submitted to San Diego Gas and Electric, Co., 152 pp. + appens. Jenkins, S. A. and J. Wasyl, 1998a, Analysis of Coastal Processes Effects Due to the San Dieguito Lagoon Restoration Project: Final Report, submitted to Southern California Edison Co., 333 pp. Jenkins, S. A. and J. Wasyl, 1998b, Coastal Processes Analysis of Maintenance Dredging Requirements for Agua Hedionda Lagoon, submitted to San Diego Gas and Electric Co., 176 pp.+ appens. Jenkins, S. A. and D. L. Inman, 1999, A Sand transport mechanics for equilibrium in tidal inlets, Shore and Beach, vol. 67, no. 1, pp. 53-58. Jenkins, S. A. and J. Wasyl, 2001, Agua Hedionda Lagoon North Jetty Resoration Project: Sand Influx Study, submitted to Cabrillo Power LLC., 178 pp. + appens. Jenkins, S. A. and J. Wasyl, 2003, Sand Influx at Agua Hedionda Lagoon in the Aftermath of the San Diego Regional Beach Sand Project, submitted to Cabrillo Power LLC., 95 pp. + appens Jenkins, S. A. and J. Wasyl, 2005, Hydrodynamic Modeling of Dispersion and Dilution of Concentrated Sea Water Produced by the Ocean Desalination Project at the Encina Power Plant, Carlsbad, CA. Part II: Saline Anomalies due to Theoretical Extreme Case Hydraulic Scenarios, submitted to Poseidon Resources, 97 pp. Jenkins, S. A. and J. Wasyl, 2005, "Oceanographic considerations for desalination plants in Southern California coastal waters," Scripps Institution of Oceanography Tech. Rpt. No. 54, 109 pp + appendices. http://repositories.cdlib.org/sio/techreport/54/ Jenkins, S. A. and J. Wasyl, 2005, "Coastal evolution model," Scripps Institution of Oceanography Tech. Rpt. No. 58, 179 pp+ appendices. http://repositories.cdlib.org/sio/techrepmt/58/ Jenkins, S. A. and D. L. Inman, 2006, "Thermodynamic solutions for equilibrium beach profiles", Jour. Geophys. Res., v.3, C02003, doi:10.1029, 21pp. Jenkins, S. A., Inman, D.L., Michael D. Richardson, M.D., Thomas F. Wever, T.F. and J. Wasyl, 2007, "Scour and burial mechanics of objects in the nearshore", IEEE Jour. De.Eng, vol.32, no. 1, pp 78-90. Jenkins, S. A. and J. Wasyl, 2011, "Hydrodynamic Approach to Wetland Restoration by Optimization of Bridge Waterways", Department of Transportation, District 11 Environmental Division, MS-242, Technical Report# 1 lAl 766, 258 pp. Jenkins, S. A. and S. Taylor, 2015, "Storm and Tidal Conditions Determination for Coastal Drainage Design," submitted to Office of Highway Drainage Design Division of Design, MS 28, California Department of Transportation Jenkins, S. A., 2017, "Beach Equilibrium Analysis of North Beach Disposal Options for Dredged Sands from Agua Hedionda Lagoon, Carlsbad, CA," submitted to Cabrillo Power I LLC, 66 pp. Merkel, 2008, " Agua Hedionda Outer Lagoon Flood Shoal Maintenance Dredging Sediment Characterization Report, Tech Rpt # ACOE-2001100328-SKB, submitted to Cabrillo Power, LLC., 45 pp. NOAA, 1998, A Verified/Historical Water Level Data@ http://www.opsd.nos.noaa.gov/data res.html NWS, 2009, "National Weather Service Daily Climate Reports," http://www.wrh.noaa.gov/sgx/obs/rtp/carlsbad.html U.S. Army Corps of Engineers, 1985, "Littoral zone sediments, San Diego Region, October 1983 -June 1984", Coast of California Storm and Tidal Wave Study, CCSTWS 85-11. U.S. Army, Corps of Engineers (USACE), 1991, "State of the Coast Report, San Diego Region," Los Angeles District, CA: Coast of California Storm and Tidal Waves Study, Final Report 1. U. S. Army, Corps of Engineers (USACE), 2006, "Coastal Engineering Manual," Engineering Manual 1110-2-1100, U.S. Army, Corps of Engineers, Washington, DC, (in 6 volumes). U.S. Department of Commerce National Ocean Service, 1986, "Tide tables 1986, high and low water predictions for west coast of North and South America", 234 pp. USGS, 1997, "USGS Digital Data Series DDS-37 at INTERNET URL," http://wwwrvares.er.usgs.gov/wgn96cd/wgn/wq/region18/hydrologic unit code. Van der Meer, J.W., 2002. Wave Run-up and Overtopping at Dikes. Technical Report, Technical Advisory Committee for Water Retaining Structures (TAW), Delft, the Netherlands Appendix-A: Maintenance Dredging History for Agua Hedionda Outer Lagoon Dredging Disposal Year Date Volume cubic Influx Volume Comments Basin *Location Start Finish yard Days Yds3/Day cubic yard 1955 Aug-55 Sep-55 90,000 Outer 90,000 s Maintenance 1957 Sep-57 Dec-57 183,000 822 223 Outer 183,000 s Maintenance 1959-60 Oct-59 Mar-60 370,000 821 451 Outer 370,000 s Maintenance 1961 Jan-61 Apr-61 227,000 396 573 Outer 227,000 s Maintenance 1962-63 Sep-62 Mar-63 307,000 699 439 Outer 307,000 s Maintenance 1964-65 Sep-64 Feb-65 222,000 703 316 Outer 222,000 s Maintenance 1966-67 Nov-66 Apr-67 159,108 789 202 Outer 159,108 s Maintenance 1968-69 Jan-68 Mar-69 96,740 700 138 Outer 96,740 s Maintenance 1972 Jan-72 Feb-72 259,000 1,067 243 Outer 259,000 s Maintenance 1974 Oct-74 Dec-74 341,110 1,034 330 Outer 341,110 M Maintenance 1976 Oct-76 Dec-76 360,981 731 494 Outer 360,981 M Maintenance 1979 Feb-79 Apr-79 397,555 851 467 Outer 397,555 M Maintenance 1981 Feb-81 Apr-81 292,380 731 400 Outer 292,380 M Maintenance 1983 Feb-83 Mar-83 278,506 699 398 Outer 278,506 M Maintenance 1985 Oct-85 Dec-85 403,793 1,006 401 Outer 403,793 M Maintenance 1988 Feb-BB Apr-88 333,930 852 392 Outer 333,930 N,M,S Maintenance 1990-91 Dec-90 Apr-91 458,973 1,095 419 Outer 458,973 M,S Maintenance 1992 Feb-92 Apr-92 125,976 366 344 Outer 125,976 N Maintenance 1993 Feb-93 Apr-93 115,395 365 316 Outer 115,395 M Maintenance Outer 74,825 N , 1993-Dec-93 Apr-94 158,996 365 436 Outer 37,761 M Maintenance 1994 Outer 46,410 s Outer 106,416 N 1995-96 Nov-95 Apr-96 443,130 731 606 Outer 294,312 M Maintenance Outer 42,402 s 1997 Sep-97 Nov-97 197,342 579 341 Outer 197,342 M Maintenance Dec-97 Feb-98 59,072 92 642 Middle 59,072 M Modification 1998 120,710 M Feb-98 Jul-98 214,509 150 1,430 Inner Modification 93,799 s 1999 Feb-99 May-99 155,000 304 510 Outer 155,000 N Maintenance 141,346 N 2000-01 Nov-00 Apr-01 422,541 701 603 Outer 195,930 M Maintenance 85,265 s 2002-03 Dec-02 Apr-03 354,266 730 485 2004-05 Jan-05 Mar-05 348,151 704 495 2006-07 Jan-07 Apr-07 333,373 763 437 2006-09 Dec-08 Apr-09 299,328 733 408 2010-11 Der.-10 Apr-11 226,026 736 307 2014-15 Dec-10 Apr-11 294,661 736 400 TOTAL 8,528,842 MAINTENANCE TOTAL 8,255,261 *Location: N= North Beach; M = Middle Beach; S = South Beach Green= pre-back-passing sand influx rates Red = post-back-passing sand influx rates 161,525 N Outer 131,377 M Maintenance 61,364 s 100,487 N Outer 170,515 M Maintenance 77,149 s 149,166 N Outer 121 ,036 M Maintenance 63,167 s 104,141 N Outer 102,000 M Maintenance 93,185 s 62,030 N Outer 93,696 M Maintenance 70,300 s 64,968 N Outer 156,056 M Maintenance 73,637 s 8,526,640 APPENDIX-B: Details of the TIDE_FEM Tidal Transport Model A finite element approach was adapted in preference to more common finite difference shallow water tidal models, e.g., Leendertse (1970), Abbott et al (1973), etc. Finite difference models employ rectangular grids which would be difficult to adapt to the complex geometry of the systems of channels of the Agua Hedionda. It is believed that large errors would accumulate from attempting to approximate the irregular boundaries of the Agua Hedionda system with orthogonal segments. On the other hand, finite element methods allow the computational problem to be contained within a domain bounded by a continuous contour surface, such as the Sf contours stored within the bathym file. The finite element mesh used to model the tidal fluxes at Agua Hedionda Lagoon using the TIDE _FEM model is shown in Figure B-1. TIDE_FEM employs a variant of the vertically integrated equations for shallow water tidal propagation after Connor and Wang (1975). These are based upon the Boussinesq approximations with Chezy friction and Manning's roughness. The finite element discretization is based upon the commonly used Galerkin weighted residual method to specify integral functionals that are minimized in each finite element domain using a variational scheme, see Gallagher (1981). Time integration is based upon the simple trapezoidal rule [Gallagher, 1981]. The computational architecture of TIDE_FEM is adapted from Wang (1975), whereby a transformation from a global coordinate system to a natural coordinate system based on the unit triangle is used to reduce the weighted residuals to a set of order-one ordinary differential equations with constant coefficients. These coefficients (influence coefficients) are posed in terms of a shape function derived from the natural coordinates of each nodal point. The resulting systems of equations are assembled and coded as banded matrices and subsequently solved by Cholesky's method, see Oden and Oliveira (1973 and Boas (1966). We adapt the California coordinates as our global coordinate system (x, y) to which the nodes are referenced, with x (easting) and y (northing). The vertical coordinate z is fixed at 0.0 ft NGVD and is positive upward. The local depth relative to 0.0 ft NGVD is h and the mean surface elevation about 0.0 ft NGVD is IJ. The total depth of water at any node is H = h + '1· The vertically averaged xy-components of velocity are (u, v). The continuity and momentum equations may be written from Connor and Wang, (1974), as: (Bl) Extreme, High Water (EHW) 7.65 ft MLLW ---0 500 1000 1500 2000 2500 HORIZO TAL SCALE IN FEET Extreme High Water (EHW) 7 .65 ft MLL W Extreme High Water (EHW) 7.65 ft MLLW DR. SCOTT A. JENKINS CONSULTING Agua Hedionda Finite Element Mesh SCOTT A. JENKINS PhD & JOSEPH WASYL Figure B-1: Finite element mesh used to model tidal transport at Agua Hedionda Lagoon Here qx, qy are mass flux components 1/ qx=pfudz -h 1/ qy =pfvdz -h (B2) (B3) and qi is the mass flux through the ocean inlet due to water surface elevation changes in the estuary: (B4) Fp is the pressure force resultant and Fxx, Fxy, Fyy are "equivalent" internal stress resultants due to turbulent and dispersive momentum fluxes ,, pgH2 F = fpdz =-- P 2 -h (BS) o o Fyx = Fxy = e(-qy +-qJ oy ox and c is the eddy viscosity. Bx and By are the bottom stress components oh Bx =rx +pgH-ox oh BY =ry + pgH oy (B6) In Equation (B6), 'tx and Ty are the bottom shear stress components that are quasi- linearized by Chezy-based friction using Manning's roughness factor, no: (B7) where Cz is the Chezy coefficient calculated as: C = l.49 H 116 (B8) z no Boundary conditions are imposed at the locus of possible land/water boundaries, Sr in the bathym file and at the ocean inlet, So. Flux quantities normal to these contours are denoted with "n" subscripts and tangential fluxes are given "s" subscripts. At any point along a boundary contour, the normal and tangential mass fluxes are: 1/ qn = f pundz=a.xqx+anyqy -h 1/ qs = f pusdz = -anxqx + anyqy -h a nx = cos(n, x) any= cos(n,y) (B9) Components of momentum fluxes across a boundary are equivalent to internal force resultants according to: Fnx =anxCFxx -FP)+anyFyx Fny =an/Fyy -FP)+anxFxy On land boundary contours, the flux components are prescribed on land (BIO) (Bl 1) On the ocean boundary, the normal boundary forces (due to sea surface elevation) are continuous with ocean values, and the mass exchange is limited by the storage capacity of the estuary. Hence qnm = qi at inlet (Bl2) In the problem at hand Fnn is prescribed at the inlet by the ocean tidal elevation, 170 , and the inlet sill depth, ho according to -pg( )2 Fnm = -1/o + ho 2 on So (B13) Ocean tidal forcing/unctions 170 were developed in Section 3. The ocean boundary condition as specified by Equation (B12) places a dynamic boundary condition on the momentum equations and a kinematic boundary condition on the continuity equation that is constrained by the storage rating curve. Solutions are possible by specifying only the dynamic boundary condition, but then mass exchanges are controlled by the wetting and drying of individual grid cells with associated discretization and interpolation errors which threaten mass conservation. The technique of over specifying the ocean boundary condition with both a dynamic and kinematic condition is discussed in the book by Weiyan (1992). The governing equations and the boundary conditions are cast as a set of integral functionals in a variational scheme, [Boas, 1966]. Within the domain of each element of the mesh, Ai the unknown solution to the governing equations is simulated by a set of trial functions (H, q) having adjustable coefficients. The trial functions are substituted into the governing equations to form residuals, (RH, Rq). The residuals are modified by weighting functions, (~H, ~q). The coefficients of the trial functions are adjusted until the weighted residuals vanish. The solution condition on the weighted residuals then becomes: A, Jf Rq~qdA = 0 A, By the Galerkin method of weighted residuals, [Finlaysen, 1972], the weighting functions are set equal to nodal shape functions, <N>, or: The shape function, <N>, is a polynomial of degree which must be at least equivalent to the order of the highest derivative in the governing equations. The shape function also provides the mechanism to discretized the governing equations. The shape function polynomial is specified in terms of global (California) coordinates (Figure B2) for the first nodal point, N 1 of a generalized 3-node triangular element of area Ai,. Wang (1975) obtained significant numerical efficiency in computing the weighted residuals when the shape functions of each nodal point, Ni, are transformed to a system of natural coordinates based upon the unit triangle, giving Ni ~ Li, see Figure 8b. The shape functions also permit semi-discretization of the governing equations when the trial functions are posed in the form: Specifying the Shape Function <N> for any 3-Node Triangular Element a) Global (California) Coordinates y .,__ __________ X <N> = (N1, N2, N3) Coordinate ______ ;_..> Ni = Li Transform N1 =[(X2Y3 -X3Y2 )+(y2 -y3 )x+(x3-Y2 )y ] /2Ai 2Ai =(x1-x3)(Y2 -y3 )-(x2-x3)(Y1 -y3) 1.0 b) Natural Coordinates y L1 L3 ____________ X 1.0 x=L1 x 1 +L2 x 2 +L3 x 3 y=L1 Y1 +L 2Y2 +L3y3 L1 +L2+L3=l.O Figure B2: Shape function polynomial and transform to natural coordinates for a generalized 3-node triangular element; (a) 3- node element in California coordinates; (b) 3-node dement in natural coordinates. H(x,y,t) = LH/t)N;(x,y) I q(x,y,t) = Lq1(t)N/x,y) J (B14) Discretization using the weighting and trial functions expressed in terms of the nodal shape functions allows the distribution of dependent variables over each element to be obtained from the values of the independent variables at discrete nodal points. However, the shape function at any given nodal point, say N1, is a function of the independent variables of the two other nodal points which make up that particular 3-node triangular element. Consequently, the computations of the weighted residuals leads to a series of influence coefficient matrices defined aii = ~-fJ N;N1dA , I ff 8N . S--=-N .-1 dA lJ A, I ax 1 ff oN. t,,=-N .-1 dA !, A,.. I 8y (B15) g .. k =-1 ff NN. aNk dA " A ' 1 a· i X h.k = _!_ff N.N. oNk dA " A; ' 1 By The influence coefficient matrices given by equation (B 15) are evaluated in both global and natural coordinates. Once the influence coefficients have been calculated for each 3-node element, the weighted residuals reduce to a set of order-one ordinary differential with constant coefficients. The continuity equation becomes: I( a !i d:,) = -~~[guk(Hiqxk + HkqJ+hyk(H;qyk +Hkqyj I( a, d!;1 )--~~[g,,(q,,q, )+ h,,(q,,q.,, )]+ N, ~ N1S, + g~s,H, L(al} ddqy}) = -LL [gijk (qxjqyk )+ hijk (qyJqxk )]+ N; L NJS 17 + g LfuH; { j k J I (B16) Equations (B 16) are essentially simple oscillator equations forced by the collection of algebraic terms appearing on the right hand side; and are therefore easily integrated over time. The time integration scheme used over each time step of the tidal forcing function is based upon the trapezoidal rule, see Gallagher 1981) or Conte and deBoor (1972). This scheme was chosen because it is known to be unconditionally stable, and in tidal propagation problems has not been known to introduce spurious phase differences or damping. It replaces time derivatives between two successive times, ~t = tn+l -tn, with a truncated Taylor series. For the water depth it would take on the form: dH -= 17(t) dt Hn+I -Hn = ~ fon+l +17J+EM E=__!__(M)2 d217 12 dt2 (B17) To solve equation (Bl 7), iteration is required involving successive forward and backward substitutions. The influence and friction slope coefficient matrices together with the trapezoidal rule reduce equations to a system of algebraic equations [Grotkop, 1973] which are solved by Cholesky's method per a numerical coding scheme by Wang (1975). For more details, refer to the TIDE_FEM code in Appendix-I of Jenkins and Wasyl (1996), and Gallagher (1981) or Oden and Oliveira (1973). APPENDIX-C: Equilibrium Beach Profile Algorithms: JOURNAL OF GEOPHYSICAL RESEARCH, VOL. Ill, C02003, doi:10.1029/2005JC002899, 2006 Thermodynamic solutions for equilibrium beach proftles Scott A. Jenkins 1 and Douglas L. Inman2 Received 26 January 2005; revised 17 November 2005; accepted 10 December 2005; published 11 February 2006. [1] Solutions are developed for beach profiles using equilibrium principles of thermodynamics applied to simple representations of the nearshore fluid dynamics. Equilibrium beaches are posed as isothermal shorezone systems of constant volume that dissipate external work by incident waves into heat given up to the surroundings. By the maximum entropy production formulation of the second law of thermodynamics (the law of entropy increase), the shorezone system achieves equilibrium with profile shapes that maximize the rate of dissipative work performed by wave-induced shear stresses. Dissipative work is assigned to two different shear stress mechanisms prevailing in separate regions of the shorezone system, an outer solution referred to as the shorerise and a bar-berm inner solution. The equilibrium shorerise solution extends from closure depth (zero profile change) to the breakpoint, and maximizes dissipation due to the rate of working by bottom friction. In contrast, the equilibrium bar-berm solution between the breakpoint and the berm crest maximizes dissipation due to work by internal stresses of a turbulent surf zone. Both shorerise and bar-berm equilibria were found to have an exact general solution belonging to the class of elliptic cycloids. The elliptic cycloid allows all significant features of the equilibrium profile to be characterized by the eccentricity and the size of the ellipse axes. These basic ellipse parameters are evaluated by process-based algorithms and empirically validated parameters for which an extensive literature already exists. The elliptic cycloid solutions displayed wave height, period and grain size dependence and demonstrate generally good predictive skill in point-by-point comparisons with measured profiles. Citation: Jenkins, S. A., and D. L. Inman (2006), Thermodynamic solutions for equilibrium beach profiles, J. Geophys. Res., lll, C02003, doi:l 0.1029/2005JC002899. 1. Introduction [2] Comprehensive reviews of the previous geomorphi- cally and statistically based attempts to define and param- eterize equilibrium beach profiles are contained in Dean [1991], Inman et al. [1993], and Short [1999]. Here we follow the prior conceptual treatment of beach profiles, where an equilibrium profile is defined as the shape attained by a beach in response to steady wave forcing over long periods of time. Field measurements and hydraulic models show that beach profiles change shape with changing intensity of wave action, but given sufficient time under steady forcing the beaches attain a constant form [e.g., Shepard and La Fond, 1940; Bagnold, 1947; USACE, 1947; Inman and Filloux, 1960; Nordstrom and Inman, 1975; Inman et al., 1993]. These findings led to the temlS "summer/winter" beach cycles and the concept of an equilibrium beach profile [Inman, 1960; Inman and Bagnold, 1963]. 1Marine Physical Laboratory, Scripps Institution of Oceanography, Unlv.rslly ofCallfoinlt, SM Dicto, La Jolla, C~Hfo,nla, USA. 11n!egrmivc Oce<\nog:rnphy Division, Scripps lnilitullon of Oceanogra- phy, University of California, San Diego, La Jolla, CaUfomia, USA. Copyright 2006 by lhe American Geophysical Union. 014 8-0227 /06/2005JC002899$09 .00 [ 3] The first general formulation for a continuous train of waves that break 011 a beach appears to be that of Inman and Bagnold [1963]. Their equilibrium formulation was based on a balance between the net upslope gradient in wave energy versus the downslope gradient in potential energy due to gravity. The energy gradient for their model was subsequently refined [Bailard and Inman, 1981] by including velocity skewness moments for the upslope wave energy gradients. In order to calculate equilibrium profiles, these models require an a priori set of velocity time series with their spectral moments. [4] A mass flux balance formulation of the equilibrium condition was introduced by Bowen [1980] and later applied by Stomis et al. [2002] to barrier beach migration in the Caspian Sea. Bowen's formulation requires a balance between the upslope and downslope transport rates using relations of bedload and suspended load transport after Bagnold [1963]. Bowen's transcendental solutions had several asymptotic f01ms of interest. One was based on the assumption of vanishing suspended load in the presence of wave asymmetry from the Stokes second harmonic and had the fonn, h = .4.'t:us. h1 these limiting fonns, the profile factor A increases with increasing sediment fall velocity (or grain size). Like the Bailard and Inman [1981] fonnulation, these solutions depend on numerous point- dependent variables that are seldom known. C02003 I of 21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 [s] Dean [1977] developed an equilibrium profile fotmu- lation h = Ax215 based on the assumption that the dissipation of wave e11e-18y per unit area of beach is cons_tallt, and the form h = Ar whe11 the dissipation of wave e11ergy per unit volume of water is constant. Later, Dean [1991], following the work of Bruun [1954], elaborated on the h = .Ax213 solution, indicating that A is an increasing function of grain size. This derivation was based on the assumption that the wave heigh!' decreases linear!.y with depth. This assumption limits the domain of the .?3 solution to the inner 1>m-tion of the beach profile, shoreward of the wave break point (although it is commonly applied to the entire profile in engineering practice). [6] Inman et al. [1993] developed an equilibrium profile consisting of two conjoined curves, both of the fonn h = Ax"' that were best-fit to 60 separate beach profiles. They considered distinct summer and winter equilibrium states because these correlate with seasonal wave climate and give recurrent curves for beach profiles taken in the field In the region between the breakpoint and the berm crest ("bar- benn"), it was found that only one exhibited a curvature exponent of m -213, and that was for the transitional profile between the winter and summer equilibrium profiles. The best fit results produced variations in curvature between slllllmer and winter equilibria of 0.29 S: m S: 0.55 in the bar-benn, and 0.21 S: m S: 0.5 in the shorerise. Inman et al. [1993] show that the conjoined shorerise and bar-benn curves gave significantly better matches to the data, with rms en-ors in depth generally under 35 cm nud volume errors ranging from 13 to 134 m3/m. Comparable en·ors for the single fitted curves using the Deau [199 I] formulation were 90 cm and 225 to 321 m3/m [l11ma11 el al., 1993). 2. Thermodynamic Formulation of Beach Equilibrium [7] Equilibrium states in natural systems are governed by tlle second law ofthennodynamics [Moore, 1962; Halliday and Resnick, 1967; Anderson, 1996]. Thermodynamics has been used to solve equilibrium problems in other geophys- ical processes such as volcanism, mineral fonnation, and melts [Newton et al., 1981; Salje, 1988; Anderson, 1995], thermohaline and mesoscale ocean circulation [Sverdrup et al., 1942; Eckart, 1962; Imawaki et al., 1989; Val/is et al., 1989; Salmon, 1998], atmospheric and marine layer dynam- ics [Bane, 1995; Bohren and Albrecht, 1998], global climate models [ Ozawa et al., 2003] and heterogeneous fluid and sediment systems [ Grinfeld, 1991 ; Casas-Vazquez and Jou, 1991]. The longshore sand transport solutions oflnman and Bagnold [1963] and Komar and Inman [1970] are also thennodynamic applications, where steady longshore trans- port is an equilibrium response to the rate of work perfonned by longshore directed radiation stress. Further, thennodynamics offers the advantage of providing solu- tions from measurable fundamental properties (macroscopic variables). These properties typically represent averages over time of the gross characteristics of a system. In the shorezone they include incident wave height Hoo, wave period T, beach sand sizes D, and the location of the profile relative to some benchmark, X1• [s] We begin with a simple representation of the universe consisting of two domains, the shorezone system bounded in green (Figure la), and everything outside the shorezone system referred to as the surroundings. In the system the sand elevation h is measured positive downward from mean sea level and cross-shore position x is positive in tlle offshore direction (Figure 1 b ). The system boundaries are stationary, enclosing a constant volume V that contains a fixed volume of sand v. sufficient for equilibrium to occur in the presence of a maximum incident wave height if 00• Equilibrium profile states are fully contained within the system boundaries. The seaward poundary is a vertical plane at the critical closure depth he corresponding to the maximum incident wave [e.g., Kraus and Harikai, 1983]. The landward boundary is a veLtical plaite at U1e bean crest (cross), a distance ,1'1 from a bench mark. The cross-shore length of the system from the benn crest to closure depth is .:t. The distance from the point of wave breaking to closure depth is Xc2 such that Xe = Xc2 + X2, where .X2 is the distance from the benn crest to the origin of the shorerise profile near the wave breakpoint. We consider equilibrium over time scales that are long compared with a tidal cycle and profiles that remain in the wave dominated regime where the relative tidal range (tidal range/H) < 3 [Short, 1999]. Under these conditions, the curvilinear coordinate that defines the profile referenced to mean sea level (MSL) vertical datum is, d~ = /(dx)2 + (dh)2 = V1 + _;,2 dh = Vl+hddx x = dx. h' = dh (1) dh' dx where d~ is calculated separately for inner (bar-benn, d~1) and outer (shorerise, d(i) portions of the conjoined profile. [9] Fluxes of energy into and out of the shorezone system are shown by arrows crossing the system boundaries in Figure la. Work W per unit length of shoreline .:ly is performed on the system by the incident waves that provide energy to the system at a rate given by where E = pgH2/8 is energy per unit longshore surface area, p and g are water density and acceleration of gravity, Cg is wave group velocity, and His local rms wave height. The waves shoal and break inside the shorezone system, dissipating wave energy into an increment of heat dQ. This evolution of heat produces an incremental entropy change dS=dQ r. where Ta is absolute temperature. Heat is removed from the shorezone system to tlle surroundings primarily by advection and turbulent diffusion in tlle nearshore circula- tion system [Inman et al., 1971] and secondarily through heat of vaporization in sea spray. [10] The second law of thermodynamics (often referred to as 'the entropy law') is a necessary condition for equilibrium and requires that a natural process that starts in one equilibrilllll state and ends in another will cause the 2 of 21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 a. Entropy lncntaee of surroundings, dS = dQ IT SuR"OUndlnga dQ ( heat of vaporization l llllfft) dW ( ~,k~) =g dQ ( latent heat l -systlffland volume, Y. I b. fa: 1: Shorllrise 'I' Bffenn Xci x,, -~ "' x~ -5 "z 0 1 MSl E ..; 5 " l,o Ito Figure 1. Definition schematic for (a) coordLnates of a thermodynamic system in the shorezone, taken as n rnp1·ese11tative cross section ofa unifonn three-dimensional control cell, and (b) physical Cartesian coordinates [after Inman et al., 1993] that apply to the thennodynamic system. Note that symbols with a circwnflex (e.g., X.,) extend to system boundaries while !hose without occur within boundm·ies. entropy of the system plus its surroundings (universe) to increase, {)S) _ {)S) + {)S) > O (2) 8t UNIVERSE -{)/ SYSTEM {)t SURllOUNDING [11] The dcci ive i ue with 1· pect to the thcnnodynamic state of the shorezone system is the fate of the heat evolved within it and whether the entropy increase associated with tbat .heat evolution is l'Ctained by the system or expo1ted 10 its surroundings. We adopt a heat transport/entropy produc- tion f01mulation for the shorezone system that is an ana- logue of that used to describe dissipation in global climate state models [Paltridge, 1975, 1978; Ozawa et al., 2001, 2003]. The derivation for entropy production in a fluid system is found in de Groot and Mazur [ 1984] and Landau and Lifshitz [1980] and can be written, ~~)UNIVERSE=/* [{)(~;Ta)+ v' · (pcvT,u) + pv' · u] dV +Jq·n, dA (3) r. where cv is the specific heat at constant volume, u is the fluid velocity, p is the fluid pressure, q is the diabatic heat flux taken as positive when oco1.1rring outward ncross system boundaries and n, is the unit normal vector on the system boundary. The volume integral in (3) is the rate of change of entropy of the fluid system and represents the first L 11 on the right side of (2). The surface integral. in (3) is taken over the system boundary (green line in Figure la) and represents the discharge rate of entropy into the surrow1dings due to heat flu: across the system bowid&y. When a fluid system is in a steady state, Chandrasekhar [1961] has shown that the first law of thermodynamics reduces the volume integral in (3) to: {)S) I I I 4' -=--v'·qdV+ -dV {)t SYSTEM T, T, (4) where qi is the dissipation function representing the rate of viscous dissipation of kinetic energy per unit volume. The first term on the right side of (4) represents the entropy change due to transport of latent heat by advection and diffusion. If latent heat is transported out of the system 3 of 21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 (q positive), the system losses entropy. Because entropy is a state function of the system, it must remain constant if the system is to achieve equilibrium [Landau and Lifshitz, 1980]. Therefore, any equilibrium fluid system that exports latent heat to its surroW1dings must compensate for the associated entropy loss through the production of new entropy at an equivalent rate. [12] When heat fluxes across the system boW1dary, the entropy of the surroW1dings changes at a rate given by the surface integral in (3). This surface integral can be expressed in terms of a volume integral through the appli- cation of Gauss's theorem, giving the entropy change of the surroW1dings in terms of heat fluxes within the system volume 8S) ;q·n, JI -= -dA= -'v·qdV 8t SUAAOUNDING Ta T, + J q·'v(i)dv (5) The first term on the right side of (5) is the entropy change occurring in the surroW1dings due to the latent heat that was imported from the system, while the second is due to heat conduction along temperature gradients formed within the system between regions of hot and cold Entropy changes in the surrouridings due to latent heat transpo1t are equal and opposite in sign to those occurring in the system (4), and taken together, produce no net change in the total entropy of the universe. hlstead, the entropy of the universe can only be changed by temperature gradients and viscous dissipation occurring within a fluid system, as foWld after inserting (4) and (5) in (2), 88) = f .! dV + J q · 'v(_!_)dv > 0 (6) 8I UNIVERSE T. T. [13] When applying (6) we assume the shorezone system is isothe1mal and hence the second term on the right is vanishingly small. We support this assumption by noting that the body of empirical data from the field has never shown warmer water Wlder breaking waves than foWld elsewhere in the shorezone, nor have episodes of high waves been correlated with episodes of elevated surfzone temperatures. If the shorezone is isothermal, then no entro- py production is possible from the heat conduction mech- anism and the second law by (6) reduces to, 8S) =f_! dV>O 81 UNIVERSE T, (7) [14] Under these circumstances the first law of thermo- dynamics (4) requires the rate of entropy production by viscous dissipation inside the system to be in balance with the rate at which entropy is discharged to the surroW1dings by latent heat transport, !_! dV = f .!...'v · qdV T. T. (8) This balance maintains constant entropy inside the system, 8S) _0 8t SYSTEM - (9) The particular value at which the shorezone entropy remains constant is determined by the number of grains of sand contained within that system (Appendix A). [1s] If the entropy flux balance between system and surroW1dings in (8) were not upheld, then equilibrium states would not be possible in the shorezone because the system could not maintain constant state function as in (9). Without latent heat transport from the shorezone, it will W1dergo a progressive build up of entropy as the train of incident waves continually work on the system, adding more and more heat while those waves are dissipated. In this circum- stance, the shorezone becomes the thermodynamic equiva- lent of Joule's experiment [Zemansfy and Van Ness, 1966] for which the first law of thermodynamics requires dQ = dW. Assuming a constant bottom slope as a lowest order approximation, this equation can be solved for the increase in local water temperature l!i.Ta that would occur in time /!;.t as a consequence of continuous working by a steady train of incident waves: Here Hb is breaker height, 'I is a factor relating the depth of wave breaking hb to breaker height Hb = '/hb, tan~0 is the mean beach slope, and c" = 3,941.3 J/kg °K is the specific heat of sea water at 29l°K (18°C) and 35%0 salinity [Cox and Smith, 1959]. For a breaking wave of height Hb = 1 m with 'I -4/5 and a nomirial beach slope of tanj30 = 0.025, the temperature of the nearshore waters would increase continuously by about 3°K (or 3°C) every 24 hours. Such warming is not observed in nature, and surf temperatures are not known to increase with wave height, indicating that shorezones do not violate (8). The fact that shorezones appear to be isothe1mal (when solar flux is negligible) suggests that thermodynamic equili- brium is a common and persistent state for these systems in nature. [16] Systems like the shorezone that achieve equilibrium through the dissipation of external work into the heat of a reservoir in the surroW1dings, belong to the general ther- modynamic system known as external mechanical irrevers- ibility, [Zemansfy and Van Ness, 1966]. While (7) and (9) are conditions for equilibrium and (8) defines the type of equilibrium, these are not sufficient conditions to define a unique equilibrium state for the shorezone system. To obtain unique solutions, we adopt the criteria of maximum entropy production (MEP) that has been successfully applied to certain steady state equilibrium climate states by Dewar [2003], and Ozawa et al. [2003]. The MEP criteria is a particular form of the second law (3), that requires the entropy of the Wliverse not only increases when the system proceeds from one equilibrium state to the next, but that the entropy increase is a stationary maximum. The validity of the MEP criteria is based on observational and numerical evidence showing that in general, non-linear systems having many degrees of freedom for dynamic equilibrium tend to 4 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 those states, among all possible states under the second law, that maximize entropy production. [17] When the MEP criteria is applied to (7) subject to (9), we neglect reflection and talce the state variable asso- ciated with the external work done by incident waves as the independent v01iable 011d then seek a stationary maximum for the viscous dissipation, which becomes the dependent state variable. Since the shorezone system volume 011d sediment volume have been fixed, the fluid/sediment inter- face (bottom profile) is the only remaining thermodynamic coordinate that is unrestrained 011d available to maximize dissipation. From Batchelor [1970], the average rate of dissipation of mechanical energy per unit volume in a 2- dimensional, incompressible fluid system is <I>= 2µ(w · w) = ('i' · w), where ( ) denotes time averaging; Gi is the fluid vorticity generated by the action of viscosity µ 011d ;-is the time varying wave induced shear stress, including bottom shear stresses, internal shear stresses 011d granular friction at the fluid sediment interface [Bagnold, 1956; Inman and Bagnold, 1963]. We assume that no vorticity or dissipation (due to bottom ventilation) occurs within that portion of the system occupied by the sediment mass. Let dV' = aydA' represent a volume increment of the remaining portion of the system that contains the fluid vorticity, where dA' is 011 increment of area bounded by the closed contour C arow1d the fluid portion of the system (red contour, Figure la) and ay is a unit length of shore.line. Applying these assumptions and definitions to Stokes theorem, fw · n,dA' =fit· dC, the average dissipation rate of the system becomes: [1s] When (10) is used to maximize entropy production in (7), only the segment of contour integration talcen along the bottom profile produces a ch011ge in the state variables of heat and work, as all remaining segments are comprised of fixed system boundaries. fu the shorezone system, the bottom profile defines the pathway along which heat 011d work are evolved 011d both state variables are path depen- dent in an irreversible process. Hence, the MEP formulation of the second law in (7) reduces to maximization of a simple line integral: 8~ = ~Yj(-r. u)d(;. > o 8t) UNIVERSE Ta (maximum) (11) [19] Our equilibrium problem now becomes that of find- ing the profile curve C = C(h, x, x') after (1) that makes the integral in (11) a stationary maximum. This c011 be accom- plished with calculus of variations using a ch011ge of variables in the integrand of (11) in terms of a generalized functional F(h, x, x') written F(h,x,x') = (f · u): (12) With the functional in (12) the entropy integral in (11) is maximized by solving the Euler-Lagt311ge equation [Boas, 1966], ~8F _ 8F =O dh8x' 8x (13) General solutions to (I 0) are given in section 3 for the shorerise 011d bar-berm profiles, while particular solutions are found in section 4. 3. General Solutions [20] We will pose separate formulations for the viscous dissipation in the shorerise (shoaling zone) 011d bar-berm (surf zone) portions of the shorezone system (Figure lb). When applied to (13), these separate formulations will yield general solutions for the shorerise profile C2 011d bar-be1m profile C1 that happen to belong to the same class of equation. The solutions for the shorerise and bar-berm are conjoined at the wave break point. [21] The simplest surrogate for the shorerise is one that is uniform in the alongshore direction in the region between closure depth 011d the wave breakpoint (Figure I b ). The fluid dynamics in this region are approximated by the linear shoaling transformation of the shallow water Airy wave, u = Um cos( at -lex) _H(x) CF Um -2 y h(x) R ( )1/4 H(x) = vTci h(x) k=-(J- Jgh(x) (14) Here, k = 21t/wavelength is local wave number, CJ = 21t/ period is radi011 frequency, um(x) is velocity amplitude at the sea floor boundary layer, 011d H00 is incident wave height. Local wave height 011d depth, HM and h~"), are talcen with respect to local curvilinear coordinate of the bottom profile Ci, as shown in Figure 1 b. The Airy approximation in (14) has been shown in Mei [1989] to be valid over sloph1g bottoms if the following mild slope condition is satisfied: (15) where tan ~ = dhldx is the local bottom slope. Exactly how much smaller th011 unity (15) must be is not definite, but its largest value is at the breakpoint where kh = <J(H,J-y g)1'2• We assume there is some N » 1 such that (15) is satisfied everywhere in the shorerise by requiring dh ~ Edx ~ O(L) E = ~ (Hb) 1/2"" rtlS (Hoo) 2/5 N yg 21/SN 8'Y (16) where L is a characteristic length scale, 1;. is a stretching factor proportional to the Airy wave mild slope factor N. 3.1. Shorerise Profile [ 22] futemal she01· stresses are neglected in the shorerise due to the absence of wave breaking. A simple power law formulation is used to prescribe the bottom shear stress, f = T0 cos(crl -lex+ <p) To = pqu;;, = pK,.U::, (17) 5 of21 C02003 JENKfNS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 where 'To is the shear stress amplitude; i,p is the phase angle between the bottom shear stress and the oscillatory potential flow velocity from Airy theory; c1 is the quadratic drag coefficient and n is the shear stress velocity exponent referred to as shear stress linearity. The particular value of n varies with the dependence or c1 011 pnrnmeters of dynamic similitude e.g., oscillatory Reynolds nwnber (It.= U:,,lcrv), grain Reynolds number (Reg= urr:J)lv), Keulegan-Carpenter (inverse Strouhal) number (St= u.,/cr,i,), where 11. is bottom rougiu1ess, [ e.g., Taylor, 1946; Ke11lega11 011d Carpenter, 1958; S/eath, 1984). !fwe generalize er.-(.R'~88;'), then n = 2 + 2} + l + w, and the shear stress amplitude can be written in terms of a proportionality factor KT that is independent of Um and consists of a collection of other factors contained in R0, Reg and S, that make (17) dimensionally correct. [23] From (14) and (17), the dissipation rate per unit length of profile varies with depth h as (--) AV /H·l 'tU '!"' II = l'"T COS<p I'm = /rJ(.+1)/4 (18) pK.,. COS !p H'.'+I gJf,H+l )/4 1:u-"" -80!•+11/2 where n = fJ/L°' and the second integration constant is -1i/4n. The first root is given by, R = R, = (2J2} [4!1h" -4!121?". +-2-(1 -4!1h" + 4!12112")] !+CY. and the second root by, R = Rb =(21~1,) 2 [1 -4!1h" +4n2h2". +-2-(4!1/i" -4!12h2")] !+CY. (22a) (22b) Here f.1l, f.2l are elliptic integrals of the first and second kind, respectively. [24] The general solution given by (22) belongs to a class of equations known as elliptic cycloids [Boas, 1966]. We can show that by making a transformation into polar coordinates (r,0) with a substitution of variables: where the work factor -ro is independent of h, x and x'. When 9 = arc cos(! -2!1h") (23) we select from (I) an infinitesimal arc length of the shorerise profile having the fonn, from which we get dC.2 = vi+ x'2dh then the integral in (I 1) for which we seek a stationary maximum becomes: (19) In terms of nond:imensional variables denoted by an underscore, x = x/L; h = h/L, the following functional F is collected from the order-I te1111s in (19) for use in the Euler- Lagrange equation in (13) (20) Because fJFlfJ!, = 0, the first integration of(13) using (20) gives, q=f (21) where ex = 3 ( n + 1 )/2 and il is the first integration constant. We can rationalize the integrand of (21) using two separate Euler substitutions after Gradshteyn and Ryzhik, [1980, sections 2.261 and 2.264). These provide a general solution with two roots that has the following dimensional form, x=------h2"+-arccos(l-2!11i") 0c ... -1i, .. [ ~;,--1 ] &./R n 2n (22) ~ -h2" = -1-sin2 9 n 4!12 (24) With (23) and (24), equation (22) reduces to two types of elliptic cycloids having the general polar coordinate form: 2ri'ki•) x=x2 =-0-(9-sin9) m (25) where r is the radius vector measured from the center of an elli1,se whose sem:imajor and semiminor ax.es are a, b and e·~) is fhc elliptic integral of the first or second kind (k1.z> depending on which of the two cycloid types we resolve from (22a) and (22b). We limit our discussions to the solutions for the type-a cycloids that result from the first root in (22a) because these were shown to be in good agreement with field data. The polar equivalent of the type-a cycloid from (25) has a radius vector whose magnitude is [ ti21l _l 112 av1-=-e2 r = r. = n2 6i112 9+ b2 cos2 OJ = Jsin2 9 + (! -e2) cos2 9 (26) where e is the ~enu'icity of the ellipse given by e = .jl -(b2 / a2). TI1e polar fo1m of the type-a cycloid in (22.a) is based on the elliptic integral of the second kind that has an analytic approximation, P.1 = (1i12)./{2 -e2)/2, see Hodgman [1947]. The integration constant n in (22) is determined from the dimensions of the ellipse axes by noting when 0 = arc cos (1 -2nh"') = 1i, then nh"' = 1. This gives: (type-a cycloid) (27) 6 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 type-a cycloid \ \ \ .... Boundary Conditions x@P(J), Xi =X1 (8=0) Xa =Xa x@P(3), (8=1t) P(3) ... h Figure 2. EUiptic cycloid solution (~·ed line) fo1· the equilibriwn be~h profile traced by a point on a rolling ellipse with semimajor and semiminor axis a, b, eccentricity e = J 1 -(b2 / a2) polar COOl'dinafe vector r, and angle ofrotation e, and e: = 1.0. rt is api>are.itt lhat tlie root R in (22a) is equivaJei,t to ,-f/•~)/b in (25) when 2 I -e2 = (1 + a) Hence, the eccentricity of the elliptic cycloid is govcmed by the shear stress lil1eority, 11 (28) The inverse of(22) subject to (25) gives the companion polar equation for the elliptic cycloid, h = h2 = 1'rEX2 (l -c~so) = r(I -cos9) 21,(1ti,,) 9-smO (29) [2s] A geometric representation of the type-a elliptic cycloid used in the general solution for the shorerise profile is shown in Figure 2 as traced by an ellipse having eccentricity e = 0.75 and e: = 1. The equilibrium beach profile is given by the trajectory of a point on the semimajor axes of an ellipse that rolls seaward in the cross-shore direction under the plane of h = h2 = 0. This trajectory defines the el.liptic cycloid and tl1e segment traced by the first half of a rotation cycle (0 < 0 < 1'r) of the rolling ellipse is the equilibrium beach profile (solid red curve). The depth of water at the seaward end of the profile (0 = 1t) is h = 2a in the case of the type-a cycloid. The length of the profile Xis equal to the semi-circumference of the ellipse, X = 2aJJ21 e!!! 1'r a J2 -e2 e E 2 at 9 = 1'r (type-a cycloid) (30) The detemunation of the particular values for a or b semi.axes and e, will be established from boundary conditions and field measurements in section 4. 3.2. Bar-Berm Profile [26] The bar-berm spans the region of shorezone between the wave break poh1(. aud the becm crest. (Figure lb) where the surf zone dissipation may be represented by several simple analytic formulations. The simplest is a dissipation formulation based on a depth-limited bore, ('f ·u) = !scg I E=g pgH2 Cg =ygh H = K. h~ (31) where K. is the bore decay factor relating bore height to some powel' Tl of the local wat<:r dep1h after formul:1tio1ts derived from measurements in open channel flow detailed in Chow [1959] and He11rle1·s011 [1966). From (I) we lake the altemative represemation for nn infmitesinlal arc length of the bar-bemi profile, (32) With (31) and (3 2), the integral in (11) that must be made stationary to maximize entropy production becomes, 7 of21 C02003 JENKINS AND INMAN: THERlvIODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 where _!_EC = K h(4<,tJ/2 dh1 (lX[ g I [ dxt K _ 411+ I l/2 2 I -16 pg K, In terms of nondimensional variables, the order-I terms collected from (33) give the following form for the functional to be used in the Euler-Lagrange equation in (13) as, ( h'2 ) I+ =-e:2 (34) Since {)F/0!_1 = 0, the first integration of (13) using (34) yields q=J (35) where CJ is the dimensionless integration constant. The indefinite integral in (35) for the depth-limited bore is equivalent to (21) when a= 411 -1 and integrates into an elliptic cycloid of the form (22) having the alternative polar form (23)-(25) with an eccentricity, where 11 is defined in (31 ). (21] The integration constant is CJ= fi,J/L4'fl-J = (2a)-4T1+1_ For a linear bore (11 = 1) after Bowe11 et al. [1968], the eccenllicity _of the bnr-benn cycloid becomes e = ./ffj."" 0.707 While the integ,:ation COllStant is C1 "'(2ar3 for the type-a bar-berm cycloid. [2s] An empirical approach to the formulation of surf zone dissipation is obtained from the work of Thornton and Guza (1983] that produced estimates of the intemal dissi- pation per unit area due to wave breaking. By fitting empirical relations to measured breaking wave distributions, they formulated two separate dissipation functions. Elliptic cycloid solutions can be derived for the bar-berm of these two relations using calculus similar to that outlined in (32)-(35). The details of these derivations are given in Appendix B. The first of these two empirically based solutions follows from the hypothesis that waves break in proportion to the distribution of all waves and produces a type-a bar-benn ·cycloid widt ecceutri.cily e = ;j97IT ,..,., 0.904 based 011 a dimensional integratio11 constant of (20)-10. Alten1atively, Tliomton 011d Guza [1983] developed a second dissipation function from the assumption that waves break in proportion to the distri- butio11 of the lUl'gost waves. This asswnption resulted in a bar- bemi cycloid having llll eccentricity e = J'ffi "" 0.845 and integration constalll of(2a)-6 for the type-a cycloid. (29] These three possible general solutions have been developed here and in Appendix B for the bar-berm equilibrium profile from separate formulations of surf zone dissipatioJL Each of these solutions is represented by type-a elliptic cycloids of a specific eccentricity but arbitrary size, (a, b). In the following section, field data will be used to resolve the cycloids that are most commonly found in beach surveys, while boundary and matching conditions will help resolve the dimensions of these cycloids. 4. Particular Solutions [3o] In this section we apply bow1dary and matching conditions to the general solutions developed in the previ- ous section to obtain particular solutions for the shorerise and bar-berm profiles that conjoin at the breakpoint, X3 (Figure 1 b ). The general solutions developed in the previous section admit to an arbitrary number of equilibriwn profiles depending on the type of elliptic cycloid and its eccentricity. Here we use field measurements to resolve the eccentricity and select the best-fit cycloid that conforms to natural beaches. A Taylor series expansion of (33) about x = 0 gives a simple analytic approximation to the general elliptic cycloid solution that is equivalent to the equilibrium profile formulations developed earlier by Dean [1977, 1991] and Inman et al. [1993]. The leading order terms of this Taylor series expansion are, h =Ax"'+ 0(£!/(l+<>)) where terms O (e:111+") are neglected and 2 2(1 -i:2) m = (2 + ~) = (3 -el) (36) (37) For the type-a cycloid, the profile factor A in (36) becomes A-3n(2a) • [ _ op] 2/(2+a) -4/J2J (38) [31] Both the profile factor A and the curvature exponent m of the classical paranietric representation in (37, 38) are functions of the eccentricity of the elliptic cycloid. How- ever, only the profile fuctor A varies with cycloid size. We use these dependencies in combination with the extensive data base on (A, m) derived from best-fits to beach profile measurements [Inman et al., 1993] to establish a criteria for the selections for a and e, that give particular solutions. 4.1. Shorerise [n] Beginning with the shorerise profile, the origin of the cycloid at P(l) in Figure 2 is positioned at mean sea level where 0= 0 811d h = h2 = 0 atx =x2 = 0. Itis apparent that the elliptic cycloid must converge on closure depth he within one-halfrevolution of the cycloid wheel, h = h2 --+ he as 0 --+ it. Hence, the size of the shorerise ellipse ax.es are given by a= a2 = hc/2 (39) This means that the closure depth formulation is decisive in achieving a particular solution to the shorerise equilibrium profile. However, the quantification of closure depth appears to be somewhat vague in the literature. (33] The general notion of closure depth he is the max- imum depth at which seasonal changes in beach profiles are measurable by field surveys, most commonly using fathom- eters [Inma11 and Bagnold, 1963]. Closure depth for sea- sonal profiles repeated over a period of a year or more is usually taken as the depth of closure of the envelope of profile changes, e.g., where the depth change vs depth 8 of 21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 4.0 3.5 a 3.0 tl:18 I 2.5 I 2.0 I 1.6 1.0 0.5 a..In1i211D1,1'111 100 160 200 250 300 I 2 4 11 I 10 12 14 111 Clomre depth,\,,. 360 400 Figure 3. Closure depth he dependence on grain siz.e D2 and incident wave height H00 for waves of 15 sec period ('I' -0.33, K0 -2.0, D0 -100 µm). decreased to a common background error [e.g., Kraus and Harikai, 1983; Inman et al., 1993). When observations ere limited to comparison of two or three surveys, the closure depth becomes the depth ai which ihe survey iines converge with depth, a point of some uncertainty [ e.g., Shepard and Inman, 1951; Nordstrom and Inman, 1973; Birkemeier, 1985). Hallermeier [1978, 1981) derived a relation for closure depth, by assuming a relationship for the energetics of sediment suspensions based on a critical value of the Froude number, giving he"" 2.28H,, -6.85(H;,/gT2) (40) where H.. is the nearshore stonn wave height that is ex:ceeded only 12 homs eaoh year and T is the associated wave period. [34) Birkemeier [1985) suggested different values of the constants in equation ( 40) and found that the simple relation h0 -1.57 H,. provided a reasonable fit to his profile measurements at Duck, North Carolina. Cowell et al. [1999) reviews the Hallermeier relation for closure depth he and limiting transport depth h, and extends the previous data worldwide to include Australia. Their calculations indicate that he ranges from 5 m (Point Mugu California) to 12 m (SE Aus!rnlia), while h, ranges from 13 m (Nether- lands) to 53 m (La Jolla, California). They conclude that disorepenoies in date ond oo.loulation procedures ma..'<e it "pointless to quibble over aocuracy of prediction" in h0 and h;. In the contex:t of planning for beach nourishment, Dean [2002) obseives that "although closure depth ..... is more of a concept than a reality, it does provide an essential basis for calculating equilibrated ... beach widths." [,s) While it may be reasonable to apply (40) or its simpler form after Birkemeier [1985] to the shorerise boundary condition (39), comparisons with the Inman et al. [1993) beaoh profile data set show that these relations tend to underestimate closure depth. We propose an alter- native closure depth relation. This relation is based on two premises: (1) closure depth is the seaward limit of non-zero net transport in the cross-shore direction; and (2) closure depth is a vortex: ripple regime in which no net granular exchange occurs from ripple to ripple. Inman [1957) gives observations of stationary vortex ripples in the field and Dingler and Inman [1976] establish a parametric relationship between dimension~ of stationary vorte]( ripples and the Shield's parameter 0 in the range 3 < 0 < 40. Using the inverse of that parametric relation to solve for the depth gives, h, = K.Hoo (Do)\'¥ sinhkh, \D1 (41) where K. and 'I' Bl'e nondimensional empirical parameters, D2 is the shorerise mediun grain size; and D0 is a reference grain size. With Ke ~ 2.0, \jl ~~ 0.33 and D0 N 100 µm, the empirical closure depths reported in Inman et al. [1993] are reproduced by (41). Figure 3 gives a contour plot calculated from (41) showing the rates at which closure depth increases with increasing wave height and decreasing grain size. Because of the wave number dependence, closure depth also increases with increasing wave period. Figure 3 is based on T-15 sec, typical of storm induced waves on exposed high-energy coastlines. [36) Using (41), the distance to closure depth X02 can be obtained from (30), L ,(2) -L f:7-·2 k=~C:it~ _-_e-_ E 2£ 2 (42) where Xc2 is measured from the origin of the shorerise located a distance X2 from the berm and a distance X3 X2 inside the breakpoint (Figure 2). This will be determined subsequently from the matching condition. It is apparent from (41) and (42) that the shorerise profile dimensions grow with increasing wave height and period, and with decreasing grain size. [,1] A family of shorerise equilibrium profiles calculated from (25)-(29) for a type-a cycloid is plotted in Figure 4a using the closure depth formulation in (41) for 15 second period incident waves. Toe aroitrary constunt in (16) that satisfies the mild slope condition in (15) is set at N -10. This particular selection has been found suitable for the narrow-shelf beaches of southern California reported in Inman et al. [1993]. Although tan ~ --+ oo as x2 -, 0, the shorerise bottom slope remains less than the angle of repose (~, :::a 33°, tan ~r :::a 0.65) at the breaker depth (hb -Hbfy), where the shorerise profile must match the bar-berm profile. In fact, shore1.i.se bottom slopes are tOD ~ ~ 0.1 for nny depth h > 1.0 m. Altogether, the family of cycloids shown in Figure 4 offer considerable diversity in the potential size and shape of the equilibrium shorerise profile. [ ,8] It is interesting to note that when cx -1 the general solution (22) reduces to that for a trochoid, a cycloid 9 of 21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES a) profllea: eccentricity and 1hear ltrell llneartty ----e=0.845;n=3 e = 0.798; n = 2 ----e=0.107;n=1 ----e=0.447;n=0 ----e = O ; n = -0.33 ; brachllltochrone IOlullon 700 700 800 500 400 300 200 Cross-Shore Distance x2, m b) slope: eccentricity and shear atreaa llneartty ----e=0.845;n=3 e = 0.798 ; n = 2 ----e=0.707;na 1 ----e=0.447;n=-0 ----e = 0 ; n = -0.33 ; brachlstochrone BOlutlon 800 500 400 300 200 Cross-Shore Distance x2, m 0 2 4 8 ~ a 8 ,c;.. 10 ! 12 14 100 0 0.1 0.08 cc. 0.08 j } 0.04 fll 0.02 0 100 0 Figure 4. Family of type-a elliptic cycloid solutions in the shoreise: (a) profile; (b) slope. Cycloids scaled for H00 = 4 m; N = 10; T = 15 sec; D2 = 100 µm; h0 = 14.4 m. C02003 produced by a rolling circle. The trochoidal solution in polar fonn after (25)-(29) has an eccentricity e = 0, and a radius r =a= b = 1/20 = h/2. This fonn of the shorerise solution is the equivalent of the brachistochrone solution, one of the first known problems in calculus of variations [Baas, 1966]. Its analytic appmximation from (36)-&38) gives the popular beach profile fom1uJation h = Ax21 after Dean [1991, 2002]. From (28) and (37) the relation between shorerise profile exponent m2 and shear stress linearity n is bottom 51.ress that decreased with increasing wave height. Jirom this we conclude that the h = Ax213 formula does not represent an equilibrium state in the shorerise portion of the shorezone, nor does any profile for which m2 > 0.571, for these would likewise require a shear stress inversion (n < 0). Accordingly, we reject any shorerise cycloid solution having an eccentricity e < 0.447. [39] To select a preferred set from the remaining cycloid solutions for the shorerise, we examine the relation between (37) and the best-fit profiles to measured shore- rises. Figure 5a gives a histogram of the shorerise profile exponent m2 derived from best fits to 51 measured profiles from nine beaches comprising the basic data set in Inman et al. [1993]. Of these 51 measured profiles, 20 were reported to represent summer equilibrium profiles, 20 4 m2 =--7+3n 4 7 or, n=---3m2 3 (43) Consequently, a profile for which m2 = 2/3 corresponds to n = -1/3, an insensible outcome since it would require a 10 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 ( eccentricity, e ] profile exponent, 111 20 a) 20 c) ~ 16 16 I,. 12 'a 8 8 1 4 4 z 0 0 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 m2 e m 1 e 10 b) (iiia;,;protl~le~fa~ctori;,7A) 10 d) 8 8 4 2 0 0 0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5 A2 A1 Flgttre S. Histograms of profile exponent m (red) and profile factor A (blue) from best fit to beach surveys [after Inman et al., 1993], corresponding eccentricity of elliptic cycloid (green). represented winter equilibrium, while 11 were said to be non-equilibrium "translational" profiles. The red histo- gram bars in Figure 5a are numbers of realizations of m2 from the best fits and the green bars are the corresponding eccentricities per (37). Only two of the 11 non-equilibrium profiles reported for the basic data set in Table I of Inman et al. [1993] produced best fits with m2 > 0.571 (outside the theoreticnl limit for shorerise equilibrium); while all 40 of the equilibrium profiles resulted in best fit m2 < 0.571. The mean shorerise profile exponent for the entire ensem- ble of 51 profiles (equilibrium and non-equilibrium alike) was m2 = 0.362, which corresponds to a shear stress linearity of n = 1.35 and a cycloid eccentricity of e = 0. 74 7. The mean m2 of the 40 equilibrium profiles is "iih = 0.365, giving n = 1.32 and e = 0.744. BRsed on these averages (3 7) suggests thnL the bottom shear stress amplitude over the shorerise typically varies :i..~ -r0 ~ u~3 in (17), coincident with the formulation of Kajiura [1968] for bottom friction due to oscillatory flow over rough beds when St= u,,./crTJ, < 30. This scale regime is consistent with the vortex ripple regime reported by Dingler and Inman [1976] when the roughness height TJs is taken to be equivnlem to the ripple heigh1. [4o] Neglecting the two anomalous translational profiles in Figure 5a (m2 > 0.571) the range of variation in the shorcrise exponent for the remaining 49 profiles was 0.21 ~ m2 ~ 0.5. AL the upper limit of this range, as m2 -0.5, the bottom shear stress varies weakly as -r O ~ u;/,3 and the shorerise cycloid has an eccentricity of e = 0.577. Al; m2 -0.4, the bottom shear stress becomes linear (n = 1), and the resulting cycloid solution for the shorerise has an eccentricity of e = 0. 707. Linear bottom shear stress -r O ~ u,., would be expected with a laminar Stokes oscillatory bound- 81)' 1.-iyer common to smaU amplitude oscillati.011s over a perfectly smooth, impenneable bed, for which c1 ,... R; 112• [Stow, 1851 ; Lm11b, 1932; Batchelar1 1970· Sleatl,, 1984]. Intermediate in the upper portion of the range of variability, where m2 ~ m2 ~ 0.4, the periodic bottom shear stress 11 of 21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 a) 4.0 a :::L 3.5 Ja.o ,2.5 · u 2.0 iu J 1.0 ) 0.15 0.0 0.4 0.8 1.2 2.0 Shandie Pmile PIOIOr A.J, ,,,OA" b) Breaker coefficient, y 0.2 0.8 1.0 1.4 1.8 4.0 [ 3.6 ~8 3.0 f 2.5 = I 2.0 1.5 I 1.0 0.5 o.o 0.4 0.8 1.2 Figure 6. (a) Shorerise profile factor A2 versus incident wave height H00 and median grain size D2 based on closure depth per Figure 3 with m1 = 0.365 and N= 10. (b) Bar-berm profile factor A1 versus incident wave heightH00 and wave breaker coefficient-y for T= 15 sec; N= 10; r = 0.76; A= 0.81; and mi= 0.4. can be characterized by a shear stress relation for smooth granular beds after Taylor [1946] that uses a drag coef- ficient fonnulation, c1 ~ R-;g115S;315, and gives a shear stress linearity of n = 6/5, corresponding to m1 = 0.377 with a cycloid eccentricity e = 0.731. In the lower potion of the range of variability below the mean, say m1 ~ 0.25, tho shear stress becomes cubic -r O ~ ~ giving a cycloid eccentricity of e = 0.845. At the lower limit m1 ~ 0.21, the non-linearity of the shear stress increases to n = 4 and the cycloid eccentricity becomes e = 0.875. Generally, over the lower portion of the range of vari- ability, 0.21 s m1 s m2, the shear stress takes on a relatively high degrees of non-linearity, typical of form drag over non-stationary rippled beds at high Reynolds numbers R. 2': 104, [Sleath, 1982, 1984], or ventilated oscillatory boundary layers over porous beds [Conley and Inman, 1994 ]. [41] From the above consideration of the empirical evidence, we conclude that the shorerise equilibrium profile is a set of elliptic cycloids whose eccentricity is limited to the range 0.447 s e s 0.875, depending on bed roughness and dynamic scale regime, with semimajor and semiminor axes specified by (39) and (41). The most common out- come has an eccentricity e = 0.744, that can be represented by an analytic J!Pproximation given by (36) with m = m2 = 0.365 and A= A2 calculated from (38) using a= 3.479. The wave height and grain size dependence of this A2 solution is shown in Figure 6a for type-a cycloids. The shorerise 12 of 21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES a) profiles, IOluUon and eccentr1clty -----.-w dllllll)llllon allerlhom1on & Gura (1983); •., 0.904 mmnum-wave dlalpatlon an. Thomlon a Gia (1983); • = 0.846 ----lln_. bonl d_,pallon after Bowen et Ill. (1968); • • 0.707 ----depth-lmltld bore dllllpatlon ., a-(1969);" = o.858 ; • = o.491 ----~aalullon;,aO -4 -2 ~ o a 4 .___,__.__~_.___,__.___,__.___,__.___,__.___,__.___,__.___,___, 6 450 400 350 300 250 200 150 100 Cross-Shore Distance, x1 , m b) elope, aolutlon and accentrlclty ----IV8fWOI ,_ dlNlpatlon aftarThomtan & Gim (1113) ; r = 0.804 fflUffl.1-dlalpallan llllar ThortDn & Qua (1983) ; • • 0.846 ----1,-boradlNlpallon lftlr 8-1 atal. (11188); • • 0.707 ----deplh,6nlacl bora dallpallon alllr Chow (1959); 11 • 0.8611: • • 0.491 ----~ aolullan ; r = 0 450 400 350 300 250 200 150 100 Cross-Shore Distance x1 , m 50 0 0.1 0.08 C0. 0.08 j 0.04 ! l;l.l D.02 D 50 0 Figure 7. Family of type-a elliptic cycloid solutions in the bar-berm: (a) profile; (b) slope. Cycloids scaled for: H00 = 4 m; T= 15 sec; N= 10; 'Y = 0.8; r = 0.76; A= 0.81. C02003 profile factor A2 increases with increasing wave height and decreasing grain size. A2 also increases with increasing wave period because its dependence on the ellipse axes a (3 8) is stronger than its dependence on 1:.. The magnitude and dynamic range of the A2 solutions in Figure 6a are in general agreement with the best-fit re.suits to beach surveys from Inman et al. [1993] shown in Figure 5b. The distri- bution in Figure 5b has a mean of 1.097 with a standard deviation of 0.457. Figure 2 is elevated above mean sea level at h1 = Z1 (Figure lb). Accordingly, the bar-berm cycloids from (35), (B4) and (BS have U1e foHowlng vertically offset polar form 4.2. Bar-Berm [42] General solutions to the bar-berm in section 3.2 and Appendix B follow from a fa-st integral m the Euler- Lagrange equation (13), and consequently admit to the addition of an arbitrary constant [Boas, 1966]. This we must do in the bar-berm because the origin at P(l) in 'l\"£ X1 ( [ -COS Q) h=h1 =-( -) -0 , 9 +Z1 =r(l -cos0) +Z1 2l/'~ -sm (44) wl1ere r is given by (26). We prescribe the vc1tical oflsol· by the maximum runup elevation from Hunt's Formula [Hunt, 1959· G111t.a a11d 711onuo11, 1985; Raubenheimer and G11za, 1996], (45) where r is the runup factor taken herein as r = 0. 76. The bar-berm cycloid must converge on the breaker 13 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 depth hh Hhl"Y within one-half revolution of the cycloid wheel, (46) Because of the vertical offset to the bar-berm profile origin, the ellipse axes must span the distance between Z1 and hh as 0 -+ 1'<:. Consequently, the breakpoint boundary condition in (46) leads to the following sizing of the bar- berm type-a cycloid ellipse axes, (47) where A is the shoaling factor from (B5). The field work of Raubenheimer et al. [1996] shows that 'Y has an observed range of variability 0.2 < 'Y < 1.6. Enforcing the breakpoint boundary condition in (47) on (44) gives the distance to the break point X3 from (30), X3 = (11b -Z1)/,(2>;;,,, ~Hoo(~+ r) /z~ (4S) e. 2AE '/ v-2 [ 43] Figure 7 gives a family of particular solutions for the bar-berm equilibrium profiles calculated from the various surf zone dissipation assumptions that produced the sepa- rate cycloid solutions in (35), (B4) and (B8). These sol- utions were applied to (44), subject to the boundary conditions in (47). The solution envelope includes cycloids produced by the linear bore (e = 0. 707), the maximum wave dissipation model (e = 0.845) and the average wave dissipation model (e = 0.904) after Thornton and Guza [1983] in Appendix B. Each dissipation model produces both type-a and type-b cycloids, but we focus our discus- sion primarily on type-a cycloids. The brachistochrone solution is included for·comp111ison, because it-is equivalent to the solution h = Ax').13 by Dean [1991, 20021 and because it is a particular case of the depth limited bore solution from (35) with TJ = 0.5. Intermediate between the brachistochrone and the linear bore solutions is another particular case of the depth-limited bore solution based on TJ = 0.659 that produces a cycloid with eccentricity of e = 0.491. These solutions are calculated for the same incident wave con- ditions (H00 = 4m, T = 15 sec) with the same arbitrary constant (N = 10) used for the shorerise in Figure 4. The shoaling factor assumed for these bar-berm solutions (A = 0.81) was based on uniform shoaling of the incident wave conditions, while a mean value was chosen for gamma ('Y = 0.8) from the data reported by Raubenheimer et al. [ 1996]. [44] Although the equilibdum profiles from the bar-be1m cycloids in Figure 7 are shorter than the shorerise cycloids in Figure 4 calculated for the same conditions, many qualitative features remain common to both, with profile length increasing with decreasing eccentricity for the type-a cycloids. Local slopes along the bar-berm profiles in Figure 7b are generally comparable to slopes found along the shorerise profiles in Figure 4b. All bar-berm cycloids in the solution envelope trend to a flat bottom (zero local slope) at the break point where matching with the shorerise is enforced. [4s] TI1e mean value ofm1 for the 51 best fits to the bar- berm profiles reported in the data set of Inman et al. [1993] is m1 = 0.411 (Figure 5c), of which only one (a translational profile) gave m1 -+ 0.66, the equivalent of the brachisto- chrone solution. The mean m1 of the 40 equilibdum profiles was m1 = 0.400, which from (37) con·esponds to a cycloid with eccentricity e = 0.707, coincident with the linear bore solution (35). The range of variability for the 40 equilibrium profiles was 0.29 :S m1 :S 0.55, corresponding to cycloids with 0.491 :S e =S 0.813. The cycloid dedved from the maximum wave dissipation formulation (e = 0.845) corre- sponds roughly to the lower limit (m1 -+ 0.29) of this statistical spread. This limit can also be obtained from the depth-limited bore solution in (35) using TJ = 1.47 to produce a cycloid with eccentricity of e = 0.813, exactly. The upper limit of the 40 equilibrium profiles (m1 -+ 0.55) can be obtained with a depth limited bore solution using TJ = 0.659. [46] The data in Figure 5c indicates that the envelope of paiticular solutions for the bar-berm are limited to a set of type-a and type-b cycloids of the form in (44) derived from the depth-limited bore (0.5 :S TJ ::; 1.47) and maximum wave dissipation formulations from (35) and (B8), respec- tively. Within this envelope of solutions, the most commonly occurdng is the linear bore (e = 0. 707) that can be approximated by (36) with m = m1 = 0.400 and A = A1 calculated from (38) using ex= 3.0. Figure 6b shows the dependence of this A I solution on incident wave height and the gamma ("() factor over the observational range of vadability reported in Raubenheimer et al. [1996]. This solution is based on the same wave pedod (T= 15 sec) and scaling assumption (N = 10) as the shorerise solution for A2 in Figure 6a. Runup and shoaling factor asswnptions (I' = 0.76; A = 0.81) are the same as those chosen for the bar- berm cycloids in Figure 7. The resulting magnitude a11d dynamic range of the A I solutions in Figure 6b are consis- tent with the best-fit results to beach surveys from Inman et al. [1993] shown in Figure 5d. The mean of these field results was A1 = 0.868 with a standard deviation of 0.386. The bar-berm profile factor A I increases with increasing wave height and decreasing 'Y. Unlike the shoredse, the ellipse axes a and bin (47) are independent of frequency in the bar-berm. Consequently A1 decreases with increasing wave period due to the frequency dependence of thee factor in (38). Therefore the mean bar-berm slopes become flatter with longer period waves. [47] The bar-berm A1 solution in Figure 6b shows very similar compound variation with wave height and 'Y as the shorerise A2 solution. A relation arises between 'Y and beach grain size through the matching condition discussed below. The basis for this relation is a finding in Raubenheimer et al. [1996] who show that 'Y increases with increasing bottom slope at the break point, tan I¾, (49) where K,, K1 are empirical constants taken as K0 = 0.2 and K1 = 6.0. The bottom slope at the break point, tan flx, is given by the slope of the shorerise profile at x2 = X3 - 2X2• 4.3. Matching the Shorerise and Bar-Berm [4s] To complete the particular solutions, we must match the bar-berm and shorerise profiles at the break point 14 of 21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 (Figure 1 b ). The point of conjoinment is the seaward end of the bar-berm profile at the break point, a distance X3 from the origin (48). Here, the shorerise and bar-berm solutions must both equal the breaker depth, (50) at x2 = X3 -X2 and x1 = X3, respectively. To obtain this match, the origin of the shorerise profile must be located a distance X3 -X2 landward of x1 = X3• fu terms of the polar form of the shorerise cycloid, this distance is, 2r2J(l) X1 -Xi= ---;:--(Bi, -sin Bi,) (51) Bi,= arccos[l -2(~7,.) ~] where r2 is given by (26) and evaluated at 0 = 0i,, and ~ is related to shorerise cycloid eccentricity by (28). With (51) the shorelise bottom slope at the breakpoint can be solved, sinflt, +e2(cos8b -l)sin9t,cos~ tanf3x, = I -cosl\+~(siuBt,-O.,)sin!li,tosOi, (52) Because Iii, is grain size dependent through the closure depth relation in (41), 'Y in (49) also becomes grain size dependent. This in tum makes the bar-berm profile sensitive to grain size variability through (47) or by its alternative analytic approximation (36)-(38). In terms of the analytic approximation, the matching condition (50) is satisfied by, X2=X1-1 1 "" [ A X"" -rn /h] 11"" Ai giving the bottom slope at the break point as, tan~.lz =m2A2(Xi-Xir'-1 (53) (54) where X3 is given by (48). Point-by-point comparisons of type-a and type-b cycloids show that type-a provided the best fits to measured beach profiles. 5. Discussion (49] The cycloid solutions were derived from 2-dimen- sional representations of the shorezone as a consequence of the assumption that external work is done on the system by a train of uniform, normally incident waves. Under this assumption, viscous dissipation was maximized along a boundary surface within the system, in the cross-shore normal plane (Figure la). However, on natural beaches, waves are seldom uniform and normally incident, resultu,g in nearshore circulation cells with longshore currents. Near- shore circulation cells are entirely compatible with equilib- rium conditions defmed by (8) and (11) because they provide a mechanism to ventilate latent heat into the surroundings. There is no violation of thermodynamic equilibrium principles through partitioning heat and work evolution between the cross-shore and longshore shear stress components, respectively. By this paititioning, the cross-shore directed shear stresses perform the dissipative work that satisfies the maximum entrnpy production for- mulation on the second law (11 ), while the longshore directed radiation stresses remain in balance with the long- shore directed shear stresses driving the nearshore circula- tion. The only consequence of such partitioning is that the rate of working by longshore directed stresses can not alter the thermodynamic coordinate ~ that defines the cross-shore profile. [so] The cycloid solutions result in vanishing bottom slope at the seaward limit of both the shorerise and bar- berm. It could be argued that field survey techniques are not s11fficiently accllfate to determine whether this is a realistic outcome, especially since the flattening at the end of the cycloid trace is highly localized. Figures 4b and 7b show that cycloid slopes depart from zero at small distances inshore of the seaward limit of both the shorerise and bar- berm. The seaward boundary of the shorezone system at closure depth excludes any dynamics associated with the shelf slope from entering the equilibrium formulation of the shorerise. The slope discontinuity formed at the motch point between the shorerise and bar-berm cycloids produce a reasonable representation of a breakpoint bar. The validity of the foregoing comments and resulting equations for equilibrium beach profiles are supported by the goodness of fit in the following detailed comparisons. [s1] The following point-wise comparisons between par- ticular shorerise and bar-berm solutions from thermody- =ic theory with profiles measured by the U. S. Army Corps of Engineers are made to determine the predictive skill of cycloid solutions. The profiles were measured at beach survey range PNl 180, near Oceanside, CA during a six year period between March 1981 and September 1987 (USACE, 1985, 1991]. This range and setting are described in Inman et al. (1993]. [s2] Wave climate was measured in 6 hour intervals by the pressure sensor array located off Oceanside, CA [ CDJP, 19 80-198 8]. These directional wave data were back refracted into deep water from their measurement location and forward refracted to 10 m depth in the neighborhood of the PNl 180 beach range to correct for local shelf and island sheltering effects [O'Reilly and Guza, 1991, 1993]. Under the hypothesis that equilibrium is dete1mined by the persis- tent large waves, the refracted wave time series were filtered for the highest 5% waves and then time-averaged over each survey period to provided the forcing history of the H=, T, needed to calculate temporal variability of the boundary conditions on each profile. Grain size data from Inman (1953], Inman and Rusnak (1956] and Inman and Masters (1991] were used in these calculations. All empirical parameters used to quantify boundary conditions were set according to those used in Figures 3, 4, and 7. The horizontal location of the bar-berm origin relative to the survey benchmark X1 (Figure 1 b) was taken directly from the database of each survey. [ 53] A procedure for achieving a high resolution fit of the cycloid solutions to field data is detailed in Appendix C. A point-by-point comparison between tl1ese high resolution solutions and the six measured beach profiles surveyed at PNl 180 is shown in Figure 8 for the type-a cycloids. The legend of each panel gives the average significant wave 15 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 a) Surw, Ran .. PN 1111 lllroh 1N1 ,._1911 prullle; H. = 22 m; T= 18 NC ---type-a cydold; "., 0.71M ---type.a cydold; er= 0.702 2 4 8 8 10 '---------'---'---'------'-_.__...____...._____._ _ __,.___,,2 1000 800 800 400 200 b) lurw, Ran .. PN 11IO JulJ 1112 ffllMUred pRJllle; H_ = 2A m; r .. 13,5 NC ---type.a cydold; .... 0.1101 ---type.a cydold; • = 0.855 .... ,.,..- ,,/ ,, ,,, ,/ / ,' i'i' ' , /,' 0 -2 0 2 4 8 8 10 '---------'---'---'------'--'---'--'------........ --'---'12 1000 800 800 400 200 11) Burwy Ran .. PN 1tl0 Apll 1- m-.ired praftle· H = 2.25 m· T" 15 eac ---type-e~;e=0.751 ---type-a cydold; e • 0.743 0 4 8 8 10 "'--___._ _ __,.__.._____. _ _.__.....___._____.__......___.12 1000 800 800 400 200 0 ~red prullle; H." 1.5 m; T= 18 NC ---type.a cydold; "= 0.71M ---type.a cydold; • • 0.788 .,, ..... ,r' / ,;'' , / ,, I -2 I , 0 /' ,, 2 4 8 8 10 ._____.__.......__..________.. _ _.__...____..___.__.....____,12 1000 800 800 400 200 a) lurw, ~ PN t1IO Aprll 1N1 ~ prollle; n. • 2.8 m; T• 13 NC type,e cydold; • = 0.788 ---type,e cyckJld; • = 0.700 0 10 ._____._ _ __,.__.,_____. _ _.__....._____...._____._ _ __,.___,12 1000 800 800 400 200 meaeured prullle;H. • 1.7m; T• 18.5 NC ---type.a c:yclold; • ., 0.846 ---type.a c:yclold; "= 0.758 0 2 4 8 8 10 ..._____.. _ _.__..._____,'-----_.__.....__._____._ _ __,.___. 12 1000 800 800 400 200 0 Cross-Shore Distance x, m Fignre 8. Comparison of beach profile survey data (gray) from Oceanside, CA, versus the elliptic cycloid equilibrium solutions for the shorerise (blue dashed) and bar-benn (red dashed). Calculated profiles based on CDIP wave monitoring data and median grain size Dz= 100 µm withN = 10, and 'Y = 0.8. height and period of the highest 5% waves that occurred between successive survey periods. The type-a cycloids (Figure 8) gave a predictive skill factor of R = 0.83 to 0.95, where R is adapted from Gallagher et al. [1998] and applied to the mean-square error in depth. The eccentricity of the type-a shorerise cycloids varied over a relatively narrow range 0.655 ~ e ~ 0.766 with a mean ofe = 0.721, generally consistent with the shear stress fonnulations for smooth to moderately rough beds after Kajiura [1968] and Taylor [1946], and in overall agreement with the larger ensemble statistics that were inferred from the analytic approximation in Figure 5a. The type-a bar-berm cycloids displayed a somewhat wider range of variability in eccen- tricity, 0.704 ~ e ~ 0.901 with a mean of e = 0.785. The range spanned all theoretical possibilities between the depth limited bore fonnulation (35) and the average wave 16 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 Sarvq Ranpa PN 1180 ud PN 1MO, Ocnndde, CA -4 meuun:d profiles (12 ea, 1950-1987) -2 ---cycloid; H00= 3.0 m; T= 14 sec 0 ---cyc;loid; H00= 2.0 m; T= 15 sec ---cyc;loid; H00-1.0 m; T-15 sec 8 10 1000 600 600 400 200 0 CTOM-Shore Distaneo .r, m Figure 9. Envelope of variability of 12 measured bench J>IOfilos {1950-1987) at Oceanside CA (gray) compared with the ensemble of type-a elliptic cycloid solulions (colored) for solcct.ed incident. wave heights (llld periods with D2 = 100 µm, N = 10, 'Y = 0.8, (llld A= 0.81. dissipation fonnulation (B4), while the mean was still within the main peak of the large ensemble distribution found in Figure 5c. Therefore, the point-by-point accmacy of the type-a cycloid solution is not only good, but was achieved with ellipse parameters that are compatible with both process-based and empirically-based computations of those parameters. [s4] If tho ]>rocoduro in Appendix C is abbreviated ot the first step and the cycloids are simply computed using the mean of the empirical distributions in Figure 5 (e = 0.744 in the shorerise and e = 0.707 in the bar-berm), then t..1ie predictive skill of the type-a cycloid solutions is less but still acceptable for most engineering applications, ranging from R = 0.76 to R = 0.88. [55] The type-b cycloids (not shown) produced a skill factor generally less than 0.5. The type-b cycloid solutions perfonned poorly on a point-by-point basis at PNl 180 beca11se Chey give a broad bench or trough and bar for eccentricities e ~ 0.65.This feature is not generally found on Jlall'OW shelf beaches such as those along the southern California coast. However, on wide shelf beaches such as the U. S. Gulf Coast, this feature may improve the predic- tive skill of the type-b cycloid. [56] When a series of type-a cycloid solutions for a broad range of wave heights are overlaid 011 an ensemble of many beach profile measurements, a well defined envelope of variabi.lity becomes apparent as illusmned in Figure 9. This figure combines 12 profiles meiism·ed over a 37 year period from two adjacent beaches near Ocean- side that have geomorphic equivalence. Such comparisons with the cycloid solutions suggest that the volume of sand associated with long term beach profile variations are directly calculable by integration of the cycloid solution between the limits of wave climate vaiiation for a partic- ular site. Irt this exnmple, the wa.vo height integrated cydoids indicate that 1,180 m3 of sand per merer of shoreline are involved with long tenn beach variability. At lowest order, this volume represents the minimum sand volume that the beach must retain in the long te1m in order for it to maintain a sus1oinable equilibrium ,vith clu111ging seasons. 6. Conclusions [ 57] Solutions for equilibrium beach profiles are obtained from the maximum entropy production formulation of the second law ofthcnnod;,Twinics, when applied to a..vi isotJ1er- mal shorezone system of constant volume and constant wave forcing. The entropy of the shorezone system remains constant while that of the surroundings is increased. The equilibrium profile was found to be the shape that max- imizes the discharge of entropy into the surronndings by maximizing the dissipation rate within the shorezone system. This formulation of thennodynrunic equilibrium, expressed in shorezone coordinates (Figures la and lb), provides exact solutions for the shorerise and bar-be1m portions of the c011joined beach profile. The shorerise, is an outer zone of wave shoaling where dissipation arises from the rate of working by bottom shear stresses. The bar-betm inner zone nnderlies the zone of broken waves where turbulent dissipation results in a progressive loss of bore height with decreasing water depth, ultimately pro- duci11g runup over the beach face and fonnation of tl1e berm crest. It is notable that thennodynamic solutions for these zones of widely differing mechanics provide one of the rare instances of a geophysical problem having an exact solution. [58] The thermodynamic solution to the equilibrium beach problem belongs to a class of analytic functions known as elliptic cycloids, fonned by the trajectory of a point 011 the perimeter of a rolling ellipse [Boas, 1966]. The general shape of the equilibrium profile is determined by the 17 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 eccentricity and starting position of the ellipse (Figure 2). Two types of cycloids are produced, resulting in two distinctly different classes of profile shape. Type-a cycloids are formed when the rolling ellipse starts with the semi- major axis (a) aligned with the vertical axis of the coordi- nate system; while type-b cycloids start with the semiminor axis (b) aligned with the vertical axis. While equilibrium profile solutions were developed for both types of cycloids, only the type-a cycloids were found to be well correlated with beach profile measurements from southern California that typically do not exhibit complex or large-amplitude bar formations. [ 59] The aspect of profile shape controlled by eccentricity is related to the degree of non-linearity of the bottom shear stress in the shorerise, and to the rate of bore height decay in the bar-berm. As the shear stress in the shorerise becomes increasingly non-linear, the eccentricity of the ellipse increases, and the profile develops more variability in curvature and bottom slope (Figure 4 ). The shear stress formulation of Kajiura [1968] for oscillatory flow over rough beds was found to give the best general average of measured shorerise profile curvature for fine sand beaches. In the bar-be1m, the profile eccentricity increases with increasing rate of bore height decay with depth. The linear bore after Bowen et al. [1968] and the maximum wave dissipation formulation of Thornton and Guza [1983] produced bar-berm solutions with the most commonly observed curvature. [60] The overall dimensions of the equilibrium beach profile are governed by the dimensions of the semimajor and semiminor axes (a and b) of the cycloid ellipse (Figure 2). These dimensions are resolved from boundary conditions that require the cycloids converge with closure depth at the end of the shorerise profile, and converge with the breaker depth at the end of the bar-berm profile. The formulations used to specify these two depth limits allow a variety of transport physics to be incorporated in the particular solutions. We have demonstrated the viability of a new closure depth formulation that gives the shorerise solution wave height, period and grain size dependence and resulted in a high degree of predictive skill in simulating measured profiles (Figure 8). We have shown generally that profile dimensions grow with increasing wave height and period, but decrease with increasing grain size. The grain size dependence of the shorerise is imprinted on the bar-berm through the matching condition and the entropy associated witl1 both increases with decreasing grain size as S ~ D-3• These solutions are easily integrated over any given range of seasonal wave height to estimate the minimum volume of sand that the beach must retain in order to maintain a sustainable equilibrium for a given wave climate (Figure 9). [ 61] The elliptic cycloid solution allows all the signifi- cant features of the equilibrium profile to be characterized by the eccentricity and the size of one of the two ellipse axes. This general formulation is especially versatile because these two basic ellipse parameters can be related to either process-based algorithms or empirically based parameters for which an extensive literature already exists. This feature gives the cycloid solution a modular charac- ter, allowing it to be continuously upgraded by new science, or modified by alternative formulations according to user preference. Appendix A: Shorezone Entropy [ 62] In its statistical mechanics formulation, entropy is shown to be a measure of the disorder or randomness of a system [Fermi, 1936]. From Halliday and Resnick [1967] and Landau and Lifshitz [1980] we write: S =KBln(3) (AI) where K8 = 1.38 x 10-23 J/°K is the Boltzmann constant and S is a disorder parameter that gives the probability that the system will exist in a particular state relative to all possible states. [ 63] The probability of finding a single grain of sand in the systems is: where J is a constant defined in (A5) that depends on the geometty of the system boundaries. The number of sand grains in the system is (A3) where N0 ~ 0.6 is the volume concentration equal to I-porosity, and D is a representative diameter of the mobile sand in the shorezone system of volume v. (Figure 1). When ~ is applied to the bar-bean and shorerise portions of Vs, D is a representative median diameter. lnserting the probability state function (A2) into (Al) and collecting te1ms from (A3) gives the beach entropy: where the system boundary scale factor is J=f~ Xe (AS) The integral in (A5) defines the boundary dependent constant in (A2). With equation (A4), the equilibrium state of the beach may be written in tenns of a state integral. Beach entropy increases with decrnasing grain size because greater numbers of grains permit larger numbers of possible states for the system. For the same reason, beach entropy increases with increasing size of the shorezone system. Appendix B: Bar-Berm Solutions From Surfzone Dlssipation [64] The first of two empirically derived dissipation functions developed by Thomton and Guza [1983] is based on the hypothesis that waves break in proportion to the distribution of all waves, yielding a surf zone dissipation per unit area Eb of: (--) 8 EC K.,. Eb= T·U =-;,-g=p </Xi '1 (Bl) 18 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 where K -3J7i 3 7 ave-16-y"pgB crH00 The dimensionless functional in (13) recovered from order-1 terms in (B6) is FIJl.b!'.) = ft-3 /1 + (E.t.'/ (B7) '.111-d ~ ~ !he 0(1 breaker lru:tor. We select from (1) an mfimtesunal arc length of the bnr-berm piofile having the With (B7) the first integration of (13) gives form, which gives 1:1te following form of the integral in (11) that we make stationary to maximize entropy prudul-tiun, From (B2) the dimensionless fhnctiona! in t.11e Euler- Lagrange equation (10) becomes F(ll,~,i_) = 11-5 VI+ (q_)2 (B3) Inserting (B3) in (13) we get oFl&x = 0 and the first integration of (13) gives -' d ,,,o l + dh 1 -if ,,10 --1- (B4) where £,11 is the dimensionless form of the integration constant. The bar-berm profile given by (B4) is equivalent ~o (22) when ex "' IO and integra,es to an ell~cycloid as m (23) ru1d ~29) witJi an eccemricity e = y19/1 I "'0.904. The d1menst0nal form of the integration constant is c; = !ti!L10 = (2a)-10 for tl1c type-a bar-benn cycloid and c( = (2b)-10for U1c type-b cycloid. [6s] T~e second form of the dissipation function put forward m Thornton and Guza [1983] is based on the hypothesis that waves break in proportion to the distribution of the largest waves, giving, where Kmax = :~B3pgcrH!,{I -I s/z } (! -i:ti.2) A. = 22/s H';J,s ( cr2 / FrY) i;s and A. is the shoaling factor relating breaker height to incident wav_e height A =. HoclHb, fur a shoaling Airy wave. By a smnlar formulation to (B2), the integral in (11) that we make stationary to maximize entropy production is (B8) where !f{ is a dimensionless integration constant. The bar-berm profile given by (B8) is equivalent to (22) when ex = 6, ru1d gives an elliplic cycloid wilh an eccentricity e = -[ffi -0.845 according to the general Cruiesiw1 relations in (22). The dimensional fonn of the integration constant is c'; = c'{IL6 = (2a)-6 for the type-a har-bel'.Dl cycloid and c'( = (2b -6 for tJ1e type-b cycloid. Appendix C: Matching Field Data to Cycloid Solutions [66] The cycloid solutions utilize boundary conditions based on several empirical parameters. We fix these param- eters iliroughout the iterative procedure described below ?Ccor'!ing to characteristic values assigned by previous field 111vest1gators. These parameters include: the Airy wave mild slope factor N after Mei [1989] set atN= 10 in (16); closure depth coefficients in (41) taken as 1.(,, = 2.0, D0 = 100 µm, 'I'= 0.33 after Dingler and Inman [1976); the runup factor r = 0.76 used in (45) based on Hunt's Formula [Hunt, 1959; Guza and 111ornlon, 1985; Raubenheimer and Guza, 1996]; and the slope factors set at K0 = 0.2, K1 = 6.0 in the breaking wave criteria for 'Y in (49) after Raubenheimer et al. [1996). These values were used to calculate Figures 3, 4, and 7. The shoaling factor A was calculated from the time-averaged wave climate data (Hoo, 1) over each survey period based on shoaling Airy wave theory as written in (B5). [ 67] In addition, the best fit cycloid solutions were calculated by the following series of iterative steps: [ 68] 1. Shorerise profiles were calculated from (25), (26), (29), (39), and (41) using an initial assumption of a cycloid eccentJ:icity of e = 0.744, per the Kajiura [1968) shear stress fonnulation; [69] 2. Bar-berm profiles were calculated from (25), (26), (44), and (47) using initial assumptions that 'Y = 0.8 and e = 0.707, per the linear bore dissipation f01mulation; [10] 3. The initial profiles from steps 1 and 2 are matched by the X3 -X2 shift of the shorerise profile calculated from (51); [11] 4. The bottom slope at the breakpoint is calculated from (52) and used to correct the initial assumption of 'Y by means of (49); [_n] 5. The bar-berm profile is re-calculated per step 2 usmg the corrected value of -y; [73] 6. Repeat steps 1-5 making iterative adjustments to the eccentricity to minimize the mean squared error between the cycloid solutions and the measured profile. (B6) [ 74] Ackoowledgmeoh. This study was earned out as a collaborative research project under contract with the Kavli Institute, Santa Barbara, 19 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFILES C02003 California, with additional support provided by the Office of Naval Researoh (code 321 CG and code 321 OE). The authors appreciate discussions with John Miles, Rick Salmon, Meryl Hendershott, Robert Guza, and Falk Feddersen on principles of thermodynamics and shallow water wave theory. We acknowledge Joseph Wasyl for contributions in computer programming and graphics. We also thank the reviewers and Associate Editor for comments and guidance that helped us to clarify the thermodynamic formulation of the problem and its practical purview. References Anderson, G. M. (1996), Thermodynamics of Natural Systems, John Wiley, Hoboken, N. J. Anderson, 0. L. ( 1995), Equations of State of Solids for Geophysics and Ceramic Science, 405 pp., Oxford Univ. Press, New York Bagnold, R. A. (1947), Sand movement by waves: Some small-scale experiments with sand of very low density, J. Institution Eng. Civil Eng. Div., 27(5554), 447-469. Bagnold, R. A. (1956), The flow of cohesionless grains in fluids, Philos. '/rans. R. Soc. London Se,: A, 249(964), 235-297. Bagnold, R. A. (1963), Mechanics of marine sedimentation, in The Sea, Ideas and Observations on Progress in the Study oft/ie Seas, vol. 3, The Earth Beneath the Sea, History, edited by M. N. Hill, pp. 507-528, Jolin WLley, Hoboken, N. J. Bailard, J. A., and D. L. Inman (1981), An energetics bedload model for a plane sloping beach: Local transport, J. Geophys. Res., 86(C3) 2035- 2043. ' Bane, J. M. (1995), The California coastal marine layer: Winds and thermo- dynamics: Technical and data repor1: June 1994 nircmfl mc=ment program, Tech. Rep. CMS-95-1, 289 pp., Univ. of North Carolina at Chapel Hill, Chapel Hill. Batchelor, G. K. (1970), An Introduction to Fluid Mechanics, Cambridge Univ. Press, New York. Birlcemeier; W. A. (1985), Field data on seaward limit of proftle change, J. Waterw. Port Coastal Ocean Div., IJ/(3), 598-602. Boas, M. L. (1966), Mathematical Methods in the Physical Sciences, 778 pp., John Wiley, Hoboken, N. J. Bohren, C. F., and B. A. Albrecht (Eds.) (1998), Atmospheric Thermody- namics, 402 pp., Oxford Univ. Press, New York. Bowen, A. J. (1980), Simple models of nearshore sedimentation, beach profiles and Longshore bars, in The Coastline of Canada: Littoral Processes and Shore Morphology, Pap. 80-10, edited by S. B. McCann, pp. I -11, Geo!. Surv. of Can., Ottawa, Ont. Bowen, A. J., D. L. hunan, and V. P. Simmons (1968), Wave 'set-down' and set-up, J. Geophys. Res., 73(8), 2569-2577. Bruun, P. (1954), Coast erosion and the development of beach proftles, Tech. Memo. 44, U.S. Army Corps of Eng., Beach Erosion Board, June. Casas-Vazquez, J., and D. Jou (Eds.) (1991), Rheological modelling: Ther- modynamical and statistical approaches, in Proceedings of the Meeting held at the Bellaterra School of Thermodynamics, Autonomous Univer- sity of Barcelona, Sant Feliu de Gulae11te;xo/s, Catalonia, Spain, 378 pp., Springer, New York. CDIP (1980-1988), Coastal Data Information Program, Summary Rep. 50-150, U.S. Army Corps of Eng., Calif. Dep. of Boating and Waterw. Chandrasekhar, S. (1961), Hydrodynamic and Hydromagnetic Stability, 652 pp., Clarendon, Oxford, U.K. Chow, V. T. (1959), Open-Channel Hydraulics, 680 pp., McGraw-Hill, New York. Conley, D. C., and D. L. Inman (1994), Ventilated oscillatory boundary layers, J. Fluid Mech., 273, 261-284. Cowel~ P. J., D. J. Hanslow, and J. F. Meleo (1999), The shoreface, in Handbook of Beach a11d Shore/ace Mo,phodynamics, edited by A. D. Short, pp. 39-71, John Wiley, Hoboken, N. J. Cox, R. A., and N. D. Smith (1959), The specific heat of sea water, Proc. R. Soc., A, 252, 51-62. Dean, R. G. (1977), Equilibrium beach profiles: U.S. Atlantic and Gulf Coasts, Ocean Eng. Rep. 12, 45 pp., Univ. of Delaware, Newark, N. J. Dean, R. G. (1991), Equilibrium beach proftles -characteristics and appli- cations, J. Coastal Res., 7(1), 53-84. Dean, R. G. (2002), Beach Nourishment: Theory and Practice, Advanced Series on Ocean EnginueJ'lng, vol. I 8, 399 pp., World Sci., River Edge, N. J. de Groot, S: R., and P. Mazur (1984), Non-Equilibrium Thermodynamics, Dover, Mineola, New York. Dewar, R. (2003), Information theory explanation of the fluctuation theo- rem, maximum entropy production and self-organized criticality in non- equilibriwn stationary states, J. Phys. A Math. Gen., 36(3), 631-641. Dingler, J. R., and D. L. Inman (1976), Wave-formed ripples in nearshore sands, paper presented at 15th Coastal Engineering Conference, Am. Soc. of Civ. Eng., Honolulu, Hawaii. Eckar1, C. (1962), The equations of motion of seawater, in The Sea, Ideas and Observations on Progress in the Study of the Seas, vol. 1, Physical Oceanography, edited by M. N. Hill, pp. 31-41, John Wiley, Hoboken, N. J. Fermi, E. (1936), Thennodynamics, Dover, Mineola, New York. Gallagher, E. L., S. Elgar, and R. T. Guza (1998), Observations of sand bar evolution on a natural beach, J. Geophys. Res., 103(C2), 3203-3215. Gradshteyn, I. S., and I. M. Ryzhik (1980), Table ofintegra/s, Serles, and Products, Elsevier, New York. Grinfeld, M. (Ed.) (1991), Thermodynamic Methods in the Theory of Heterogeneous Systems, 399 pp., John Wiley, Hoboken, N. J. Guza, R. T., and E. B. Thornton (1985), Velocity moments in nearshore, J. Waterw. Port Coastal Ocean Eng., 111(2), 235-256. Hallermcier, R. J. (1978), Uses for a calculated limit depth to beach erosion, paper presented at I 6th Coastal Engineering Conference, Am. Soc. of Civ. Eng., Hamburg, Germany. Hallermeier, R. J. (1981), A profile zonation for seasonal sand beaches from wave climate, Coastal Eng., 4, 253-277. Halliday, D., and R. Resnick (1967), Physics for Student,· of Science and E11gineering, Parts I and II, John WLley, Hoboken, N. J. Henderson F. M. (1966), Open Channel Flow, 522 pp., Macmillan, New York. Hodgman, C. D. (1947), Standard Mathematical Tables, 358 pp., Chemical Rubber, Cleveland, Ohio. Hunt, I. A., Jr. (1959), Design of seawalls and breakwaters, J. Waterw. Harbors Coastal Eng. Di:v., 85, 123-152. Imawak~ S., P. P. Niiler, J. D. llleman, W. G. Largo, and T. K. Chereskin (1989), Trajectories of longline drifters in the eastern tropical Pacific in 1983-1986, Ref. Ser. 92-05, 27 pp., Univ. of Calif., San Diego, Scripps Inst. of Oceanogr., La Jolla, Calif. Inman, D. L. (1953), Areal and seasonal variations in beach andnearshore sediments at La Jolla, California, Tech. Memo. 39, 234 pp., U.S. Army Corps of Eng., Beach Erosion Board. Inman, D. L. (1957), Wave-generated ripples in nearshore sands, Tech. .M.mo. JOO, 65 pp., U.S. Army Corps afBng., .Boaoh Bro!lon Board. lnrunn, D. L. (1960), Shore processes, in McGnnv-Hfl/ Eneydopedia of Science and Technology, vol. 12, pp. 299-306, McGraw-Hi!~ New York. Inman, D. L., and R. A. Bagnold (1963), Littoral processes, in The Sea, Ideas and Observations on Progress in the Study of the Soos, vol. 3, The Earth Beneath the Sea, edited by M. N. Hill, pp. 529-553, John Wiley, Hoboken, N. J. Inman, D. L., and J. H. Filloux (1960), Beach cycles related to tide and local wind wave regime, J. Geo[., 8(2), 225-231. Inman, D. L., and P. M. Masters (1991), Budget of sediment and prediction of the future state of the coast, in State of the Coast Report, San Diego Region, Coast of Califomia Storm and Tidal Waves Study, chap. 9, 105 pp., U.S. Army Corps of Eng., Los Angeles District, Calif. Inman, D. L., and G. A. Rusnak (1956), Changes in sand level on the beach and shelf at La Jolla, California, Tech. Memo. 82, 64 pp., U.S. Army Corps of Eng., Beach Erosion Board. Inman, D. L., R. J. Tait, and C. B. Nordstrom (1971), Mixing in the surf zone, J. Geophys. Res., 76(15), 3493-3514. Inman, D. L., M. H. S. Elwany, and S. A. Jenkins (1993), Shorerise and bar-berm profiles on ocean beaches, J. Geophys. Res., 98(CIO), 18,181-18,199. Kajiura, K. (1968), A model of the bottom boundary layer in water waves, Bull. Earthquake Res. Inst., 46, 75-123. Keulegan, G. H., and L. H. Carpenter (1958), Forces on cylinders and plates in an oscillating fluid, J. Res. Natl. Bur. Stand., 60(5), 423 -440. Komar, P. D., and D. L, Inman (J 970), Longshore sand transport on beaches, J. Geophys. Res., 75(30), 5914-5927. Kraus, N. C., and S. Harikai (1983), Numerical-model of the shoreline change at Oarai Beach, Coastal Eng., 7(1), 1-28. Lamb, H. (1932), Hydrodynamics, Dover, Mineola, New York. Landau, L. D., and E. M. Lifshitz (1980), Statistical Physics, Part I, 3rd ed., 544 pp., Pergamon, Tarrytown, New York. Mei, C. C. (1989), The Applied Dynamics of Ocean Surface Waves, 740 pp., World Sci., Teaneck, N. J. Moore, W. J. (1962), Physical Chemistry, 977 pp., Prentice-Hal~ Engle- wood Cliffs, N. J. Newton, R. C., A. Navrotsky, and B. J. Wood (Eds.) (1981), Thermody- namics of Minerals and Melts, 304 pp., Springer, New York. Nordstrom, C. E., and D. L. hunan (1973), Beach and cliff erosion in San Diego County, California, in Studies on the Geology and Geologic Hazards of the Greater San Diego Area, Califomia, edited by A. Ross and R. J. Dowlen, pp. 125-131, San Diego Assoc. of Geol., San Diego, Calif. Nordstrom, C. E., and D. L. Inman (1975), Sand level changes on Torrey Pines Beach, California, Misc. Pap. 11-75, 166 pp., U.S. Army Corps of Eng., Coastal Eng. Res. Cent. 20 of21 C02003 JENKINS AND INMAN: THERMODYNAMIC SOLUTIONS FOR BEACH PROFrLES C02003 O'Reilly, W. C., and R. T. Guza (1991), Comparison of spectral refuu:1ion and refiaction-difliaction wave models, J. Wate,w. Port Coastal Ocean Eng., 117(3), 199-215. O'Reilly, W. C., and R. T. G= (1993), A comparison of2 spectral wave models in the Southern California Bight, Coastal Eng., 19(3-4), 263- 282. Ozawa, H., S. Shimokawa, and H. Sakuma (2001), Thermodynamics and fluid turbulence: A unified approach to the maximum transport proper- ties, Phys. Rev., E(64), 026303, doi:IO.II03/PhysRevE.64.026303. Ozawa, H., A. Ohmura, R. D. Lorenz, and T. Pujol (2003), The second law of thermodynamics and the global climate system: A review of the maximum entropy production principle, Rev. Geophys., 41(4), I018, doi: l 0.1029/2002RGOOO 113. Paltridge, G. W. (1975), Global dynamics and climate-system of minimum entropy exchange, Q. J. R. Meteorol. Soc., 101(429), 475-484. Paltridgo, G. W. (1978), S1endy-stn10 format of global climate, Q. J. R. Mi!rcorol. Soc., /04(442), 927-945. Raubenheimer, B., and R. T. Guza (1996), Observations and predictions of run-up, J. Geophys. Res., 101(11), 25,575-25,587. Raubenheimer, B., R. T. Guza, and S. Elgar (1996), Wave 1nlllsfonnation across the inner surf zone, J. Geophys. Res., 101(11), 25,589-25,597. Salje, E. K. H. (&I.) (1988), Ph)'s/C/JI Propl!Tili,,r a11d Thermodynamic Behaviour of Mil1erols, 707 pp" Springer, New· York. Sabnon, R. ( 1998), Lectures on Geophysical Fluid Dynamics, 3 78 pp., Oxford Univ. Press, New Yorlc. Shepard, F. P., and D. L. hunan (1951), Sand movement on the shallow inter-canyon shelf at La Jolla, California, Techn. Memo. 26, 29 pp., U.S. Army Corps of Eng., Beach Erosion Board. Shepard, F. P., and E. C. La Fond (1940), Sand movements along the Scripps Institu1ion Pier, Am. J. Sci., 238, 272-285. Short, A. D. (Ed.) (1999), Handbook of Beach and Shorejace Morphody- 11amics, 379 pp,, John Wiley, Hoboken, N. J. Sleeth, J. F. A. (1982), The suspension ofsand by waves, J. Hydraul. Res., 20, 439-452. Sleath, J. F. A. (1984), Sea Bed Mechanics, 335 pp., John Wiley, Hoboken, N. J. Stokes, G. G. (1851), On the effec1 of internal friction of fluids on the mo1ion of pendulums, Cambridge Phllos. Soc. './rans., 9, 8-106. S1onns, J. E. A., G. J. Weltje, J. J. Van Dijke, C. R. Geel, and S. B. Kroonenberg (2002), Process-response modeling of wave-dominated coastal systems: Simulating evolution and stratigraphy on geological 1imescales, J. Sed. Res., 72(2), 226-239. Sverdrup, H. U., M. W. Johnson, and R. H. Fleming (1942), The Oceans, Their Physics, Chemistry, and Ge11eral Biology, 1087 pp., Prentice-Hall, NewYorlc. Taylor, G. I. (1946), Note on R A. Bagnold's empirical formula for the critical water motion corresponding with 1he firs1 disturbance of grains on a flat surface, Proc. R. Soc. Lolldon, Ser. A, 187(1008), 16-18. Thornton, E. B., and R T. Guza (1983), Transformation of wave height distribu1ion, J. Go':;Jpll_v,;. Ri!S., 88(10), 5925-5938. USACB (1947), l.ahomlol)' 5!udy ofcquilibrimn beach profile, Bull. 1, pp. 5-ll, U.S. Amiy Corps. of Eng, B!!Mh Erosion Bo:ml. USACE (1985); Littoml zone scdimonlS, Sllll Diego Rcgi011: Dana Point to Mexican border, October 1983-June 1984, l11terim Data Rep. 85-11, Coast of Calif. Storm and Tidal Waves Study, U.S. Anny Corps of Bng., Los Angele.\ District, Calif. USACE-(1991 , Staleoflhc Coast Report, San Diego Region, Final Rep. 1, Coast of Calif. Stonn and Tidal Waves Study, U.S. Army Corps of Eng., Los Angeles District, Calif. Vallis, G. K., G. F. Carnevale, and W. R Young (1989), Extremal energy properties and construction of stable-solutions of the Euler Equa1ions, J. Fluid Mech., 207, 133-152. Zemansky, M. W., and H. C. Van Ness (1966), Basic E11gineering Thermo- dynamics, 380 pp., McGraw-IHI~ New York. D. L. Inman, Integrative Oceanography Division, 0209, Scripps Institution of Oceanoll"'phy, University of California, San Diego, 9500 Gibnan Drive, La Jolla, CA 92093-0701, USA. (dinman@ucsd.edu) S. A Jenkins, Marine Physical Laboratory; 0701, Scripps Institution of Oceanography, University of California, San Diego, 9500 Gibnan Drive, La Jolla, CA 92093-070 l, USA. (sjenkins@ucsd.edu)