Quntittive ssessment of surf-produced se spry erosol Filip P. Neele', Gerrit de Leeuw', Mrtijn Jnsen2 nd Mrcel Stive2' 1TNO Physics nd Electronics Lbortory, The Hgue, The Netherlnds 2Delft Hydrulics, Delft, The Netherlnds Fculty of Civil Engineering nd Geosciences, Delft, The Netherlnds ABSTRACT The first results re presented from quntittive model describing the erosol production in the surf zone. A comprison is mde with erosol produced in the surf zone s mesured during EOPACE experiments in L Joll nd Monterey. The surf erosol production ws derived from erosol concentrtion grdients mesured downwind from the surf zone, fter correction for the bckground size distribution tht ws mesured upwind from the wve breking zone. The erosol production model ws originlly developed from mesurements performed long the Bltic cost1. The model predicts the erosol production from the totl energy dissipted in the wve breking zone, clculted from the costl bthymetry nd deep-wter surfce wve field. n the present work, the prmeteristion of the erosol production in the wve breking zone is mintined, but the energy dissiption in the wve breking zone is clculted using different model tht produces more relistic surf zone widths. Wve dt were obtined from buoys off the Clifornin cost, while bthymetry dt were supplied by the Scripps nstitute of Ocenogrphy. Observed nd predicted erosol production in the surf zone re in good greement, for both sites. The predicted erosol flux reproduces the dy-to-dy vritions nd even some of the observed vritions on time scle of severl hours. 1. NTRODUCTON The erosol produced in surf zones hs recently gthered widespred ttention. Surf zones re reltively nrrow nd, in spite of the likely high production, hve been considered negligible in ccounting for trnsmission losses of visible nd infr-red (R) rdition cross the costline. Studies tking the surf zone into ccount with mesurements both shore nd on ship suggested tht indeed the surf zone could be ignored2. However, recent mesurements prove tht the effect of the surf on erosol production my strongly depend on the loction nd ocenogrphic nd meteorologicl conditions. For exmple, the erosol produced t L Joll my significntly ffect electro-opticl propgtion cross the surf zone, especilly for pths oblique to the cost nd off-shore winds. The erosol production in the surf zone hs been one of the focl points of recent mesurement cmpigns long the Clifornin cost. Aimed primrily t the effectiveness of electro-opticl systems in costl res, the Electro-Opticl Propgtion Assessment in Costl Environments (EOPACE) efforts contined three mesurement cmpigns tht included the study of erosol production in the surf zone. These three Surf experiments were conducted in vriety of environmentl conditions t two costl sites in Cliforni: L Joll, ner Sn Diego, in JnuryfFebrury 199 nd in Mrch 1997, nd in Moss Lnding, Monterey By in Mrch 199. The finl gol of these experiments ws to provide n ssessment of the effect of the surf on electro-opticl propgtion in costl environment. t ws common effort of severl reserch groups from the USA, the UK nd The Netherlnds involving mesurements of erosols, bubbles nd meteorologicl prmeters including turbulence nd visibility, s well s lser visulistion of the erosol plumes over the surf, nd spectrl trnsmission mesurements. nitil results on the observed erosol production t the two sites nd computed JR extinction coefficients were presented by De Leeuw et l., Gthmn nd Smith nd Neele et l.. Appliction of simple erosol dispersion model showed tht the effect of the surf-produced erosol my extend to more thn 2 km downwind from the surf zone. This result is confirmed by clcultions with costl erosol trnsport model. This illustrtes tht the surf zone cn be n importnt fctor controlling electro-opticl (EO) propgtion in costl environments. A physicl model describing the surf source function will be n importnt tool to ssess effects on EO systems. Prt of the SPE Conference on Propgtion nd mging through the Atmosphere S Sn Diego, Cliforni July 1998 SPE Vol. 277-78X/98/$1O.OO
Recently, Petelski & Chomk' nd Chomk & Petelski7 developed model relting erosol production in the surf zone to energy dissipted in the wve breking process, following simple scling rguments. Aerosol flux mesurements performed t loction long the Bltic cost were used to define the constnt of proportionlity in the model. Below, the ppliction of wve breking model to predict the energy dissiption in the surf zones t L Joll nd Moss Lnding is described. The surf erosol production fluxes re clculted using the prmeteristion of Chomk & Petelski7. The results re in good greement with the experimentl dt. 2. AEROSOL PRODUCTON: DATA The mesurements were performed t the piers of the Scripps nstitute of Ocenogrphy (1), L Joll, Sn Diego, in Jnury-Februry 199 nd Mrch 1997 nd Moss Lnding, Monterey By in Mrch 199. At both loctions, se breeze conditions govern the wind pttern: winds from lnd during the night nd erly morning, light winds from se during the fternoon. Aerosol probes were instlled t both ends of the piers to exploit these conditions. For winds from se, dt from prticle counter t the end of the pier provided bckground size distributions in the well-mixed ir mss before it entered the surf zone. At the bse of the piers, three prticle counters mesured the increse in erosol concentrtion due to production in the surf zone, t different heights bove se level. The production in the surf zone cn be estimted from the verticl grdient mesured by the three counters. Stndrd meteorologicl prmeters (temperture, humidity, pressure, wind speed nd direction) were recorded t vrious positions on the piers. The reder is referred to refs. [] nd [] for detiled description of the experiments nd the dt processing. Figure 1 shows n exmple of the dt mesured during the experiments, showing clerly the effect of the surf zone on the erosol concentrtions. The lower prt in figure 1 shows concentrtions of prticles with dimeter of rim, during the penod of 8 through 11 April, 1997 t L Joll. The time in the figure is given in GMT, cusing the periods of winds fromse to occur in the lte evening. The concentrtions re reltive to the concentrtions mesured by the prticle counter t the pier end which, for westerly (onshore) winds, represent the bckground erosol. Also presented in figure 1 re meteorologicl prmeters such s ir temperture, reltive humidity, wind speed, wind direction nd wve height nd period. The erosol concentrtions (shown in ' log formt) show cler correltion with wind direction: for westerly winds, the concentrtions t the bse of the pier (downwind from the surf zone, mesured by instruments pmsl, pms2 nd pms) re up to one order of mgnitude higher thn upwind from the surf zone. Strong verticl grdients re observed, with highest erosol concentrtions close to se level. For winds from lnd, the surf effect is visible in the higher erosol concentrtions observed t the pier end. The correltion is prticulrly cler on dy 1. n the erly morning of this dy, the wind turned from NNW to NNE nd bck severl times. This cused the ir to rrive from lnd (for NNE directions), rther thn cross the surf zone (for NNW directions). This chnge in trjectory is promptly visible in the erosol concentrtions t the bse of the pier, reltive to those t the end of the pier. The dt mesured during the surf experiments were used to estimte the erosol production in the surf zone. Combining the dt from the three surf experiments, wind speeds up to bout 9 rn/s re covered. The totl mount of erosol produced in the surf zone cn be estimted from the erosol concentrtion grdients mesured for winds from se t the bse of the piers. The size distributions from the instrument upwind from the surf zone (t the pier end) provide the bckground erosol concentrtions (before entering the surf zone). The erosol surf production derived for the L Joll nd Monterey sites ws presented in ref. [], where lso the method followed to compute the erosol source function is illustrted in more detil. Figure 2 (reproduced from ref. [1) shows the erosol source function for the L Joll nd Monterey sites, obtined for wind speeds of - ms. Also shown in the figure re severl source functions derived from mesurements t the open ocen. The open-ocen erosol source functions pply to wind speed of 2 ms; the surf source function hs been corrected for the difference in whitecp rtio between the surf (ssumed to be 1%) nd the open ocen t 2 m/s (1%, using the reltion given in ref. [8]). Good greement is observed between the erosol source function derived for the L Joll nd Monterey surf zones nd those obtined for open-ocen conditions, for erosol sizes between 1 nd 1 rim. Of the source function of Monhn et l.9 only the droplet-medited prt is reproduced.
Reltive erosol concentrtions, meteo dt 1 1 '. 2 2 1. 1 12 8 1 8 27 2 E 1 o -2 99 1 11 12 Dy number F Figure 1. Exmple of dt mesured t the SlO pier t L Joll. Shown re reltive erosol concentrtions (dimeter rim) nd relevnt meteorologicl prmeters in the period of 8 through 1 1 April 1997. The bottom pnel shows erosol concentrtions (in 1log formt) t the bse of the Scripps Pier reltive to those from the prticle counter t the end of the Pier (pms): ' log(pmsl,2,) - 1log(pms). For onshore winds (westerly winds, during the fternoons), pms ('reference') mesured the erosol concentrtion in the well-mixed ir mss before it enters the surf zone. The upper pnels show the vrious meteorologicl nd wve prmeters mesured during the cmpign. Prmeters plotted on the left hnd side xes re plotted s solid lines, those on the right hnd side xes s dotted lines. There is strong correltion of erosol concentrtion with wind direction (see, e.g., dy 1). During winds from se, erosol concentrtions re more thn 1 times higher thn in the ir mss unffected by the surf zone. The height bove men se level of the three sensors t the bse of the pier is indicted in the legend in the lower pnel..1 Model. AEROSOL PRODUCTON The erosol production in surf zone cn be clculted from the costl bthymetry nd the incoming wve field using the model proposed by Petelski & Chomk1 nd Chomk & Petelski7 (herefter referred to s CP97). The model is bsed on number of scling rguments. CP97 ssume tht the erosol is produced minly by bubbles, entrined in the wter by breking wves. CP97 ssume tht erosol production (emission flux F ) is linerly relted to the surfce bubble density L1!,
-1. Miller Monhn et l. Smith et l. Surf-1+ Surf2-11 E E E E -12-1 1 2 log( rdius t formtion ) (mu) Figure 2. Aerosol source functions for the surf zone nd the open ocen. The surf zone erosol source functions derived from the Surf-1, nd Surf-2 experiments hve been scled to whitecp rtio of 1% (wind speed 2 ms). There is good greement with the source function of Monhn et l.9 nd Miller', while the source function of Smith et lower concentrtions. predicts FeiLp (.1) The emission flux is mesured in units of g/m2/s. The fctor 1 describes the number of erosol droplets produced per bubble. The bubble density is relted to the volume V of the breking wve: V L=2 irr (.2) where r is the chrcteristic bubble rdius nd 2 constnt. Wve geometry considertions led to reltion between the volume V nd the wve height reduction AH due to the wve breking V=(LsH) (.) with constnt. The decrese in potentil energy (energy dissiption) due to the wve breking is AE=pgVM (.) Equtions (.2)-(.) cn be combined to obtin n expression for L s function of AE, which is then inserted in (.1), yielding Fe C(<dE>)" (.)
where <de> is the verge energy dissipted per unit re in the surf zone nd the constnt C contins ll constnts (including Pw nd g) from expressions (-1) to (-). CP97 estimted the constnt C using mesurements of surf-produced erosol fluxes t loction on the Bltic cost', deriving the following reltion for the emission flux Fe 99.(<dE>)"+ (.) For the clcultion of the emission flux from eqution (.) the verge energy dissiption in the surf zone needs to be clculted. CP97 use wve model bsed on wve energy flux blnce theory of Bttjes & Jnssen' nd Thornton & Guz'. n the present study, the ENDEC* module of UNBEST wve breking model is used. This model describes wve refrction nd wve energy dissiption by wve breking nd bottom friction. The code is lso bsed on Bttjes & Jnsen12. A description of the UNBEST-TC model cn be found in ref. [1J. The results from using the ENDEC module nd those from the wve model proposed by CP97 re compred in Tble 1. The inputs to the models were se bottom of constnt slope (either tn =.2 or tnc =.) nd wve field with significnt wve period of. s nd significnt wve height of. m. Tble shows tht the width of the wve breking zone clculted with the wve model used by CP97 is much lrger thn ccording to the ENDEC model. n the CP97 model, significnt energy dissiption (nd, hence, erosol production) strts t wter depth of bout 2 m, for both se bottom slopes. n contrst, the ENDEC model predicts erosol production in wter depths of m nd less. As rule of thumb, wve is ffected by the se bottom for wter depths less thn bout1. times the wve period squred. For the wves used in these clcultions, this would be bout 2 m. This implies tht in the CP97 model significnt energy dissiption strts in reltively deep wter (bout 2 m). t does not seem relistic to ssume tht erosol production strts t tht point, when the wves re not yet significntly ffected by the se bottom. The surf width predicted by the ENDEC model is more relistic. The totl erosol production in the surf zone clculted with the CP97 model is lrger thn ccording to the ENDEC model, for both se bottom slopes. The differences between the CP97 nd EN- DEC models re further illustrted in Figure, which shows the erosol production profiles perpendiculr to the cost, for both wve breking models, for the se bottom with slope of tncr =.2. Tble 1. Comprison between the wve breking models of CP97 nd the ENDEC model1. The wve field used hs dominnt wve period of. s nd significnt wve height of. m. The se bottom hs constnt slope of either tn =.2 or tn =.. The distnces re mesured from the shore line. CP97 ENDEC CP97 ENDEC Se bottom slope tncc =.2 tnci =.2 tncx =. tnct =. Width wve breking zone (m) 1 1 2 8 Totl erosol flux (ig/m/s) 1. i.7. itt' 7. 1 i' 2.8 i Averge erosol flux (g/m2/s) 9 1 2 Mximum erosol flux (pig/m2/s) 1 8 1 7 17 Position mximum flux (m) 2 1 11 The results in Tble 1 nd Figure confirm the conclusions of CP97. The loclly highest erosol fluxes re produced by steeper se bottom. However, the totl erosol production is highest for shllower se bottom slope, which lso cuses wider wve breking zone. Furthermore, the mximum erosol production, nd the most vigorous wve breking occur close to the shoreline..2 Appliction to L Joll nd Moss Lnding. The CP97 model of erosol emission in the surf zone is pplied to the conditions t L Joll nd Monterey. The se bottom slope t L Joll ws ssumed to hve constnt slope of tncz=.1, to sufficiently deep wter. For Moss Lnding, the verge se bottom slope ws tn=.. The wve field dt (significnt wve height nd period) were obtined from NOAA * nformtion bout this model is vilble t http://www.wldelft.nl/mci/en/softwre/19.shtml (UNBEST) nd httpj/www.wldelft.n/mci/en/softwre/12.shtml (ENDEC module). 7
nd CDP buoys off the Clifornin cost. For the L Joll site, buoy 2 (Ctlin islnd) ws used, for Moss Lnding dt from the CDP buoy in Snt Cruz hrbour. t ws ssumed tht the dt from these buoys re representtive for the deep wter wves cusing the surf t the sites where the mesurements were mde. During ech period with winds from se, totl erosol fluxes were computed t one-hour intervls. As both piers re oriented roughly West, t right ngles to the cost, periods of wind from se were defined s those intervls when wind direction ws Emission flux Endec nd Chomk & Petelski, se bottom: lf=.2 1 9 'v;.. 8 7.2 g. U). 2 Ll 1. CD -1 1 2 7 8 distnce from shore line (m) Figure. Aerosol emission flux profiles over the wve breking zone, for se bottom profile with slope of tn=o.2 (shown s hevy solid line). Shown re the emission profiles predicted by the wve breking model of CP97 (dshed line) nd the Endec model (thin solid line). The Endec model predicts much nrrower wve breking zone thn the CP97 model; erosol fluxes re loclly higher in the Endec results between 18 nd. For wind directions oblique to the cost, the effective surf width increses. To ccount for this effect, the predicted fluxes were divided by the fctorf 1 wdir 27 27 where wdir is the wind direction in degrees. This fctor ws chosen to void singulrities of correction fctor like 1/cos(wdir-27). Figures -c show the results for surf-i, surf-2 nd surf- experiments. The lst pnel in figure c corresponds with the time period shown in figure 1. Thetotl erosol production is given in units of.tg/m/s, where the length unit is length of shore line. There is generlly good greement between predicted nd mesured erosol fluxes. The observed dy-to-dy vritions in the observed erosol flux re resonbly well reproduced by the model, while t times the model lso predicts the correct trend. However, in other cses the model clcultions my be up to one order of mgnitude different from the observed fluxes. This shows tht more reserch is needed to refine the model. Tests re being mde on the prediction of the wve breking model itself, while on the other hnd lso the bubble-medited erosol production model could be improved.. Wve dt from the buoys mentioned in the text re vilble t http:llsebord.ndbc.no.govfmps/swstmp.shtml (for the NOAA dt) nd http://cdip.ucsd.edu/wc/sntcruz.html (CDP dt). 8
z ';s : v;; 2 2 27 28 S. j S.. 1. 29 f S tksc c S 1 oo 8'o9. d'. 2 7 Dy number Figure. () Time seril plots of predicted (solid dots) nd observed (open circles) totl erosol fluxes during the Surf-l experiment. Note the good greement on dys 2 to 29, when even the trend during the dy is reproduced by the model. 7 - J T... #.... C- 7 :. S %:oc:;;c: c 7 $. S. ;8 _ S 8.. q:c 9 7 i _ 71 72 7 7 7 7 77 Dy number Figure. (b) As figure, now for the Surf-2 experiment. 9
-.s.*. o. Dc: 91 92 9 9 u - QX *. ', % w f9... S....1 9 9 97 98 :; jo ' 99 1 11 12 Dy number Figure (c). As figure, now for the Surf- experiment. The lower pnel shows the sme time period s figure 1. The greement between observed nd modelled erosol production for the Surf-2 experiment in Moss Lnding is not s good s for the Surf-i nd Surf- experiments. Aerosol production is overestimted by one to two orders of mgnitude. The windspeed corrected erosol source functions for the Moss Lnding nd L Joll sites gree quite well (figure 2), nd there is resonbly good greement between model nd dt for the L Joll site. Therefore, it is unlikely tht the discrepncy between model prediction nd experimentl dt is due to experimentl errors. A source of uncertinty my be the wve dt used, which were tken from buoy ner Snt Cruz. The wve dt given by buoy in Monterey by (NOAA buoy 2) were not used, s these gve deep-wter wve heights in excess of m, while the surf during the experiment did not show evidence for such wve heights. Without wve dt representtive for the re where the erosol mesurements were mde, no definitive conclusions cn be drwn. n view of the shpe of Monterey By, the wve field t the site is likely sheltered from open ocen wves nd much energy my lredy hve been dissipted upon rrivl t Moss Lnding. The vlue of the results presented here lies in the fct tht they show the generlity of the surf-zone erosol flux model of CP97. The CP97 model ws developed for site long the Bltic cost, using erosol flux dt to determine the constnt relting totl erosol flux from breking wves to the energy dissipted in the surf zone.. SUMMARY AND CONCLUSONS The results presented in this pper show tht the erosol emission from costl wve breking zone cn be resonbly predicted, when the se bottom profile ner the cost nd the deep-wter wve spectr re known. A reltively simple reltion involving the wve energy dissiption7 describes the totl mount of erosol produced in the surf zone. n the present work, the CP97 model ws improved by replcing the wve breking model proposed by Chomk & Petelski by the ENDEC module of the UNBEST-TC code. Results from the ppliction of the two wve breking models showed tht the CP97 wve breking model predicts wves to brek in reltively deep wter. The wve breking zone widths predicted by the ENDEC
code re more relistic nd smller thn those from the CP97 model. Appliction of the erosol production model to two sites on the Clifornin cost shows the generlly good greement between model nd observtions for both sites. The model of Chomk & Petelski7 ws developed for the erosol flux t site long the Bltic cost. The greement between model nd dt for the Clifornin cost indictes tht the pproch leds to resonble results nd, more importnt, shows the generlity of the model. The reltion between totl energy dissiption in the wve breking process nd totl erosol production provides resonble description of the erosol source function. The two most importnt prmeters controlling erosol production in the surf zone (when spume production is ignored) re the costl bthymetry nd the deep-wter surfce wve field. Once these prmeters re known, the erosol flux cn be resonbly predicted. Aprt from these two prmeters in the production process, wind speed plys role in the trnsport of erosol to the prticle counters. Neele et l. showed tht there is positive correltion between wind speed nd erosol production, which might be explined by more efficient trnsport of erosol. The increse in erosol production between wind speeds of -2 m/s nd -8 m/s ws found to be bout n order of mgnitude. The current model does not include such wind dependence. mprovements re expected from better prmeteristion of the erosol production. Bubble size spectr hve been mesured in the surf t vrious occsions, nd the results will be used in future efforts in this topic. The effects of wind speed on the trnsport of the surf-produced erosol should lso be tken into ccount. The vlue of surf erosol production model will lie in pplictions in erosol trnsport models which re used, e.g., to predict tmospheric effects on the performnce of JR sensor systems. nitil studies show the lrge impct of the surf erosol on the erosol concentrtions in the littorl region in off-shore winds. ACKNOWLEDGEMENTS The prticiption of TNO-FEL in EOPACE is supported by the Netherlnds Ministry of Defence, ssignment A9KM729, nd the US Office of Nvl Reserch, Grnt N1-9- -81. REFERENCES. 1. T. Peteiski nd M. Chomk, "Mrine erosol fluxes in the costl zone -BAEX experimentl dt, Ocenologi, 8, 9-8, 199. 2. A. Kohnle, nd J.H. Richter, "Result of the Fll 1981 North Se mesurements" NATO AC/2 (Pnel V/RSG.8) Report 198-1, 198.. G. de Leeuw, F.P. Neele, A.M.J. vn Eijk, E. Vignti, M.K. Hill nd M.H. Smith, "Aerosol production in the surf zone nd effects on JR extinction", SPlEproceedings volume 12, 1-27, 1997.. S.G. Gthmn nd M.H. Smith, "On the nture of surf-produced erosol nd their effect on electro-opticl systems", SPE Proceedings Volume 12, 2-12, 1997. F.P. Neele, G. de Leeuw, A.M.J. vn Eijk, E. Vignti, M.K. Hill nd M.H. Smith, "Aerosol production in the surf zone nd effects on JR extinction", pper presented t the SET symposium on 'E-O propgtion signture nd system perf ormnce under dverse meteorologicl conditions considering out-of-re opertions ', Nples, 1998.. E. Vignti, G. de Leeuw nd R. Berkowicz, "Aerosol trnsport in the costl environment nd effects on extinction", in: Propgtion nd mging Through the Atmosphere, SPE, Sn Diego, July 1998 (this issue). 7. M. Chomk nd T. Petelski, "Modelling the se erosol emission in the costl zone", Ocenologi, 9, 21 1-22, 1997. 8. E.C. Monhn, nd.g. O'Muirchertigh, "Opticl power-lw description of ocenic whitecp coverge dependence on wind speed", J. Phys. Ocenogr., 1, 29, 198. 9. E.C. Monhn, D.E. Spiel nd K.L. Dvidson, "A model of mrine erosol genertion vi whitecps nd wve disruplion", in "Ocenic whitecps", edited by E.C. Monhn nd G. Mc Niocill, pp. 17-19, Reidel, 198. 1. M.A. Miller, "An investigtion of erosol genertion in the mrine plnetry boundry lyer", M.S. thesis, 12 pp., Dep. of Meteorol., P. Stte Univ., University Prk 1987. 11. M.H. Smith, P.M. Prk nd.e. Consterdine, "Mrine erosol concentrtions nd estimted fluxes over the se", Qurt. J. Roy. Meteorol. Soc., 119, 89-82. 199. 1
12. J.A. Bttjes nd J.P. Jnssen, "Energy loss nd set-up due to the breking of rndom wves", Proc. 1thmt. Conf Cost. Eng., Am. Soc. Civ. Eng., New York, 9-78, 1978. 1. E.B. Thornton nd R.T. Guz, "Energy sturtion nd phse speeds mesured on nturl bech",.1. Geophys. Res., 87, -79, 1982. 1. J. Bosboom, S.G.J. Arninkhof, A.J.H.M. Reiniers, J.A. Roelvink nd D.J.R. Wistr, "Unibest-TC 2., drft version.1", Delft Hydrulics, Deift, 1997. 2