Beach Dynamics of Sebago Lake

A Report on the Results of Beach Profiling

by Robert A. Johnston and Martha N. Mixon

Maine Geological Survey Natural Resources Information and Mapping Center DEPARTMENT OF CONSERVATION Robert G. Marvinney, State Geologist

Open-File 98-122 Beach Dy namics of Se bago Lake

A Report on the Re sults of Beach Pro fil ing

Robert A. Johnston Maine Geo logi cal Sur vey 22 State House Stati on Augusta, Maine 04333

Martha N. Mixon F. M. Beck, Inc. One Main Street Yarmouth, Maine 04096

Maine Geo logi cal Sur vey Natu ral Re sources In for ma tion and Map ping Cen ter DE PART MENT OF CON SER VA TION Rob ert G. Marv in ney, State Ge olo gist

Open- File 98- 122

1998 Related Publi ca tions

Dickson, S. M., and Johnston, R. A., 1994, Se bago Lake State Park beach dynam ics: a report on re sults of beach profil ing: Maine Geologi cal Survey, Open File Report 94-4, 189 p.

Johnston, R. A., 1998, Shore line classi fi ca ti on of Sebago Lake, Maine: Maine Geologi cal Survey, Open-Fil e Map 98-123, scale 1:24,000.

Lewis, E. B., and Johnston, R. A., 1998, Slope stabil ity / shoreli ne classi - fica ti on map of the Songo River, Maine: Maine Geologi cal Survey, Open-Fil e Map 98-124, scale 1:4,000. Table of Content s

Page Ab stract ···················································1

In tro duc tion ·················································2 Pur pose···············································2 Ge ol ogy of Se bago Lake ······································2 Shore line Proc esses ········································3 Waves ···········································3 Wind ···········································4 Storms ··········································4 Lake Lev els ········································5 Ice·············································5 Hu man In flu ences ····································5 Cur rents and Long shore Drift·······························5

Meth ods ···················································5 Shore line Clas si fi ca tion ······································5 Beach Pro file Sites ·········································5 Grain Size Data···········································7 Em ery Method of Beach Profil ing ·································7 Plots of Beach Pro files ·······································7 Data Analy sis and Er ror Es timates ·································7 Data Cor rec tions ··········································8 Weather···············································8 Ice Ac tion ·············································8

Re sults ····················································8 Shore line Clas si fi ca tion Map····································8 Analy sis of Pro files ········································9 Songo Beach No. 3 ····································9 Songo Beach No. 4 ····································9 Songo Beach No. 5 ···································10 Songo Beach No. 7 ···································10 Halls Beach 11, Tas sel top Park ·····························10 Halls Beach 12, Tas sel top Park ·····························11 Frye Island 2·······································11 Frye Island 3·······································11 Rock wall 7 ·······································12 Rock wall 8 ·······································12 Stan dish No. 1 ······································12 Stan dish No. 3 ······································12 Sand bar 10········································12 Sand bar 12········································12 Har mon Beach, Bar ton Residence ····························13 Har mon Beach, Banks Residence ····························13 Har mon Beach, Straw Resi dence #2 ··························13 Long Point, Sun ning dale ································13 Interannual Trends ········································13

i Re sults (con tin ued) Large Changes, Fall 1996·····································14 Sea sonal Pat terns ·········································15 Beach Fac ing Di rec tion ······································15 Lake Cur rents ···········································15 Beach Dy nam ics ·········································16

Con clu sions ················································16

Rec om men da tions ·············································17

Ac knowl edg ments ·············································17

Ref er ences ·················································18

Fig ures ···················································19

Ap pen dix I - Beach Pro files ········································43

ii List of Fig ures

Figure 1. Loca ti on map··········································19 Figure 2. Photo graph of Songo Beach ex posed tree roots. ························20 Figure 3. Beach profil e sites ·······································21 Figure 4. Beach and wave ter minol ogy ··································22 Figure 5. Ide alize d beach profil e adjust m ents to lake level changes ···················23 Figure 6. 1990 wa ter levels ········································24 Figure 7. 1991 wa ter levels ········································25 Figure 8. 1992 wa ter levels ········································26 Figure 9. 1993 wa ter levels ········································27 Figure 10. 1994 wa ter levels ········································28 Figure 11. 1995 water levels ········································29 Figure 12. 1996 wa ter levels ········································30 Figure 13. Fetch map of Songo Beach ··································31 Figure 14. Fetch map of Halls (Tas sel top) Beach ·····························32 Figure 15. Fetch map of Frye Island Beach ································33 Figure 16. Fetch map of Rock wall Beach·································34 Figure 17. Fetch map of Stan dish Beach ·································35 Figure 18. Fetch map of Sandbar Beach ·································36 Figure 19. Fetch map of Har mon Beach ·································37 Figure 20. Fetch map of Long Point (Sun ning dale) Beach ························38 Figure 21. Photo graph of ice carry ing sand and ice bull dozing sand on Sebago Lake ··········39 Figure 22. Photo graph of Spring 1997 ice ridge, Songo Beach ······················40 Figure 23. State of Maine Com prom ise Sebago Lake Plan, 1996·····················41

iii

Maine Geo logi cal Sur vey Open- File 98- 122

Beach Dynam ics of Sebago Lake A Re port on the Result s of Beach Profil ing

Robert A. John ston Maine Geo logi cal Sur vey 22 State House Stati on Augusta, Maine 04333

Mar tha N. Mixon F. M. Beck, Inc. One Main Street Yar mouth, Maine 04096

AB STRACT

Se bago Lake, lo cated in south west ern Maine, is the site of this shore line change study. This study follows up on a previous investigation (Dickson and Johnston , 1994) that re ported on three years of beach profil ing at the north end of Sebago Lake. This re port includes an addi tional three years of beach profil ing data at those sites and describes three years of beach pro filing at 12 more sites at 6 beaches around the perime ter of the lake. The ad di tional sites cover a vari ety of compass-faci ng direc tions. In addi tion to beach profil ing, a shoreli ne classi fi ca tion and mapping proj ect was com pleted to de ter mine the types and ex tents of dif fer ent shoreline en vi ron ments around the lake; and a moni toring program for erod ing bluffs was initiated. Map ping for the shoreline classi fi ca tion proj ect was done by boat and tape meas ure using 7.5 minute topo- graphic maps as a base. The re sults show the fol lowing types of sur fic ial ma te rials around the pe rimeter of the lake: till (a mixture of sand, silt, and clay), glacial out wash (sand and gravel), silt, clay, and wetland depos its (silt and de- cay ing plant ma te rial). The shoreline en vi ron ment classes mapped are marsh, sandy en vi ron ments (sand beach, seawall behind sand beach, groins with sand in between rip rap, bluffs behind sand beach, and sand beach with boul- ders present), till, arti fi cial fill, and bedrock. Till is the domi nant en viron ment mapped, followed by sand beaches. Two maps pres ent the results of the study: a shoreli ne classi fi ca tion map of Sebago Lake (Johnston, 1998), and a simi lar shore line clas si fi ca tion map showing the Songo River from the Songo Lock to Sebago Lake (Lewis and John - ston, 1998). The bluff ero sion monitor ing pro gram began in the sum mer of 1996, on six erod ing bluffs. Per ma nent mark ers (steel pins) placed at meas ured dis tances along the bluffs were sur veyed with a to tal sta tion and data was collected through the spring of 1997. This pro gram docu mented mi nor ero sion over this short time pe riod, and fur ther moni - tor ing of these sites is rec om mended be cause bluffs tend to re cede episodically. Re sults of beach pro filing showed that most beach changes are tran sient sea sonal changes related to ice pro - cesses and wa ter level fluc tua tions. Beaches were sta ble over the study pe riod un til the fall of 1996, when the south- facing beaches and one east- facing beach ex pe rienced cata strophic ero sion as a re sult of a com bi na tion of weather events that raised the lake lev els to their maxi mum and later brought high winds out of the south with asso ciated high waves. These large-scale changes are ex pected to be long-last ing. Mi nor ero sion events occurred dur ing spring high water in two years, but were not long-lasting. Erosion events that oc curred in the fall of 1992 and 1995, although they in volved large volumes of ma te rial, were re stored within a year. A picture of beach dy nam ics emerges in which long-term stabil ity is punctu ated by sud den large, long- lasting changes caused by storms during high water peri ods.

1 R. A. Johnston and M. N. Mixon

IN TRO DUC TION the lake level. At a number of places around the lake, erosion of the sand beaches was re ported due to the higher water levels . Pur pose Begin ning in 1990, the Maine Depart ment of Conser va - tion, Bureau of Parks and Recreation, raised concerns that S. D. The Maine Geologi cal Survey , F. M. Beck, Inc., and Port- War ren’s wa ter level manage ment plan was in creas ing ero sion at land Water Distri ct under took stud ies to evaluate the ero sion po- Se bago Lake State Park. In the late 1980's tree roots were ex - ten tial of spe cific sedi men tary en vi ron ments at Se bago Lake, posed along Songo Beach (Figure 2). The exposed roots sug- Maine. The work de ter mined types of mate ri als, their erodabil- gested that sand had been lost from the upper beach due to the ity, whether or not erosion or accre ti on has oc curred at a va riety higher wa ter lev els. The Bu reau of Parks and Rec rea tion asked of sites, and the influ ence of water levels and other envi ron - the Maine Geo logi cal Sur vey (both agen cies are in the Depart - mental fac tors on the shore line of the lake. These other fac tors ment of Conser va ti on) to look at the erosion problem . As a re - in clude ice ac tion, weather, hu man in flu ences, and cur rents. sult, the Survey estab li shed a network of beach profil e stati ons Se bago Lake, Maine's sec ond larg est lake, is lo cated ap - along the sand beaches in the park. Ten stati ons were estab li shed proxim ately 20 miles northwest of Portland (Figure 1). The lake in the fall of 1990 on Cub Cove, Songo, Naples, and Witch Cove cov ers a sur face area of ap proxi mately 47.5 square miles and is beaches (Figure 3). In the sum mer of 1991, two more sites were 316 feet deep. It is the deep est lake in Maine. Se bago Lake added at Halls (Tas sel top) Beach, just south of Browns Point in serves as a publi c water supply for approxi m ately 160,000 peo- Ray mond (Figure 3). Dickson and John ston (1994) publi shed an ple in the Portland area. Erosion of sand bank sedim ents due to open-file report based on the monitor ing of these twelve sites wa ter level changes may affect wa ter qual ity in the lake. during the open-wa ter sea son. The Se bago Lake en vi ron ments stud ied in clude river Due to the increa sing concerns about shoreli ne erosion, banks, lake beaches, and bluffs. Bottom sedi ment studies are in Maine Geo logical Sur vey and Port land Wa ter Dis trict staff, with prog ress (Johnston and others, 1994). The first step was devel - as sis tance from the Friends of Se bago Lake (a lo cal citizen’ s opm ent of a shore line classi fi ca ti on map for Se bago Lake to group), ini ti ated beach profil e training during the fall of 1993. identi fy sites for further study (Johnston, 1998). Based on this Another 34 sites were started in the fall of 1993 and four more map, beach and bluff sites were chosen for profil e loca ti on. The sites were added on the sand beach at the southern end of Frye Is- study examined up to seven years of sand beach profil es to deter - land. This brought the total number of sites around the lake to 50 mine the cause and ex tent of ero sion and ac cre tion at the sand (Figure 3) and expanded beach profil ing from the north end of beaches around Sebago Lake. Begin ning in 1996, six eroding the big basin and Jordan Bay to cover most of the sand beaches bluffs on the lake were studied to deter mine the seasonal patter ns on the lake. Since their initia tion, most of these sites have been of bluff erosion and its control s (Smith, 1997). In 1996 the profil ed on a monthly basis during the open-wa ter sea son, how- Songo River banks were clas sifie d accord ing to their erosion po- ever some of the sites have been profil ed only occa sion all y. The tenti al (Lewis and Johnston, 1998). Both the beach and bluff bluff sites were initi ated in 1996, when six eroding bluffs around stud ies were a co op era tive ef fort be tween the Port land Wa ter the lake were se lected for profil ing (Smith, 1997). Dis trict (PWD), the Maine Geo logi cal Sur vey (MGS), and F. M. The main pur pose of this study is to evaluate sand beach Beck, Inc. of Yar mouth, Maine. profil es to deter mine if erosion or accre ti on has oc curred dur ing Ar tifi cial man age ment of lake lev els at Sebago Lake be gan the past seven years and to ex am ine the pro cesses that cause sand in 1830 when the Basin Dam was first built by the Cum berland to move on and off the beaches. The authors exam ined over 466 and Oxford Ca nal Com pany (Wheeler, 1994). The dam was sand beach profil es through Decem ber of 1996 and 23 bluff pro- origi nally con structed to provide for bet ter naviga ti on and to di - files (1996 and 1997) for indi ca ti ons of ac creti on or erosion. vert water to a newly construc ted canal. The present full pool They con sid ered weather, lake lev els, and ice ac tion when analy - ele va tion is 266.65 feet above mean sea level. At the pres ent zing beach and bluff trends. Smith (1997) reporte d bluff erosion time the wa ter lev els in the lake are gov erned by a wa ter level result s. Refer to the secti on on Methods for detai led infor m ati on manage m ent plan devel oped by S. D. Warren, the Federal En- on the methods used. ergy Regula tory Agency, the Portland Water Distri ct, vari ous state agen cies, and lo cal citizen groups. Ge ol ogy of Se bago Lake Shoreline ero sion becam e a con cern at Se bago Lake in the mid 1980’s when the S. D. Warren Pa per Com pany, the owner of Sebago Lake is locat ed along the boundary betwee n the the dam, changed their wa ter level manage ment plan. Bef ore coastal lowland and centra l highlands of New England (Denny, 1986 there was no specif ic rule curve for the water levels . In 1982). Topo graphi c reli ef in the region is low to moder ate , with 1986, S. D. Warren began to hold more water in the lake during ele vati ons ranging from 200 feet at the southern and east ern the fall and winter months in order to obtai n more cost ef fecti ve sides to 1300 feet to the northwest . North of Indian Island, the electri c rates. This raising of the water levels brought on an in- lake is under lai n by the Se bago batholi th, a Late Carbon if er ous - crease in the number of com plaints about their manage m ent of Early Per mian in tru sion of muscovite- biotite granit e and pegm a-

2 Beach Dynamics of Sebago Lake

tite with an es timated age of ap proximately 290 million years follows: waves, wind, lake levels , storms, ice ac tion, hum an in- (Hussey, 1996). The southern porti on of the lake, below Indian ter ac tion, and cur rents. Is land, is un der lain by Si lu rian metape li tic rocks of the Rindge- Waves. Wind and waves are the most im portant agents of mere Form ati on. These rocks consis t of rusty and non-rusty gray shoreli ne change at Sebago Lake. Wind forms waves and their to sil very muscovite- biotite- quartz schist (Hussey, 1996). The size depends on the strength and dura ti on of the wind, as well as lake lies in a gla cially eroded basin. Differ en tial ero sion of the the distanc e the wave travels (fetch). The stronger the wind, the two bedrock types devel oped a larger and deeper ba sin to the longer the wind dura ti on, and the longer the fetch, the larger the northwest where the lake is under lai n by the gran itic rocks, while wave. Waves cre ated by boat wakes were not con sid ered in this the shallow and smaller basin to the south is under lai n by the report, although in certai n constri cted areas of the lake they may more re sis tant meta morphic rocks. play an im portant role. Con ti nen tal gla cia tion strongly influ ence d the lake and the In this report we consid er the beach envi ron m ent (Figure 4) surround ing landsca pe. The remnants of the conti nental ice to be the zone of modern, un con soli dated granular sedi ment that sheet left the Se bago Lake area thir teen to four teen th o u s and extends from the upper m ost limit of wave acti on to the deepest years ago. When the area was depress ed under the weight of the wa ter depth agi tated be neath waves or “wave base.” Wave base con tinen tal ice sheet, the ocean was in con tact with the re treating is a depth equal to 25% of the deep water wavelengt h (Kom ar, edge of the ice sheet. Sebago Lake, and much of the surround ing 1976). Beach profil es meas ured in this study typi cally extend land scape, was sub merged below sea level. The inland marine under wa ter , but not far enough offshore to reach wave base. We limit dur ing the most re cent gla cia tion, ap proxi mately 13,000 es tim ate wave base to be ap proxi mately 24 feet (based on es ti- years before pres ent (Bloom, 1960), was in the vicin it y of the Se - mates below). Hence, sand exchange in the beach envi ron m ent bago Lake basin. That inland limit of the ocean has been identi - by waves can include the offshore lakebed beyond the lim its pro- fied on the Sur fic ial Geologi c Map of Maine (Thomp son and filed and illus trated in this re port. Borns, 1985) based on the expo sure of marine clay (called the Wave acti on is a prim ary influ ence on the movem ent of Pre sump scot Form ati on) along the shoreli ne near Whites Bridge beach sand (Kom ar, 1976). Waves that approach the beach at an and at the northern end of Jordan Bay (Bolduc and others, 1994). angle cause sand movem ent along the shore called long shore The Pre sump scot Form ati on is not found along the shoreli ne of drift. The di recti on of wave approach is control led by the wind the big basin or on land north of the mapped marine limit direc ti on and wave refrac ti on in shallow wa ter. In addi ti on, (Thompson, 1976; Thompson and Smith, 1977). Sur fic ial de- there may be sea son al ity in the wind speed or direc ti on that may posit s pres ently exposed in the region include till, gla cial out- deter mine the volum e of sand transporte d by long shore currents. wash (sand and gravel), silt and clay, and wet land mate ri als In a study by Lorang and others (1993) in a Montana lake, the re- (Thompson and Smith, 1977). The age of Sebago Lake can be distri buti on of annual wave ener gy (from that expe ri ence d by a es ti mated to be ap proxi mately 13,000 years, a time when the natu ral lake) due to regu lated lake lev els caused in creased beach ocean retreated from the area due to the iso static rebound of the erosion. In Montana , higher lake levels caused erosion in the earth’s crust af ter the re treat of the last con tinen tal ice sheet. stormy fall season when wave en ergy was great est. Fall storms The Holocene epoch, the warm time peri od in which we may play an im portant role in moving sand onshore or offshore live, began about 10,000 years ago. Since the begin ning of the or along the beaches at Sebago Lake. Holocene , only minor changes have occurred to the sedim ents in In general , lake level restri cted to one eleva ti on (high or the Sebago Lake basin. The large sand depos it s of the Songo and low) limits wave ac tion to a rela tively nar row part of the beach Northwest River del tas (Bloom, 1959) are evidence of the re- profil e. A prolonged episode of single pe riod wave acti on on a working of glacia l depos it s by fluvial proces ses. Rivers and beach should produce a profil e of “equili brium .” In re alit y, streams depos it ed mate ria l in the lake ba sin form ing thick de pos- how ever, lev els fluc tu ate due to pre cipi ta tion, wind, baro met ric its of sandy delta sedim ents. Sedi ments de pos it ed along the pres sure, dams, etc. These fluc tu ating lake lev els and wave shore and at the bottom of the lake are the result of erosion of ma- heights will influ ence a wider range of the beach profil e than a teri als exposed along the shore line of the lake, at vari ous levels lake with a consta nt or narrow range of levels . Ideal ize d profil e throughout the Holocene . Processe s acting on those exposed de - adjust m ents due to lake level changes are shown in Figure 5. posit s of clay, sand and gravel, and bedrock include frost acti on Unli ke this ideal situati on, waves are of various heights and peri - and ice, wind and waves, long shore drift, and other fluvial pro- ods and can arrive from dif ferent di recti ons. These condi ti ons cesses. vary sea son ally, fur ther com pli cat ing pre dic tions of pro file variati on due to lake level changes. Conse quentl y, an equi lib - rium pro file, in the strictest sense, is never es tab lished. In stead, Shore line Proc esses sand is al ways shifting in response to exist ing condi ti ons. Ex- treme condi tions, such as storm waves, can cause large changes The sand beach and bluff ecosys tems are the most dy namic that may take weeks or months to erase. A beach pro file then, is of all the geologi c envi ron m ents at Sebago Lake. Im portant the com plex result of past events and include s epi sodes of both agents of shore line change that af fect those eco sys tems are as ero sion and de po si tion.

3 R. A. Johnston and M. N. Mixon

Waves can be either con struc tive (de po si tional) or destruc - (distanc e of travel across the lake), dura ti on of wind, wind direc - tive (ero sional) to a beach pro file. Wave char acter istics such as tion, wind speed, and water depth. height and length as well as beach slope and grain size deter mine Wind. The study of wind di recti on and occur rence in Dick- how sand will move under certai n condi ti ons. The proces s of son and Johnston (1994) showed that most of Maine's winds are landward movem ent of sand be neath shoal ing waves is well un- in flu enced by pre vail ing west er lies (Tables 3-6 and Figure s 9 derstood and docum ented. The proces s called Stokes Drift (Ko- and 10 in Dickson and Johnston, 1994). Winter winds have a mar, 1976) result s from wave orbit al fric tion and speed strong northwest erly com ponent, while spring winds have a dif ference s below wave crests and troughs. Sands on the outer southerl y com ponent. Sum mer winds also show a strong south- part of the beach profil e expe ri ence an oscil latory (onshore- erly com ponent, while fall winds show a distri buti on from north offshore) moti on below a passing wave. Smaller, construc tive to south (Dickson and Johnston, 1994). The strongest winds oc - waves produce a stronger landward veloc it y which result s in cur in the winter and spring, while the calm est winds are in the sand shifting landward beneat h shoal ing waves. Conse quentl y, sum mer. not all waves are det ri men tal to the beach. In fact, this land ward Another wind variable af fect ing wave heights is fetch. moti on of sand by Stokes Drift causes sand accu m ula ti on along Fetch var ies with di rec tion from Sebago Lake beaches and the upper part of beach profil es and alm ost certai nly helps main- ranges from 3.1 miles at Har mon Beach (Figure 19) to 9.1 miles tain Se bago Lake beaches. at Rock wall Beach (Figure 16). The rem aining fetches range Sand can also be car ried into deeper water where it would from 4.2 to 7.1 miles (Figures 13-20). Larger waves can be built build up the offshore porti on of a profil e. This offshore shift in over larger fetches so the largest waves breaking on Se bago Lake sand re sults in a lower-slopi ng profil e with appar ent erosion of beaches should come from the northwest onto Rock wall Beach the sub aer ial beach. The sand can be returne d landward by a dif - (Fig ure 16). ferent wave type or by lower ing the lake level. When the lake Using a theoret ical approach, the height of waves reaching level is lowered, sand can be re worked back ashore (even by Sebago Lake beaches can be approxi m ated (Dickson and John- small waves) onto the upper part of the beach profil e. Sand car- ston, 1994). Wave heights can be es ti mated graphi cally from no- ried ashore leads to a steeper beach pro file. As a re sult of pro file mo grams based on wind speed and fetch (U.S. Army Corps of steep en ing, high waves can travel closer to shore before break - Engineers, 1984, p. 3-50). A maxim um signifi cant wave height ing. This condi ti on allows for greater scour near the water line for Songo Beach is 4.4 feet and pe riod is 4.3 sec onds and re sults and erosion back to a flatter profil e. In short, the profil e response from SSE gale force winds (winds 32 mph or greater). A gale to waves and lake level conti nuall y changes (Figure 5). will produce a 3.4 foot wave over a 6 mile fetch and a 2.4 foot Because no ob ser va tional wave data were avail able for this wave over a 3 mile fetch. study, a spe cific inves ti gati on has not been made of wave height At Halls Beach the maxi mum sig nificant wave height for a and peri ods at the Sebago Lake profil e sites. Such a study would SSW gale is 3.5 feet and wave peri od is 3.7 sec onds. In the case help de ter mine which wave condi ti ons are construc tional and of ei ther the state park or Halls Beach, fully-develop ed (“deep which are ero sional. Pre dicted wave heights were cal cu lated in wa ter”) waves can be created in as lit tle as two hours. The sig - the En vi ron men tal Im pact State ment on the re li cens ing of the nifi cant wave height is the aver age height of the highest one- Eel Weir Hy droe lec tric Pro ject Dam (FERC, 1997). If lake lev- third of the waves (Kom ar, 1976). For Se bago Lake, wind dura - els rem ain high during peri ods of high waves then it is pos sible tion and water depth do not restri ct the maxim um size of waves. that erosion could occur in the upper porti on of the beach profil e. Wind speed and di recti on (and hence fetch) are the two factors If lake levels were low during peri ods of high waves, then the control ling maxim um wave heights at the sand beaches. Gale lower porti on of a beach profil e would be eroded. force winds last for only a few hours each year, but could create Beaches ex ist be cause of a bal ance be tween sedim ent sup - signifi cant wave heights in exces s of 4 feet at Sebago Lake ply and loss in di recti ons both paral lel and perpen dicu lar to the beaches. Waves will affect the beach in all seasons, but are re - shore. Shallow wa ters with sandy lake beds can supply sand to duced in im portanc e in winter due to ice cover. Signifi cant wave beaches. The delta of the Songo River may be a sand source for heights for the other sites are calcu lat ed in an envi ron m ental im - Songo Beach. During peri ods of low lake level, the delta surface pact statem ent by FERC (1997). may be reworked more vigor ously by waves and addi ti onal sand Storms. Two types of storm sys tems have an im pact on the may be car ried to ward the beach. Similarly , deeper por tions of Sebago Lake region: low pressure system s and tropical storms the beach pro file may be a source of sand to the beach. At times or hurri cane s. The most com mon type of storm is the ex tra tropi- of low lake level, waves may re work the lower beach area and cal low pres sure sys tem which gener all y tracks in a southwest to carry sand onshore or offshore. north east di rec tion. When a low pressure system tracks to the In sum mary, wind- generated waves af fect the shore, the east of the lake basin, close to or out over the Gulf of Maine, the swash zone, and the lakebed above wave base. Changing wave winds tend to blow from a northeas terl y di recti on (a condi ti ons and lake levels result in dy namic beach profil es that "nor'easter"). When a low pres sure sys tem tracks well to the never reach a sta ble equi lib rium. Wave size de pends on fetch west of the lake ba sin (for ex am ple, up the St. Law rence River

4 Beach Dynamics of Sebago Lake

valley ) winds blow from a southerl y di recti on on the lake. When sional and de po si tional ef fects on a beach. Sebago Lake ice fre - a ex tra tropi cal low pres sure sys tem tracks over the lake ba sin, quently incor po rat es and carrie s sand (Figure 21). In addi ti on, storm winds blow from an easterl y di recti on. ice physi cally plows frozen sand shore ward onto the sub aer ial Hurri cane s occur infre quentl y in Maine, but can produce beach. sig nifi cant size waves on the lake. These tropical sys tems origi - Hu man In flu ences. Hu man in flu ences on Se bago Lake’s nate in the low latitudes over warm ocean ar eas in summ er and shoreli nes include foot traf fic in sensi ti ve areas , boat wakes, fall. Hurri cane Bob tracked just east of Portland in August of dredging, sea wall and groin construc tion, and recrea ti onal ve hi- 1991 with north winds of over 40 mph and gusts greater that 60 cle tracks. Foot traf fic is no ticeable on Songo, Halls and Har mon mph. However , since there were no estab li shed profil e sites on Beaches. Boat wakes cause prob lems when boat ers cre ate any of Sebago Lake’s north-facing beaches, no erosion was re- waves when exit ing nar row channels such as along the Songo corded. River in Naples. Dredging and groin construc tion in some areas Lake Lev els. Lake lev els, par tially controlled by a dam, cause the inter rupti on of the natural movem ent of sand along the fluc tuate up to five feet annu all y (Figures 6 - 12). Water levels shore line. Rec rea tional ve hi cle tracks were oc ca sion ally visi ble deter mine where, and for how long, waves have an im pact on the on the beach at the south end of Frye Island and at the Stan dish beach. Sebago Lake’s norm al maxim um water level is 266.65 sites. feet above mean sea level. Water lev els above that eleva tion spill Current s and Long shore Drift. Se bago Lake cur rents and out over the Eel Weir Dam in Wind ham. The lake is usu ally at its long shore drift were stud ied by Char lie Page at Bates College highest level in May or June, and usuall y reaches its lowest lev- (Page, 1996). His work showed that, depend ing on which direc - els in Novem ber. The months of Novem ber through mid-Mar ch tion the wind was blowing, sand would move in either direc ti on gener all y show a water level betwee n 262 and 263 feet above sea along Songo Beach in Se bago Lake State Park. Much of the spit level, with a grad ual rise in late- March through April to the at Songo Beach appears to have been built by long shore drift and 266.65 level in May or June. the movem ent of sand by current s. Lisius and others (1990) Water lev els have been a con ten tious is sue in the lake ba sin stud ied cir cu la tion in the lake at differ ent wa ter depths in the late since the mid 1980’s when S. D. Warren changed their wa ter 1980’s and early 1990’s. Their re port showed that current s in the level man age ment plan. Reported cases of ero sion in creased at Lower Bay move in a counter clockwise direc ti on. This result about that time. Begin ning in 1994, the Federal Ener gy Regula - may af fect the movem ent of sedi ment along the southern lake tory Com mis sion (FERC) re viewed en vi ron mental con cerns in shore line. response to com plaints about S. D. Warren’ s manage m ent of wa- ter levels . FERC (1997) pre pared a final envi ron m ental im pact METH ODS statem ent which ad dressed wa ter level manage ment and beach erosion around Sebago Lake. Shore line Clas si fi ca tion The wa ter level manage ment plan now in effect is the 1996 State of Maine Compro mise Se bago Lake Water Level Plan A field inves ti gati on of the Sebago Lake shore line began in (Figure 23). It lists five recom menda ti ons: (1) earli est full pond the late spring of 1992. By boat, the various shore line sedim ent date - May 1; (2) maxim um full pond dura ti on - 3 weeks (be- typeswere identi fied. Navi gati on was by com pass, and the data tween May 1 and the sec ond week of June); (3) tar get level for was logged onto USGS 7.5’ topographi c maps. Three days were August 1 - 265.17 feet; (4) tar get level for Novem ber 1 - 262.5 needed to traverse the enti re 105 mile shore line of the lake. In feet; and (5) tar get level for Novem ber 1 - January 1 peri od: the of fice, car tog ra phers dig it ized the 7.5’ topo maps using the 261.0 feet msl or lower two times in nine years. These rec om- ARC/INFO geo graphic in for ma tion sys tem (John ston, 1998). menda ti ons were made in order to mini mize erosion during early In the spring and sum mer of 1996, a shoreli ne classi fi ca - spring high water , to have am ple water available for boat traf fic tion map was construc ted for the Songo River (Lewis and John- during the sum mer boat ing season, and to allow , during the two ston, 1998). Methods used were similar to the Se bago Lake out of nine low years, enough wa ter level rise to bring sand up the shore line clas si fi ca tion, but ar chive aer ial pho tog ra phy was used profil e onto the upper beach. to note his toric riv er bank po si tions. Ice. Ice covers Se bago Lake for approxi m ately three months each winter when strong winds are predict ed on the lake. Beach Pro file Sites Since win ter wave ac tion is re duced by ice, beach ero sion by waves on the lake should also be re duced in the winter . How- Eight een pro file sites on eight beaches were se lected for in - ever, an other pro cess that af fects sand trans port is ice action. clusion in this study. Table 1 presents descri pti ve infor m ati on on Wind stress on the ice causes it to move across the lake and pile each pro file site. The grain size of the beaches ranges from me - up along the shore. On the Ohio shore of Lake Erie, ice is sus- dium to very coarse sand (Ta ble 1). The beaches range in length pected of causing beach erosion (Bar nes and others, 1993). In from about 530 feet at Rock wall to over 3900 feet at Songo the St. Law rence River es tu ary ice may de posit sand along the Beach. Beach pro file sites were cho sen on the ba sis of ac cess, shore line (Di onne, 1993). Conse quentl y, ice can have both ero - reporte d erosion, and facing di recti on. The beaches chosen for

5 R. A. Johnston and M. N. Mixon

TA BLE 1. BEACH PRO FILE SITES

Location Site Beach Length Slope Av er age Fetch Fetch Length (feet) Phi Size Di rec tion (miles) Frye Island ...... Site 2 1370 1:10.2 0.75f(cs) S 4.2 Site 3 1370 1:10.1 0.55f(cs) S 4.2 Halls Beach ...... Site 11 1510 NC -0.15f(vcs) SSW 7.1 Site 12 1510 1:8.3 -0.05f(vcs) SSW 7.1 Har mon Beach.....Bar ton residence 2840 NC 1.20f(ms) ENE 3.1 Banks residence 2840 NC 0.75f(cs) ENE 3.1 Straw resi dence #2 2840 1:9.0 0.90f(cs) ENE 3.1 Long Point Beach ...Thomp son residence 3175 1:7.8 -0.10f(vcs) NE 6.2 Sun ning dale residence 3175 NC .NC NE 6.2 Rock wall Beach ....Site 7 530 1:10.0 -0.35f(vcs) NW 9.1 Site 8 530 1:10.1 -0.10f(vcs) NW 9.1 Sand bar Beach.....Site 10 1895 NC 0.70f(cs) NE 4.4 Site 12 1895 1:8.3 0.50f(cs) NE 4.4 Songo Beach...... Site 3 3935 1:10.5 0.75f(cs) S 6.0 Site 4 3935 NC 0.30f(cs) S 6.0 Site 5 3935 1:10.9 0.50f(cs) S 6.0 Site 7 3935 NC 0.40f(cs) S 6.0 Standish Boat Launch . Site 1 3555 NC 0.40f(cs) N 4.2 Site 3 3555 1:8.8 0.50f(cs) N 4.2 Notes: NC = not cal cu lated vcs = very coarse sand, cs = coarse sand, ms = medium sand (Folk, 1974) de tailed analysis in this study were those that rep re sented a vari - slope in 1994 was calcu lat ed to range from 1:7 to as low as 1:17 ety of en vi ron mental fac tors in clud ing wind di rec tion, fetch, lo- (Dick son and John ston, 1994). Pro fil ing re spon si bil ity here be- cati on, the number of profil es taken, and rem oteness. Some of longs to the Maine Geologi cal Survey . the sites were cho sen be cause of the lack of hu man in flu ence. Halls Beach (also called Tas sel top Beach or Tas sel top Baseline data on beach slope and grain size were col lected at Park) is lo cated well away from the state park beaches and is in each of the eight een sites for future com pari sons. Mul tiple pro - Jordan Bay (Figure 3, 14). This mainland beach is 1500 feet or files were es tab lished along each beach in or der to evalu ate long- more in length and at the profil e sites, near the center of the shore shifts in sand (e.g., erosion at one end and deposi ti on at the beach, the slope is in the range of 1:8. Profil ing began at these other). Profil e loca ti ons are shown in Figure 3. sites in July of 1991 and is the respon si bil ity of the Maine Geo- Beach slope was calcu lated with a simple lin ear re gres sion logi cal Sur vey. of profil e data points. Data points used in the regres sion were The south-facing beach at the southern end of Frye Island from the highest point clos est to the reference pin to a point ini tially had four beach profil e sites (Figure 15). Site 4 was de - marking the aver age profil e dis tance offshore for that site. stroyed in the fall of 1996 during construc tion of the new marina Songo Beach is a main land beach lo cated in the day- use entra nce. Sites 1 and 3 are include d in our analy sis. This beach is area at Se bago Lake State Park (Fig ure 13). The main beach (ex - approxi m ately 1400 feet long and has a maxim um fetch length of cluding the sand spit) is about 2000 feet long and has an aver age 4.2 miles in a south-facing di recti on. Profil ing began on Frye Is- slope of about 1:11, based on all of the profil e data. Sites profil ed land in the fall of 1993 and is the re spon si bil ity of the Maine on Songo Beach in clude Songo 3, 4, 5, and 7. Profil ing began at Geo logi cal Sur vey. these sites in the fall of 1990. This beach faces south with fetch Rock wall Beach, located on Port land Water Dis trict prop - lengths of up to 6.2 miles. At the west ern end of Songo Beach a erty, is a north west-fac ing beach with a fetch distance of 9.1 spit conti nues away from the mainland for a distanc e of about miles to the northwest (Figure 16). This beach is approxi m ately 2000 feet. This is the loca ti on of Songo 7. Slopes on the spit 500 feet long and has an aver age slope of approxi m ately 1:10. beach were not calcu lated with the newer pro file data, but its The two pro file sites lo cated on this beach are in a re mote sec tion

6 Beach Dynamics of Sebago Lake

of the lake that sees lit tle hu man in ter fer ence (swimming is pro - Em ery Method of Beach Pro fil ing hibit ed). Profil e sites were estab li shed here in Octo ber of 1993, and pro fil ing re spon si bil ity be longs to the Port land Wa ter Dis- Beach profil es are lines surveye d perpen dicu lar to the trict. shoreline that rec ord the shape of the beach at the time of meas - The two Standish profil e sites used in this study are locat ed urem ent. In this report a serie s of profil es record changes in at the southern limit of the lake along a 3500 foot sand beach beach shape over a pe riod of seven years provid ing insight into (Fig ure 17). No swimming is al lowed at this site so foot traf fic is the pro cesses that shape Se bago Lake’s beaches. Waves, cur - at a mini mum along the beach. How ever , evi dence of rec rea - rents, lake lev els, sedi ment size, ice ac tion, and hu man in flu - tional ve hic le traf fic has been noted. The beach faces north and ences affect these pro cesses. From the time se ries of pro files, the has a fetch dis tance in that di recti on of 4.2 miles. Profil ing began ero sional and accre tional his tory of the beaches can be deter - on this beach in Octo ber of 1993 and is the respon si bil ity of the mined, and the coastal proces ses that shape the landform s can be Port land Wa ter Dis trict. studied. Sand bar Beach is a north east-fac ing beach lo cated along K. O. Em ery (1961) devel oped the “visual method” of the west ern edge of Lower Bay (Figure 18). It is alm ost 2000 beach pro fil ing which in volves meas ur ing the ver ti cal dis tance feet in length and has its longest fetch dis tance to the northeas t at betwee n the tops of two graduate d poles of equal length by level - 4.4 miles. Profil ing re sponsi bil ity here belongs to the Portland ing on the hori zon. Dickson and Johnston (1994) out lined the Water Distri ct and began in Octo ber of 1993. This site steps used in meas uring and record ing profil es in this study. experiences a great deal of boat traf fic. Data were en tered into a com puter spread sheet pro gram, and cu - Har mon Beach is an 2800 feet east-fac ing sand beach that mu la tive hori zon tal dis tances and dif fer ences in ele va tion cal cu - is heavil y devel oped with sum mer cot tages (Figure 19). The lated. Data were ed ited to cor rect re cord ing er rors (see Data fetch dis tance to the east- northeast is 3.1 miles. Pro fil ers noted Analy sis sec tion be low). evidence of hum an influ ence with foot traf fic, boat traf fic, and seawalls and docks. Profil ing re sponsi bil ity here belongs to the Plots of Beach Pro files Maine Geologi cal Survey and began in Septem ber of 1993. Long Point is a north-faci ng beach on the southern edge of Beach pro files were plot ted using Tech base software. A the Big Basin (Figure 20). Sites profil ed here include Thompson large ver ti cal ex ag gera tion of ap proxi mately 20:1 ac cen tu ated (where the grain size data sample was collected) and Sun ning - changes in the eleva ti on of the beach. A plot of all the profil es at dale. The Thompson site has the steepest profil e among all the each lo cation de fined the ver tical en ve lope that the beach oc cu - sites on the lake at 1:7.8. Profil ing re sponsi bil ity here belongs pied dur ing the study pe riod. The out erm ost points that de limit with the Friends of Sebago Lake and be gan in Septem ber of this enve lope de fine a “sweep zone” (Barnes and King, 1955). 1993. The beach is over 3000 feet long. Profil e sites on Long Analy sis of the sweep zone helps to define the overall vari abil ity Beach (west of Long Point) were not include d in this re port due of a given beach profil e loca ti on which can then be com pared to to the infre quency of profil ing. indi vid ual changes seen betwee n months, sea sons, or years. In addi ti on to sweep zone plots for each site, plots were made of Grain Size Data each pro file, usu ally three to a page, with sepa rate pages for spring, sum mer, and fall plots. Where suf ficie nt data were avail- Grain size data was col lected for all eight een of the beach able, plots were made show ing all of the late summ er pro files at a profil e sites analy zed in this study (Table 1). The sam ples, ex- site. These plots helped define inter annual and long term trends cept those from Songo Beach, were col lected on Novem ber 22, at the pro file sites. Ad di tional plots were made for some sites, 1996 at the wa ter line when the wa ter level was ap proxi mately com paring the last fall profil e of one year to the first spring pro- 265.8 feet above mean sea level. The Songo Beach sam ples file of the next year, all fall pro files at a site, and other combi na - were collec ted on August 10, 1995 at the water line, with a water tions. A light table was used to com pare profil es that were not on level of approxi m ately 263.8 feet (Page, 1996). All sam ples the same page. were analyzed in the Sedi ment Lab at Bates Col lege in Lewis ton, Maine. Methods used to obtai n the median grain size are dis- Data Analy sis and Er ror Es timate s cussed in detai l in Folk (1974). Percent gravel, percent sand, and per cent silt and clay were cal cu lated for each sample. Also, the Data for each pro file site was ex am ined chrono logi cally to median phi size was deter mined from the weight of sedi ment in look for events, epi sodes, or trends af fect ing the beach. The full each phi size us ing the cum ula ti ve curve, arithm etic ordi nate data set (sweep plot) was used to as sess to tal pro file vari abil ity method (Folk, 1974). Phi(f) is a loga rith mic scale used in ana - over the study pe riod. Profil e trends by month and season were lyzing sedim en to logi cal data. This scale de fines the size of par ti- ana lyzed to compare sea sonal changes. Late summ er or fall pro - cles (f = -log2S, where S = grain size in mil lim e ters). The larger files, when wa ter lev els are low est and the long est pro files are the f unit, the smaller the par ticle. Grain size data was also col- gen er ally collected, were used to as sess in ter an nual changes and lected on the six bluff sites and is pre sented in Smith (1997). long term trends.

7 R. A. Johnston and M. N. Mixon

Since there is a large varia tion in the amount of data col - 17, 1993; Straw #2, Octo ber 18, 1993) which could not easil y be lected for dif fer ent sites (7 pro files col lected over 3 years at Frye cor rected. No cor rec tions were made in those cases, and data Island com pared to 50 collec ted over 7 years at Songo 5), the de - were not include d in the analy sis. gree of our under sta nding of proces ses and our abilit y to recog - Changes in pro file start ing lo ca tions re quired ad just ments nize trends is dif fer ent for differ ent sites. Those sites with the to some data at the Har mon Beach, Banks and Bar ton residences. highest number of profil es over the longest time peri od are the At the Barton residence , two profil es which started at the corner best un der stood. of the house could not be rec oncil ed with the others, and were Measure ment pre cision is es tim ated at less than 3 inches omitted from analysis. verti cally and a foot horizon tal ly at each data point on the profil e. Dickson and Johnston (1994) de scribe data correc ti ons In the Emery method an er ror at any meas ure ment lo cation is car - made to the 1990 through 1993 data for Songo Beach and Halls ried through the rest of the profil e. Large errors, therefore, re sult Beach pro file lo ca tions. in pro files that are out liers to the sweep zone. All out lier profil es were scru ti nized for pos si ble re cord ing or meas ure ment er rors Weather and cor rected where pos sible. Smaller ran dom er rors are as - sumed to can cel each other out. System atic error s would lead to Weather infor m ati on was obtai ned from the Nati onal greater im preci sion in the outer porti ons of profil es. Weather Serv ice of fice in Gray. Monthly summ aries of lo cal cli - Use of the Em ery method intro duces some very small sys - ma to logi cal data for Port land, Maine were used as a source of tem atic error s into the ver tical eleva tion data. These er rors arise wind speed and di recti on. Sebago Lake weather dif fers from becaus e the oppo sit e shore line of the lake is used as a hori zon for that of Portland mostly in the sum mer when the lake region expe - level ing the tops of the survey stakes. These errors result in a ri ences less of a sea breeze. The Se bago Lake re gion also ex pe ri - very slight un der es ti ma tion of the down ward-slop ing por tions ences more ex treme tem pera tures than Port land due to its of profil es and a very slight over esti mati on of the upward-sl op - dis tance from the milder ocean. Ad di tional weather data was ing porti ons. col lected and analyzed from the Portland Water Dis trict weather Where the distanc e to the hori zon or lake shoreli ne is a mile sta tion in Stan dish. Only 1995 and 1996 wind data were avail - or greater, the er ror in tro duced for a typical meas ure ment is ap - able from this lakeside site (see Smith, 1997). In both weather proxi mately +/- 0.01 foot or less (negati ve for downward sloping data sets sustai ned winds were analy zed rather than wind gusts. readings, positi ve for upward sloping or flat incre m ents) for each Wave-building winds are the longer dura ti on events, not the meas ure ment taken. Fetch maps for each site (Fig ures 13-20) gusts. show the distanc e to the oppo sit e shore line. Accord ing to these fig ures, the Halls beach sites (Tas sel top Park) have the shortest Ice Action distanc e (1.9 miles) to the oppo sit e shore line on Raym ond Neck. The re sult ing er ror for a typi cal meas urem ent would therefore be Ice may cover the lake for up to five months each year. In less than 0.01 foot, or less than 0.1 foot over a 100-foot profil e shal low wa ter sand is frozen onto the ice and is re distri buted by meas ured at 10-foot in ter vals. The er ror in creases with in creas - ice-push and ice-raft ing. In order to deter mine the role of ice on ing size of hori zontal inter vals and with closeness of the hori zon. the beaches at Se bago Lake, pro files were com pared from the Inter vals greater than 10 feet are not used in these beach profil es. late fall and early spring. Beaches were visited dur ing early In any case, these sys tem atic error s do not af fect the de tection or spring to look for evidence of ice ac tion, and photo graphs were analy sis of changes in profil es over time, which is the focus of taken of the beach to docum ent its im pact on the sand budget. this beach profil ing study. RE SULTS Data Cor rec tions Shore line Clas si fi ca tion Map Cor rec tions to the data set were made when obvi ous re- cording er rors were noted in plots of profil es. These errors in- Surfic ial mate ri als found in the Se bago Lake region in - cluded sign switches (+/-), deci mal place er rors and oth ers. The clude till, a mix ture of sand, silt, and clay; gla cial outwash , sand posi ti on of the water level re corded on the profil e served as a tool and gravel; silt and clay; and wetland de posit s, a mixture of silts to check reason able ness of the cor rected pro files. These types of and de cay ing plant ma te rial. A shore line clas si fi ca tion map of correc ti ons were made to the follow ing profil es: Halls 12, July 8 Sebago Lake (Johnston, 1998) identi fies the sedi ment types ex- 1994 (at 25.5 feet from pin, -1.75 changed to -0.75); Songo 7, posed along the shore of the lake (see Table 2). Septem ber 1, 1995 (sign change in ver ti cal incre m ent at 60 feet Till, a mix ture of sand, silt and clay, is the dominan t sedi - from pin), April 6, 1996 (verti cal incre m ents adjuste d in first two ment type found along the lake shore. The second most com mon readings to make them fit in sweep). Record ing er rors are sus- sedi ment type is the sand beach. Till and sand were depos it ed ei - pected in a few other profil es (Songo 7, July 5, 1995; Stan dish 3, ther di rectly or in di rectly from the late Wis consi nan gla cia tion. August 5, 1994; Sandbar 10, Novem ber 8, 1993 and Decem ber Till was de pos ited in di rect con tact with gla cial ice while sand

8 Beach Dynamics of Sebago Lake

Ta ble 2. at a verti cal posi ti on 3 feet below the pin. The change, de pend- Shore line Type Length ing on the slope of that por tion of the beach, could have re sulted (yards) from the addi ti on of a layer of sand a few inches to over a foot thick. The rela tive sig nifi cance of these changes is evi dent from Marsh ...... 7,240 ex ami na tion of the pro files. Sand beach ...... 26,030 Songo Beach, Site 3. A total of 49 profil es, collec ted be - Seawall behind beach ...... 8,370 tween Decem ber 1990 and May 1997, were analy zed for this Groins with sand in betwee n ...... 4,690 study. Profile s of Songo 3 docum ent the fol lowing events and Bluff behind sand beach ...... 7,470 pro cesses: Sand beach with boul ders...... 3,820 Till (sand, silt and clay) ...... 100,750 · Off shore-onshore profil e shifts in response to fal ling and Ar ti fi cial fill ...... 10,100 rising lake levels (see, for exam ple, Sum mer 1994, and Fig- Bed rock ...... 6,920 ure 5), accret ing up to 5 feet horizon tal ly to the dry beach as water levels drop in the fall, and eroding the under wa ter porti on of the profil e; was de posit ed by flowing wa ter out in front of the ice. The large · Ice-push ridges (see Spring 1995 and 1996 profil es). volum es of sedim ent avail able from local glacia l depos it s have · A major but transie nt erosion episode af fect ing the outer provided much of the mate ria l for the sand beaches around the beach betwee n Septem ber 1 and Oc tober 3, 1995 during lake. Fluvial and la cus trine pro cesses have moved the gla cial record low wa ter levels (Figure 11). The outer beach profil e mate ri als to their pres ent po si tion. Both the Songo River delta re turned to ap proxi mate pre-ero sion lev els by Oc to ber 11, and the Northwest River delta were formed by fluvial proces ses 1996 (shown on Sweep Zone and Fall 1996 plots). since gla cial time and com prise the largest vol umes of sand · A catastrophi c erosion event dur ing early Novem ber of around the lake. Bloom (1959) suggest ed that waves re worked 1996, and extre mely high wa ter levels , steep ened the profil e the delta sand along shore to create Cub Cove, Songo, Naples, consid era bly in the upper beach. Erosion of up to 10 feet and Witches Cove beaches at the state park. (horizon tal ly) at 3 to 4 feet (verti cally) below the pin, and up Shoreli nes modifie d by hum an ac tion make up the third to 3 feet (horizon tal ly) 1.5 feet below the pin, with deposi - most prevalent type. These include sand beaches with a seawall tion of this ma te rial in the outer part of the pro file (see Fall behind, groins with sand in betwee n, and ar ti ficia l fill. Bedrock 1996 profil es). This erosion episode re sulted in a zone of is exposed along al most 4 miles of lake shore. When added to- ex posed tree roots, left a cob ble layer along the beach, and gether, all sedim ent types that include sand make up alm ost 28 toppled some large trees (Figure 2). By May 16, 1997 there miles of shoreli ne. Total length of shoreli ne at Sebago Lake is was no evidence of recov ery , and there was addi ti onal loss approxi m ately 105 miles. The till-and bedrock -lined shore lines of sand from the mid and outer beach areas on that date. are the most re sis tant to erosion, while the sand, gravel and mud shorelines are eas ily reworked by wave acti on. Songo Beach, Site 4. A total of 44 profil es, collec ted be - Anal y sis of Pro files tween Decem ber 1990 and May 1997, were analy zed for this study. This site was not profil ed during the sum mer and early fall This sec tion descri bes events and proces ses docum ented at of 1996. The fol lowing obser va ti ons were made from profil e each profil e site. The appen dix contai ns plots of the profil es re - data col lected at Songo 4: ferred to in the text, where they are ar ranged by beach and site, in clockwise order around the lake. Within the profil e set for each · Off shore-onshore shifts in the profil e with changing wa ter site, profil es are arranged in chronologi cal order by sea son, with lev els. For exam ple, an offshore shift in the profil e occurred any addi ti onal profil es (com pari son of all fall profil es, for exam - in re sponse to fal ling wa ter lev els be tween Sep tem ber 1 and ple) at the end. Figure 3 shows loca ti ons of the profil e stati ons. Octo ber 3, 1995, with progra dati on of the dry beach 8 to 10 Other data referred to in the fol lowing descri pti ons is contai ned feet (hori zon tally) at 5 feet (ver tically) be low the pin. Ris - in Figures 6-12 (lake level curves for 1990-1996), Figure 4 ing wa ter levels in the spring shifted the profil e back (see (beach and wave ter minol ogy) and Figure 5 (cartoon of ideal ize d April 6, 1996 profil e). beach changes due to changing lake levels ). These figures are · A ridge run nel sys tem was pres ent at this site in the spring placed at the end of the text. and late fall of 1996 (see Figure 4, and Spring and Fall 1996 For conven ienc e, the magni tude of changes are gener all y pro files). descri bed in their horizon tal dim ension, and refer ence d to a ver- · A large ice-push ridge was present in the spring of 1997 (see ti cal posi ti on below the survey marker or pin. For exam ple, “ac - April 9, 1997 profil e, and Figure 22). The long, sinuous cre tion of 5 feet hori zon tally at 3 feet ver tically be low the pin” ridge extende d discon ti nuously for most of the length of descri bes the change along a hori zontal line through the profil e Songo Beach on that date.

9 R. A. Johnston and M. N. Mixon

· A catastrophi c erosion event in Novem ber 1996, during ex- · Transie nt changes re lated to rising and falli ng lake levels . tremely high water lev els, steep ened the pro file con sid era - · Mi nor erosion af fect ing mid beach during rising wa ter lev- bly. Con sid er able ero sion also oc curred be tween April and els, Spring 1996 (see Songo 7, also). May 1997. By Decem ber 1996 the profil e had lost 5 feet horizon tal ly in a zone from 1 to 3 feet be low the pin. By Songo Beach, Site 7. A total of 48 profil es, collec ted be - May the loss had been increa sed to 11 feet horizon tal ly 3 tween Decem ber 1990 and May 1997, were analy zed for this feet below the pin (see Fall 1996 through Spring 1997 pro- study. This profil e loca ti on, at the far west end of Songo Beach files). on the spit, shows a high degree of variabil ity com pared to other sites include d in the study, with a sweep zone around 20 feet Songo Beach, Site 5. A total of 50 profil es, collec ted be - wide in the mid beach. Elem ents of some of this variabil ity are as tween Decem ber 1990 and May 1997, were analy zed for this fol lows: study. The Songo 5 profil e passes be tween two stumps, with roots exposed on the side fac ing the lake, 20 feet from the survey · Pro gra da tion of the beach face ac com pa ny ing ris ing wa ter pin. Some of the pro file vari abil ity at this dis tance from the pin is level, from deposi ti on of sand above the water line, as docu- due to variable placem ent of the survey rod on the stumps or mented in Figure 30 in Dickson and Johnston (1994), during down betwee n the roots to the sand below . Since the erosion Spring 1992. This is also seen in Spring 1994 (May 3 to event of Novem ber 1996, these stumps have been under m ined May 31); Oc tober 1994, extend ing through July 5, 1995; con sid era bly. Pro files of Songo 5 show a progres sive loss of and Spring 1996 (April 6 to June 6). sand from the outer beach over two consec uti ve years, a major · Beach pro gra da tion ac com pa ny ing fal ling wa ter level. erosion event af fect ing the upper beach during Novem ber 1996, Sand is depos it ed in the swash zone as the wa ter line re - and other smaller or more tran sient fea tures. These are de scribed cedes. This is seen betwee n August 3 and Septem ber 9, be low: 1994, and betwee n July 18 and August 13, 1996. · Erosion of the beach face occurred be tween June 6 and July · The sweep zone plot for Songo 5 shows a progres sive loss of 18, 1996, with a slight increa se in water level. There were sand from the outer beach, beyond approxi m ately 90 feet several days of high winds during that time peri od. from the pin. This loss occurred be tween July 1995 and Oc- · Accre ti on of al most 1/2 foot of sand to the top of the spit oc- tober 1996. Profil es from earli er years do not go beyond 90 curred dur ing the very high wa ter levels of late Oc tober and or 100 feet from the pin, but some from 1992 and 1993 sug- early No vember , 1996. The lower beach face also pro- gest the outer beach may have been at a higher level. The graded ap proxi mately 5 feet. By early De cem ber 1996 the profil e of Decem ber 1996 shows a rever sal of the trend, top of the spit had reached its highest point since profil ing when a major erosion event af fect ing the upper beach began at this site in fall 1990, but the spit had also eroded dumped sand onto the outer beach. Future profil es will horizon tal ly some 10 feet in the upper beach face, a result of show whether or not this rever sal is tem porary . the high winds of Novem ber 8-9, 1996. This eroded mate - · The erosion event of Novem ber 1996 that af fected other rial was re de pos ited in the re main der of the pro file, cre ating south-fac ing beaches also af fected Songo 5. The upper to a stranded ridge and runnel (or long shore bar and trough) as middle beach lost up to 10 feet (horizon tal ) of sand be tween the wa ter fell, and caus ing con sid er able pro gra da tion of the the No vember 6 and Decem ber 3 pro files (at 3 feet ver tically outer beach. below the pin). The outer beach gained approxi m ately the · The low slope ter race in the outer porti on of the profil e ap- same amount at that time. A ridge and runnel de veloped. pears to os cillate slowly be tween 3 and 4 feet (ver tically) be - The pro file re mained rela tively stable over the win ter, but low the pin. In Fall 1992 and 1993 it appears to have been betwee n April and May of 1997 the upper beach eroded an- po si tioned around -3.25 feet. By Fall 1994 it had moved other 4 to 5 feet horizon tal ly at 1.5 feet below the pin. The down to around -3.5 feet. In Spring 1995 it moved down sand was depos it ed in the runnel of the mid beach, smooth- fur ther, ap proach ing -4 feet. Since then it has been rising. ing it out. · A small ridge and runnel were present on April 6 and were Halls Beach, Site 11, Tas sel top Park. A total of 36 pro- gone by April 18, 1996. files, collec ted be tween July 1991 and May 1997, were ana lyzed · An ice push ridge was docum ented on April 6, 1996 and was for this study. Some of the vari abil ity at this site is due to a large reworked by April 18 (Figure 22). drift wood log approxi m ately 10 feet from the pin, a large boul- · Dur ing a pe ri od of rela tively sta ble wa ter lev els in the fall of der at approxi m ately 30 feet, and vandal ism (rem oval) of the pin 1994, the mid beach prograded 10 feet in the vi cinit y of the in 1993. Profil ers have variably placed the profil e rods on top of, wa ter line (4 feet ver tically below the pin). The sub merged in front or in back of, or to the side of the log and rock, creat ing a beach eroded during this time. This gain was graduall y lost consid er able spread in the data at these loca ti ons. Some of the over the fol lowing spring. out liers on the sweep zone are due to vandal ism (rem oval) of the

10 Beach Dynamics of Sebago Lake

pin during Spring 1993. Despit e vandal ism and lack of hori zon- - An onshore shift occurred be tween the April and tal con trol, the site is fairly sta ble. May profil es, 1996. - Accre ti on and erosion oc curred with stable water · Most pro file changes at the site are tran sient ad just ments to level in the fall of 1994, pro grad ing the beach at the rising and falli ng water levels . For exam ple: water level and eroding it below . - Erosion of outer beach with falli ng water level in · Ice push ridges were present in the spring of 1992 and 1995. Sum mer 1994 (offshore profil e shift). · Dickson and Johnston (1994) noted the devel op m ent of - Accre ti on of sand to dry beach during falli ng water notches and erosion of 5 feet (horizon tal ) of sand from the level, Sum mer 1995 (offshore profil e shift). beach face in the fall and winter of 1992-1993, with return - Accre ti on above wa ter line, erosion be low, with ris- of the profil e to its original shape by May 1993. ing lake level, Spring 1996 (onshore profil e shift). · A signifi cant accre ti on event occurred be tween the August - Ac cre tion at wa ter level and ero sion below also oc - 13 and Octo ber 30, 1996 profil e dates. The middle to outer curred with sta ble wa ter level in Fall 1994. beach area was extende d by 5 to 7 feet (horizon tal ). Some of · An erosion event rem oved 6 feet of sand from the this mate rial was re worked and re de pos ited higher on the mid-beach betwee n May 3 and 31, 1994. Cli ma to logi cal profil e, above the water level, by the next profil e date, No- data show aver age winds excee ding 10 mph out of the south vem ber 22, 1996. On Oc tober 30, beach profil ers observe d and southwest on May 7, 9, and 11-13. The profil e had re - chan nels cut in the beach and small del tas, as de scribed for covered by July 1994. the Halls 11 site. This profil e may have crossed the edge of · Promi nent ice-push ridges were present in April 1992 and one of these small deltas. 1995. · Com pari son of a sweep plot of the 1991 to 1993 data with a · Accre ti on of 5 to 6 (hori zontal ) feet of sand to the mid and sweep plot of the 1994 to 1997 data suggest s a steep en ing of outer beach areas betwee n Fall 1995 and Spring 1996, per- the profil e has occurred in the upper beach area. This trend haps re lated to very low water levels of 1995 followed by may have been re versed, at least tem po rar ily, by the events water level rise in Spring 1996. of Fall 1996. · Slight ero sion fol low ing Oc to ber 20- 22 rains fol lowed by accre tion above and below wa ter level during Novem ber Frye Is land, Site 2. A to tal of 7 pro files, col lected be tween 1996; at a time when south-fac ing beaches suffered major Octo ber 1993 and Novem ber 1996 were analy zed for this study. erosion during high water , with winds from the south. Fol - Sig nificant fea tures of this pro file site are : low ing the Oc to ber 20-22 rain storm, profil ers observe d channels cut through the beach form ing small deltas out in · An ATV track pres ent in the upper beach, Octo ber 1995; ac- the lake. The heavy rains had flooded a low-ly ing area be - creti on of 13 feet (horizon tal ) below the water level, in the hind the vege tated berm (east of it). The flood water s had outer beach, betwee n Octo ber 1993 and Oc tober 1995. broken through the berm, cre ating the channels across the · Erosion from upper , mid, and outer beach areas . Loss of up beach. The early Novem ber profil e did not cross any part of to 12 feet (horizon tal ) of sand from most of the centra l por- the small delta nearby. The early De cem ber ac cre tion re - tion of the profil e, and de velop m ent of an ero sional notch sulted from rework ing of the delta sands. Novem ber field betwee n Octo ber 1995 and by Spring 1996. The notch most notes indi cat e the channels were be ing filled in by the ac- likely formed in the fall of 1995 during the long pe riod of creting sand. low lake lev els. · The May 1997 profil e shows devel op m ent of a runnel/ ridge · An erosion event in the fall of 1996, with devel op m ent of a or long shore bar/trough feature . notch and loss of up to 5 feet of beach (horizon tal ) within 15 feet (horizon tal ) of the pin; and deposi ti on of the mate ria l in Halls Beach, Site 12, Tassel top Park. A total of 37 pro- the mid and outer profil e, re sult ing in up to 8 feet of accre - files, collec ted be tween July 1991 and May 1997 were ana lyzed tion there. This event occurred be tween Octo ber 30, 1996 for this study. As at site 11, this pro file is rela tively sta ble, with and Novem ber 22, 1996, during very high water levels , most beach changes relat ed to onshore -off shore pro file shifts most proba bly during the epi sode of high winds out of the that accom pany rising and falli ng lake levels . Some exam ples of south in early Novem ber that af fected other south-fac ing these shifts, and other fea tures of the pro files are de scribed here: beaches. Some of the ero sional notch in the upper profil e had filled in by May 1997, but the ac creti on in the mid and · Some exam ples of the onshore -offsho re shifts with ris ing outer profil e had been eroded to beyond the profil e lim its. and falling lake lev els are: - Accre ti on above the shore line, erosion be low with Frye Is land, Site 3. A total of 7 profil es, col lected be tween fal ling lake levels , result ing in an offshore pro file Octo ber 1993 and Novem ber 1996, were ana lyzed for this study. shift, betwee n late July and Septem ber 1995. This site shows the follow ing feature s:

11 R. A. Johnston and M. N. Mixon

· A net gain of sand in the outer beach, with a gain of up to 13 · By Spring 1994 and Spring 1995, most of the sand gained feet of beach (horizon tal ), at 7 feet below the pin from Octo - during the previous sum mers had been eroded. This is con- ber 15, 1993 to Octo ber 4, 1996. sis tent with wind data for the Sebago area (Port land Jetport) · Simi lar to Frye Island 2, it also shows the Fall 1996 erosion that shows strong winds out of the north and west are typi cal event with devel op m ent of a notch and loss of approxi - in the spring. mately 5 feet of beach (horizon tal ) within 15 feet (horizon - tal) of the pin. The eroded mate ria l was depos it ed in the mid Stan dish Boat Launch, Site 1. A total of 24 profil es, col- beach area, re sult ing in some minor accre tion there. Like lected be tween Octo ber 1993 and Novem ber 1996, were ana - site 2, some of the ero sional notch in the upper profil e had lyzed for this study. The Stan dish 1 pro fil ing lo ca tion filled in by May 1997, but the ac creti on in the mid profil e ex pe ri enced al ter nat ing ac cre tion and ero sion of up to 10 feet area had been eroded to beyond the profil e lim its. Up to 10 (horizon tal ) in the mid-beach area over the study pe riod. The feet (hori zon tal) of beach was lost from the mid beach sec - sweep zone is compact, 10 feet wide in the mid beach area and up tion of the profil e. to 15 feet wide at the outer limits of pro fil ing. The major profil e changes at this site are: Rock wall Beach, Site 7. A total of 23 profil es collec ted be - tween Octo ber 1993 and Novem ber 1996 were analy zed for this · The beach prograded 10 feet dur ing pre dominantl y stable study. The sweep zone contai ns two sets of notches, and an out- and rising wa ter levels during Fall 1993 and Spring 1994, lier on ei ther side of the main sweep in the mid beach area. The and was sta ble unti l the Septem ber 1994 profil e, when up to major pro file changes at this site are: 10 feet was lost from the upper , middle, and outer beach ar- eas (NWS data show winds out of the north betwee n 13 and · Notches in Fall 1993. It is not clear whether the Fall 1993 20 mph for 21 hours on August 5-6, and again out of the notches are ero sional fea tures, or de po si tional fea tures since north be tween 9 and 14 mph for 15 hours on August 23, these were the first pro files col lected. They were filled in by 1994). The eroded profil e rem ained sta ble through the fall. June 1994. · By the spring (May) of 1995 the mid and outer beach had · Sum mer 1994 profil es show erosion of the outer beach be- pro graded again 10 feet, during stable to slowly rising lake tween the July and August profil es, fol lowed by accre ti on in levels . These gains were lost dur ing the sum mer, regaine d the same porti on of the profil e betwee n August and Septem - in the outer beach (Septem ber 1995 pro file), and then lost ber. New “notches” that first ap peared in the Sep tem ber again (Octo ber 1995 profil e). The beach then rem ained sta - 1994 profil e and persis ted through the spring of 1995 are ac- ble, with only minor changes near the shoreli ne, unti l Fall tu ally the re sult of ac cre tion of a flat-topped layer of sand or 1996. berm up to 8 feet wide (hori zontal ). Both erosion and accre - · Some erosion (4 feet, horizon tal ) in the mid beach area is tion occurred dur ing falli ng lake levels in the sum mer of docum ented betwee n the Octo ber 1 and Novem ber 5, 1996 1994. This was followed by further accre ti on and smooth- pro files. During this time peri od, the lake level rose 2.5 feet ing of the profil e betwee n May and June 1995 when water due to an 8- to 12-inch rainfal l on Oc tober 20-21. lev els rose slightly. · Other pro file changes are smaller and more tran sient re - Stan dish Boat Launch, Site 3. A total of 23 profil es, col- sponses to lake level changes. lected be tween Novem ber 1993 and Novem ber 1996, were ana - lyzed for this study. The Stan dish 3 site was very stable over the Rock wall Beach, Site 8. A total of 23 profil es collec ted be - study peri od. The sweep zone is straight and com pact, up to 7 tween Octo ber 1993 and Novem ber 1996 were analy zed for this feet wide for most of its length, if a sin gle out lier (sus pected re - study. The sweep of this site shows a com pact zone near the pin cording er ror) is ignored. The major profil e changes at this site that expands slightly with distanc e from the pin. There are no are: out liers. Offshore -onshore profil e shifts with falli ng and rising lake levels account for most of the profil e changes at this site. · Beach profil es were straight and stable from Novem ber The major pro file changes at this site are: 1993 through Fall 1994, except for a small amount of ero- sion from the outer beach (Septem ber to Octo ber), and ac- · Sum mer 1994, 1995, and 1996 show ac creti on to the mid creti on just above the water level (offshore shift of profil e beach area be tween the August and Sep tem ber pro files, with fal ling water level). A record ing er ror is suspect ed in with fal ling lake lev els. The ac cre tion is most sig nificant in the August 1994 profil e and the large amount of accre ti on 1994, when 8 feet (horizon tal ) of beach was added to the shown there may not be real. The profil e rem ained sta ble pro file. over the winter . · By Octo ber 1994 a strong ero sional notch had devel oped in · As water levels rose slightly in the spring of 1995 (May to the newly ac creted beach at 3.5 feet (ver tical) be low the pin. June) the enti re beach prograded up to 6 feet, and the profil e

12 Beach Dynamics of Sebago Lake

was smoothed. It then rem ained sta ble, with only a sugges - land Wa ter Dis trict weather data, collected in their sta tion in tion of offshore -onshore shifts with fal ling and rising lake Stan dish, show a few hourly ranges excee ding 8 mph from levels for the rem ainder of the study pe riod. the appro pri ate direc ti ons. None of these wind records show sustai ned high winds capa ble of produc ing the major Sandbar Beach, Site 10. A to tal of 24 pro files col lected erosion seen in this profil e. This suggest s that winds from betwee n Octo ber 1993 and Novem ber 1996 were analy zed for other direc tions gen er ate waves that are re fracted into the this study. This site has an S-shaped pro file that is rela tively sta - north end of Harm on Beach, causing the erosion observe d at ble. All variati ons are minor , transie nt onshore -off shore shifts the Bar ton resi dence. re lated to lake level ris ing and fal ling (There may be record ing errors in profil es B and C). Harmon Beach, Banks Residenc e. A total of 12 profil es collec ted be tween Septem ber 1993 and May 1997 were ana- Sandbar Beach, Site 12. A to tal of 24 pro files col lected lyzed for this study. Over the study pe riod this beach was rela- betwee n Octo ber 1993 and Novem ber 1996 were analy zed for tively sta ble, with an S-shaped pro file. No table fea tures are an this study. This site has been extre mely sta ble over the study pe- ero sional episode shown in the August 1994 profil e, a signifi - riod. The sweep zone is com pact and straight. The stabil ity of cant but short-lived ero sional event shown in the Decem ber this site and Sandbar 10 may be due to their protec ted loca ti on. 1996 profil e, and a pronounced ice-push ridge in the May 1997 The only wind di rec tion ca pa ble of gen er at ing sig nifi cant waves pro file. All of these fea tures were tran sient. is northeas t. During the study pe riod, there have not been many large storms with strong winds out of the northeas t. Har mon Beach, Straw Residence #2. A total of 12 pro- files, collec ted be tween Septem ber 1993 and May 1997, were Harmon Beach, Barton Resi dence. A total of 12 profil es analy zed for this study. The beach has an S-shaped profil e with a collec ted be tween Septem ber 1993 and May 1997 were ana- fairly nar row sweep zone. Vari abil ity is due to: lyzed for this study. In contra st to the other sites profil ed at Har- mon Beach, this one showed quite a bit of variati on, espe cia lly in · Ero sion (ap proxi mately 7 feet, hori zon tal) be tween the 1996. Confu sion about the start ing loca ti on of profil es for this August 1994 and Septem ber 1995 profil es (Oc tober 1993 profil e loca ti on led to use of two dif ferent starting points, the profil e eliminate d from analysis due to suspect ed record ing northeas t cor ner of the house, and a steel pipe east of the house. error). By May 1996, it appears some of this sand had re - Profil es which started at the northeas t cor ner of the house, but turned. did not refer ence the pipe in the field notes, have not been used. · During the spring of 1996 (May to June), approxi m ately 4 The major pro file changes at this site are: feet (horizon tal ) of erosion oc curred just below the water line, at 3 feet be low the pin (ver tically). As wa ter lev els fell · This site was sta ble through the Septem ber 1995 profil e. in the sum mer and fall, accre ti on and erosion oc curred, The next usable profil e, in August 1996, showed erosion shift ing the profil e offshore slightly. The rapid rise in water from all areas of the profil e. Maxim um erosion was at the level of Oc tober 1996 result ed in progra da ti on of the outer end of the profil e, where 18 feet (hori zontal ) of sand were beach, restor ing it to its posi ti on of the fall of 1993 and lost. By Octo ber 1996 sand had returne d to the outer porti on 1994. of the profil e. · Betwee n Novem ber 6 and De cem ber 3, 1996 a major ero- Long Point Beach, Sunning dale Resi dence. A to tal of 11 sion event af fected the upper part of the profil e, expos ing an profil es, collec ted be tween Septem ber 1993 and July 1995, were ad di tional foot (ver tical) of the steel pipe that serves as the analy zed for this study. The sweep zone for this site is gentl y profil e start ing point. The eroded mate ria l was re de pos ited S-shaped, com pact in the upper beach, spreading graduall y to 8 in the outer porti on of the profil e, where a gain of approxi - feet wide at 10 feet below the pin, in the mid porti on of the pro- mately 15 feet (horizon tal ) occurred. By Spring 1996 this file; to about 15 feet in the outer beach. Sources of variati on are: mate rial had been re moved from the outer pro file area, and a small amount of sand had returne d to the upper part of the · Approxi m ately 5 feet of progra da ti on with fal ling water pro file. The Novem ber sum mary of Local Cli ma to logi cal level in Fall 1993. Data for Portland , Maine, and Port land Wa ter Dis trict wind · Tran sient changes that are not clearly related to wa ter lev els, data were scanned for high wind occur rence s from fetch di- in 1994 and 1995. recti ons af fect ing Harm on Beach during this time peri od. The Portland data shows one reading (3-hourly readings) on In ter an nual Trends each of Novem ber 6, 7, and 9 of moder ate to high winds, from a direc ti on that could reach Harm on Beach. These Over most of the time peri od studied (3 to 6 years betwee n times bracket the time of high southerl y winds of Novem ber 1990 and 1996), most beaches were sta ble. The excep ti ons are 8 and 9, that af fected the south-fac ing beaches. The Port - descri bed below (Large Changes, Fall 1996). Most of the

13 R. A. Johnston and M. N. Mixon

changes ob served can be char acter ized as tran sient sand move - the more recent data. The mid and outer beach porti ons of ment result ing in onshore and offshore profil e shifts (Figure 5), the sweep plots over the same time pe ri od do not show the in re sponse to rising and falli ng lake levels . Other changes docu- same trend. The upper beach porti on of Halls 11 is more mented include sand ridges and piles in the upper beach that vari able, due to a large driftwood log and a large boulder were bulldozed or de posit ed by wind-blown slabs of ice as they that the profil e passes over. A trend show ing a steep en ing of melted in spring; and erosion of the outer beach with devel op - the upper beach would be more dif ficult to recog nize at ment of ero sional notches at two sites during very low lake levels Halls 11. in 1995. Similar erosion and notching were observe d in the fall of 1992 (Dickson and Johnston, 1994). Although some of these Long-term grad ual changes sug gested here, es pe cially at changes involved rela tively large volum es of sand, none of them Frye Is land, need con fir ma tion from ad di tional pro fil ing. persis ted longer than a year. The sta bil ity of most pro file sites is seen from the nar row - Large Changes, Fall 1996 ness and com pact ness of the sweep zone plots, espe cia lly the up- per beach sec tions near the pins; and from the lack of di recti onal The south-fac ing beaches (Songo and Frye Island) and one change, or long term trends in the beach profil e plots. Evi dence east-fac ing pro file site (Bar ton Residence at Har mon Beach) ex - of sta bil ity of the beaches can also be seen in the plots that show peri ence d signifi cant erosion to the upper beach areas during the all August, or other late sum mer or fall profil es for each site. fall of 1996. This erosion oc curred af ter sev eral years of stabil ity These are collected at a time when lake lev els are near their low or gradual change. At the same time, the Halls 12 site (south - for the year, so pro files are long, and are gen er ally free from the west-fac ing) ex pe ri enced a sig nifi cant amount of ac cre tion. The ef fects of any re cent storms. north ern Halls beach site ex pe ri enced mod er ate ero sion, fol- While most beaches were sta ble over the study peri od, a lowed by accre tion. few beaches showed evidence of gradual, long term change. A com bina ti on of events led to this major episode of shore- These inter annual changes were observe d by com paring yearly line change. On Oc tober 20-22, 1996 a major precipi tati on event late-sum mer profil es. Sites which showed progres sive sand gain occurred in southern Maine. Precipi tati on in the form of rain fell or loss over the study pe riod are de scribed here: in exces s of 8 inches in the Sebago Lake basin over a peri od of about 12 hours (FEMA, 1996). This rec ord-break ing rain fall re - · Songo 5 appears to show a progres sive sand loss from the sulted from a blocked northeas ter over southern Maine being fed outer beach profil e. Cover age of the outer beach is better in tropi cal moisture from hurri cane Lili, locat ed 900 miles south - more recent profil es. The profil es from 1992 and 1993 seem east of Port land. A strong high pressure system over Labra dor to suggest the outer beach may have been at a higher level kept the low pressure system from moving out of the region . The then. Profil es be tween July 1995 and Octo ber 1996 clearly Se bago Lake wa ter level rose 2 feet from Octo ber 16th through show the trend of sand loss from the outer beach. The trend Octo ber 23rd. The overall rise in lake level from Octo ber 16th was re versed by the Decem ber 1996 profil e due to the ero- through Novem ber 12th was 3.6 feet. sion event in Novem ber, in which mate ria l eroded from the While lake lev els were at or near this full pond level, a low upper beach was depos it ed on the outer beach. Future pro- pres sure sys tem track ing up the St. Law rence River val ley fil ing will show whether sand loss from the outer beach re- brought strong southerl y winds to the lake on Novem ber 8th and sumes fol lowing this event. 9th (FEMA, 1996). The com bina ti on of the strong south erly · The low slope ter race in the outer porti on of the profil e of winds and a high lake level allowed waves to reach and erode the Songo 7 gained sand during 1995 and 1996. Prior to that it upper beach areas , expos ing tree roots and toppli ng trees along may have lost sand for a few years in a row, al though these the shore line at the Songo and Frye Island sites (Figure 2). ear lier pro files are more dif fi cult to in ter pret be cause they The steepen ing of upper porti ons of profil es that oc curred are not as long as the more re cent pro files. The sand loss at at the Songo sites in the fall of 1996 conti nued into the spring of Site 5 (1995- 1996) and sand gain at Site 7 (1995- 1996) sug- 1997, when further erosion oc curred on these beaches. Ex po - gest a long shore current may be transport ing sand to the sure of tree roots at Songo and Frye Island beaches, and loss of west. the front line of trees along Songo beach, suggest that further · Frye Island 3 shows net gain of sand in the outer beach over changes in the profil es are likely before a new equi librium pro- the three-year study peri od (Octo ber 1993 to Octo ber 1996). file is reached. Becaus e the erosion oc curred so high up on the Site 2 at Frye Island does not show any clear trend. Collec - beaches, it is not likely that sand will be returne d to the eroded tion of data over a longer time peri od, and exam ina ti on of upper profil e. Conti nued profil ing at these sites will show the data from the other two Frye Island sites not include d in this ex tent of ero sion that even tu ally will re sult from the weather study, might confirm this trend and show whether accre ti on events of Octo ber and Novem ber 1996. or erosion is occur ring at the other Frye Is land sites. The ac cre tion event docu mented at the Halls 12 site ap - · Separat e sweep plots of 1991 to 1993, and 1994 to 1997 data peared to be a result prim aril y of the rain storm. Field evi dence for Halls 12 show the profil e is steeper in the upper beach in suggest ed the accre ti on was the result of erosion from channels

14 Beach Dynamics of Sebago Lake

cut into the beach and deposi ti on of the sand in small deltas , sion of the under wa ter porti on of the profil e, as at Songo 3, Sum - which profil e 12 crossed. Subse quent accre ti on at Halls 11 mer 1994. Falli ng water levels also swept sand offshore, in a re- could have re sulted from re dis tri bu tion and long shore transport verse of the spring beach-building proces s (Halls 11, Sum mer of mate ria l from site 12 to the south to site 11 to the north. 1994). The major erosion docum ented at the Bar ton Residence at Ero sion and de vel op ment of ero sional notches were ob- Har mon Beach betwee n Novem ber 6 and De cem ber 3, 1996 was served in some late sum mer and fall profil es. Sum mer and fall of unex pect ed. The beach facing di recti on is east, the winds asso ci - 1995 was a time of prolonged low lake levels (Figure 11). Songo ated with the low pressure system of Novem ber 8-9 were out of 3 de vel oped an ero sional notch, and suf fered a major erosion the south, and there was an appar ent lack of signifi cant wind episode to the outer profil e betwee n the Septem ber 1 and Oc to- from the east, northeas t, or southeas t during the profil e month. ber 3 profil es. This outer, low-slope porti on of the profil e lost 1 Wave re frac tion may play a role at this site, allow ing waves from to 1.5 feet ver tically (60 feet hori zon tally, at 8.5 feet be low the other direc ti ons to reach the north end of Har mon Beach, caus ing pin) during this month. This erosion turned out to be tem porary . the erosion observe d at the Bar ton resi dence. In addi ti on to the The sand graduall y re turned over the next year, return ing al most southerl y winds, there was over a day of conti nuous moder ate ly to its pre vious posi ti on by Oc tober 11, 1996. high wind out of the north on Novem ber 26-27. The northerl y Also during the 1995 low lake levels (betwee n Octo ber 13, winds would have had a long fetch. 1995 and May 8, 1996 profil es), Frye Island 2 de veloped an ero - The records of stabil ity at these sites, followed by the sud- sional notch at the po si tion of the low wa ter level in the fall. den large changes asso ci ate d with these fall storms, suggest that Lake lev els sometim es be come stable for a pe ri od of a few large, per manent shifts in profil es are not the norm, but are the re- months, espe cia lly in the fall. When this happens, progra da ti on sult of extre me events that oc cur infre quentl y. of the dry beach just above the water line, and erosion just below often oc cur. Exam ples of this are seen in the fol lowing profil es: Sea sonal Pat terns Songo 5 and 7, Halls 11 and 12, profil es of Oc tober 5, Oc tober 25, and Decem ber 6, 1994. Early spring obser va ti ons and beach profil es have shown Ex cept for the de vel op ment of tem po rary ice-re lated fea - ice-depos it ed ac cum ula ti ons of sand on the dry beach and in tures, profil es do not change much over the winter when the lake shal low wa ter at sev eral beaches. These sand ridges have been is fro zen (see Songo 3 and 5, Last Profil e 1996, First Profil e ob served at Songo, Halls, Har mon, and Long Point beaches. A 1997). striking exam ple of an ice-depos it ed sand ridge was profil ed and All of these sea sonal effects ob served were tran sient, per - pho to graphed at Songo beach in 1997 (Figure 22). During the sist ing for a year at most. winter of 1997, a spectac ular wind-blown ice pile de veloped in front of a residence at Wards Cove. Ice-re lated sand de pos its are Beach Fac ing Di rec tion quickly re worked with rising wa ter levels in the spring. Beaches gen er ally ex pe ri enced ac cre tion dur ing mid and This study include d beaches around the enti re perim eter of late spring when lake levels rose. The com bina ti on of rising wa - the lake, so that all facing di recti ons could be evaluate d. Sites ter level and aver age waves moved sand onshore from farther out from all facing di rec tions can be char acter ized as predomi nantly in the profil e (Halls 11, Last Profil e 1995, First Profil e 1996; stable, with tran sient changes re lated to wa ter lev els. The north - Songo 7, Spring 1994, for exam ple), and ac creted it to the beach west-fac ing Rock wall sites showed a large amount of accre tion face. Ris ing wa ter lev els also re ar ranged the sand, erod ing it with fal ling wa ter lev els in the late summer , and loss of the ac - from just be low the water line and de posit ing it just above the creted sand by the fol lowing spring (May) profil es. The prevail - water line (Halls 11, Spring 1996), cre ating an onshore shift in ing west er lies may play a role in the strong sum mer and early fall the profil e, and progra da ti on of the dry beach. In some cases the ac cre tion at this site. Strong north west winds in the early spring re ar rang ing and ac cre tion go on si mul ta ne ously (Songo 7, may play a role in the sub se quent loss of the accreted mate rial Spring 1992, Figure 30, Dickson and Johnston, 1994). (Figure 10 and Table 1 in Dickson and Johnston, 1994). Erosion was also observe d during rising wa ter levels in the Ice de po si tional feature s have been observe d in the field at spring at some beaches. Some exam ples are seen in Spring 1996 all beaches ex cept Rock wall, Sandbar , and Stan dish. These sites at Songo 5 and 7, and Spring 1995, Songo 4. The erosion at may not have been profil ed early enough in the spring to docu- Songo 4 and 5 was minor . At Songo 7 approxi m ately 10 feet ment ice fea tures. Ice fea tures are ex pected to occur at all sites to (horizon tal ) was lost 1 foot below the pin. High winds may have some degree, but are also depend ent on wind di recti on during the played a role in the springti me erosion. There were some days on time of ice break- up. which higher than aver age southerl y winds were recorded dur- The largest changes docum ented in this study were asso ci - ing the month in questi on. ated with a low pressure system moving up the St. Lawrence val- Most beaches prograded as lake levels fell during sum mer, ley, bringing southerl y winds during a time of high lake levels in and profil es shifted lake ward (Rock wall 8, Sum mer 1994). Pro- the fall of 1996. It af fected pri maril y south-fac ing beaches. A grada ti on of the dry beach was someti mes accom panied by ero- northeas ter during high water levels could have caused erosion

15 R. A. Johnston and M. N. Mixon

of the same magni tude at beaches on the west ern shores of the Fal ling wa ter lev els can also sweep sand off shore, in a re verse of lake. A longer peri od of data col lec tion may docum ent such an the spring beach-building proces s (Halls 11, Sum mer 1994). event in the fu ture. Water lev els in the lake are not an ticipated to The lowered lake levels of fall expose the outer porti ons of be that high during the storm sea son with the new lake level rule beach pro files to the effects of waves, sometim es re sult ing in curve (FERC, 1996). erosion of the outer beach and devel op m ent of notches in the The study is bi ased with respec t to facing di recti on in that profil es (Fall 1992, Songo and Halls beaches, Dickson and John- the most data has been collec ted over the longest time peri od on ston, 1994; Fall 1995, Songo 3; Frye Island 2, Fall 1995). Songo the south-fac ing Songo sites and the southwest -fac ing Halls 3 expe ri ence d a loss of over a foot of sand (verti cally) from the beach sites. outer beach during late sum mer to Fall 1995, but had nearly re- gained it a year later. Lake cur rents Lake lev els sometimes rise slightly in the late fall to early win ter bef ore the lake freezes. Ex cept for ice-re lated fea tures, Mul tiple pro file sites are in cluded at each beach stud ied so profil es do not change much over the winter when the lake is fro- that changes along a beach can be docum ented. Evi dence of pos- zen. si ble long shore cur rents were seen be tween Songo 5 and 7, and All of these seasonal ef fects, includ ing fall erosion and de- Halls 11 and 12. The profil e changes that suggest the acti on of velop m ent of notches, are transie nt, and rarely per sist beyond a long shore current s, loss of sand from a porti on of the profil e at year. Most loca ti ons have a bal anced sweep zone that exhib it s one site and gain at another , are subtl e and have not been docu- sea sonal dy nam ics with an in ter an nual sta bil ity. mented over long time peri ods. More profil ing data is needed to In contra st to the regular sea sonal oscil lati ons, storms can evalu ate the role of long shore currents. produce larger long-lasting changes in beach profil es. A storm occur ring during high water gives storm waves acces s to the up- Beach dy nam ics per beach, contri butes new mate rial to the ac tive beach, steepens the upper beach profil e, and leads to further changes in the pro- The lake freezes over each winter for about 3 months of the file un til a new equilib rium pro file is reached. year, from January through March. While frozen, the shore line is protec ted from batter ing by winter storm waves. Slabs of ice, however , can be blown by winds up onto the shore, plowing sand CON CLU SIONS in front of them. Melt ing ice piles along the shore line can leave ridges of sand as they melt in the spring, both from the sand Beaches were stable over the study peri od un til the Songo plowed up, and from sand that is frozen into the ice and re leased and Bar ton beaches ex pe ri enced cata strophic changes in the fall dur ing melting. These fea tures were ob served in sev eral years, of 1996. These changes were the result of an unusual com bi na- but striking exam ples were seen in the spring of 1997 at Songo tion of weather events, but illus tra te the role of high lake levels in Beach (Figure 22). During the winter of 1997, a spectac ular ice provid ing storm waves acces s to new ground. Although this pile devel oped at Wards Cove (not a profil e site in this study). event took place in the fall, com para ble dam age could occur at As the spring thaw occurs , the lake level rises between 2.5 any time lake lev els are high and on any beach where the combi - and 5 feet, reaching an annual maxim um usuall y in May. As the nati on of wind di recti on, speed, and fetch allow de velop m ent of lake level rises, ice push features are quickly re worked; ris ing storm waves. Relati vely minor erosion events were docum ented lake level sweeps sand onshore (Halls 11, Last Profil e 1995, during Spring 1994, and the late spring/early sum mer of 1995; First Profil e 1996), accret ing it to the beach face; and the sand in both events at times of high water and strong winds. the profil e is rear ranged by erosion from just be low the water Erosion that oc curs at low lake levels , even of large vol- line and deposi ti on just above the water line (Halls 11, Spring umes of sand, does not re sult in per manent change to the profil e. 1996). This result s in an onshore shift of sand on the beach pro- Signifi cant erosion events occurred in the fall of 1992 and 1995, file. In some cases, the re ar rangement and ac creti on go on si- at low lake levels . One of the largest changes docum ented in the mul ta ne ously (Songo 7, Spring 1992). Erosion can also occur six years of beach profil ing is an erosion event at Songo 3 be- during spring high wa ter, in com bina ti on with high winds tween Septem ber 1 and Oc tober 3, 1995, a time of low water . (Songo 5 and 7, Spring 1996). The sand lost was re stored within a year. Low lake lev els, there - After reaching its maxim um in spring, the lake level begins fore, do not pose the same risk of long term change to beaches as to drop, reaching an annual mini mum around the first of Novem - high lake levels . ber. The falling lake lev els cause ac cre tion of sand to the dry Over all, the beach pro file data pres ent an emerg ing pic ture beach as the swash zone moves down the beach, piling up sand of beach dynam ics in which long term sta bil ity is punctu ate d by just above the water level as it goes (Rock wall 8, Sum mer 1994). sudden large, long-las ting, and irre versi ble changes caused by This is often accom panied by erosion of the under wa ter porti on storms during high water . Only one such catastrophi c event was of the profil e (Songo 3, Sum mer 1994). The net ef fect is an off- observe d, in the fall of 1996, during the six years of profil ing this shore shift in the profil e im me di ate ly around the wa ter level. study covered. Predic ti ons are made of its long-las ting ef fects

16 based on the loss of large trees along the shore line, and expo sure doned, es pe cially where there are mul tiple sites at a beach. of the roots of many other trees. We rec om mend keep ing at least two sites per beach, so that The shore line clas si fi ca tion study showed ap proxi mately long shore current s and intra-beac h variati on can be evalu - 25 percent of the shoreli ne is com posed of sandy sedi ment, ated. A third site would be very benefi cia l in the event of the evenly distri buted around the lake. These are the most easil y loss of one of the two oth ers - as has oc curred else where by eroded shoreli nes on the lake. construc tion or vandal ism . Pres ently the total number of This study include d sites facing all com pass di recti ons. sites is un man age able for eco nomi cal data col lec tion, analy - The south-faci ng beaches expe ri ence d the most change over the sis, and report ing. Perhaps some sites “abandoned” can be study peri od, but a longer peri od of data col lec tion at the other profil ed in August or Septem ber (once a year) for long-ter m beaches could al ter this. The poten ti al for a “northeas ter” during trend analysis . For this reason the pins at the begin ning spring high lake levels is high. Such a storm could cause large point of each profil e site should not be rem oved. and long- lasting changes at north to east-faci ng beaches. · Beaches and sites that we con sid er impor tant to maintain The Em ery method of beach profiling is eas ily learned and are: inex pen sive in terms of capital costs. It is useful for provid ing a Songo 3, 4, 5, and 7: long record of data, south-faci ng, picture of beach changes over time, and is suitable for use by a heavy hu man in flu ence, rec rea tional value, re cent diverse group of volun tee rs. It has some short com ings, how- storm- caused ero sion. ever, as follows. Halls Beach 11 and 12: long data rec ord, southwest- · Any meas ure ment or re cord ing er rors are cu mu la tive, af - facing, rec rea tional value, re cent beach nour ish ment fect ing all of the pro file past the point where the error oc- fol low ing storm dam age. curred. If the er ror is large enough, it is easily recog nized Rockwal l 7 and 8: 3-yea r data rec ord, northwest- and of ten can be cor rected. If the er ror is of mod er ate size, it facing, no hu man in flu ence, longest fetch on the lake. may not be detected and may be er ro ne ously in ter preted as Standish 1 and 3: 3-yea r data rec ord, north-faci ng, an ero sion or ac cre tion event. One such er ror can cre ate a near Portland Wa ter Dis trict in takes, pub li c boat false con clu sion. launch. · A com mon source of error in this method is in the record ing Harmon Beach; Barton, Banks, Straw #2 Resi- of plus or minus val ues in each read ing. Use of a rod and dences: 3- year data set, east- facing, de vel oped shore- pop- level, or rela tively low- cost con struc tion la ser level line, pos si ble wave re frac tion ef fects. would eliminate these sign errors. Frye Island 1 and 3: 3- year data set, south- facing, re - cent ma rina de vel op ment be hind beach, re cent storm- REC OM MEN DA TIONS caused ero sion. Long Point: north-faci ng, well exposed with long A number of rec om menda ti ons follow from the result s of fetches to northwest , north, and north east winds. this study and from our expe ri ence in data col lec tion and analy - Long Beach: northeast-faci ng, well exposed with long sis. fetches to northwest winds, heavy hum an influ ence . · Conti nue peri odic surveys of eroding bluffs. A frequency of · First, we recom m end con tinu ing data collec tion at se lected once per year is adequate, with budget flexibil ity to in crease profil e sites around the lake. The conclu sions drawn in this the frequency follow ing any storm or other event that report are preli minary . At some sites the conclu sions are warrants closer monitor ing. based on only three years of data, someti mes with a single · Com pare two beach profil ing methods to see if Em ery profil e in one or more years. At the Songo and Halls beach method re sults cor re late well with to tal sta tion sur vey sites, where data col lec tion has been in progress for six method result s. If the two methods provide the same level years, only in the fall of the sixth year was a major of ac cu racy, then data sets can be com bined.. shoreline- changing event docum ented. Profil ing should be · Data from all sites should be col lected and analyzed when - conti nued to verify some of the conclu sions of this study, ever pos sible by the same in di vidu als. Fa mili ar ity with the and to pro vide a more complete data set for analy sis. sites is crucial to pro file in ter pre ta tion and to the de vel op- Beaches rep re sent ing all com pass direc ti ons should be in- ment of an under sta nding of beach dy nam ics at a site. cluded. Future profil ing should include many of the exist - · All beach profil e data should be maintai ned in a data base ing sites becaus e of the value of the long- term data sets. that is eas ily ac ces si ble to any po ten tial us ers. · Im me di ate ly ac ti vate a cou ple of the aban doned beach pro - · Since high lake levels expose the shore line to the poten ti al fil ing sites at Long Beach in Stan dish to collect base line for major storm-generat ed ero sional events, maxim um lake data. Current ly, there is no data from these sites. Long levels should be restri cted in eleva ti on and dura ti on at any Beach ap pears to be at risk of large- scale shore line changes time of the year when storms can be ex pected. in the event of a “northeas ter,” a com mon type of storm in · The eleva ti ons of known clay outcrop pings around the pe- the south ern Maine area. Sev eral other sites can be aban - rimeter of the lake should be surveye d so that the im pact of

17 R. A. Johnston and M. N. Mixon

dif fer ent wa ter lev els can be evalu ated. Clay exposed to Dickson, S. M., and Johnston, R. A., Se bago Lake State Park beach dynam ics; a wave erosion can add unwante d nutri ents to the lake and report on results of beach pro filing: Maine Geologi cal Survey, Open- File Report 94-4, 189 p. have an im pact on overall water qualit y. Ele va tion data on Dionne, J. C., 1993, Sedim ent load of shore ice and ice rafting poten tial , upper St. po ten tial clay ero sional sites will as sist in over all shore line Lawrence es tu ary, Quebec, Can ada: Journal of Coastal Re search, v. 9, p. pro tec tion and planning. 628- 646. Em ery, K. O., 1961, A sim ple method of measur ing beach profiles: Lim nology and Oceanog raphy, v. 6, p. 90-93. AC KNOWL EDG MENTS Fed eral Emergency Man age ment Agency (FEMA), 1996, In teragency haz ard mitiga tion report: Federal Emergency Manage m ent Agency, FEMA- We grate fully ac knowledge the work of many beach profil - DR- 1143- ME, 12 p. ers who got their feet wet on this project . Profil ers from the Port- Fed eral En ergy Regu la tory Com mis sion (FERC), 1997, Eel Weir hy droe lec tric pro ject, fi nal en vi ron men tal im pact state ment: Fed eral En ergy Regu la - land Water Dis trict, the Friends of Se bago Lake, the Maine tory Com mission, FERC No. 2984-025, Maine. Bu reau of Parks and Rec rea tion (now called the Maine Bu reau of Folk, R. L., 1974, Petrol ogy of sedi mentar y rocks: Hem phill Publishing Parks and Lands) in the Depart ment of Conser va ti on, and the Company, Austin, Texas, 182 p. Maine Geologi cal Survey (also in the Depart ment of Conser va - Hussey, A. M., II, 1996, Bedrock ge ology of the North Windham 7.5’ quadran - gle, Maine: Maine Geologi cal Survey, Open- File Report 96- 16, 6 p., tion). We would like to thank the student inter ns involved in the map, scale 1:24,000. proj ect: Char lie Page and Stephen Fourt ney from Bates Col lege, Johnston, R. A., 1998, Shoreline classi fi ca tion of Se bago Lake, Maine: Maine Car ter Smith from Bow doin College, and Beth Lewis from the Geologi cal Survey, Open- File Map 98- 123, scale 1:24,000. Univer sity of Maine in Orono. Thanks go to Dr. Mi chael Re telle Johnston, R. A., Kelley , J. T., and Belknap, D. F., 1994, Geom orphol ogy, surfic - ial sedi ments and Qua ternary stratigra phy of Se bago Lake, Maine: Geo - at Bates College, Dr. Pe ter Lea at Bow doin College , and Dr. logi cal So ci ety of Amer ica, Ab stracts with Pro grams, North east ern Dan iel Belk nap at the Uni ver sity of Maine for their help with the Section, v. 26, no. 3, p. 26. student inter ns. We would like to thank Tom Skolfield and his Ko mar, P. D., 1976, Beach proc esses and sedi men ta tion: Prentice-Hall, Inc., staff at Se bago Lake State Park and Vic Rich ards of Frye Island Engle wood Cliffs, New Jersey , 429 p. Lewis, E. B., and Johnston, R. A., 1998, Slope stabil ity / shore line classi fi ca tion for their as sis tance. map of the Songo River, Maine: Maine Geologi cal Survey, Open- File We thank Stephen Dickson, marine geolo gist at the Maine Map 98- 124, scale 1:4,000. Geologi cal Survey , for his advice and as sis tance in the of fice and Lisius, J. D., Monasky, G. A., and Mumme, K. I., 1990, Transport- purification the field. Finall y, we would like to thank Dana Perkins, Phil model of the Lower Bay of Sebago Lake for predic tion of wa ter quality : Re port to the Portland Wa ter District. Bois son neault, Jody Smith, Ron Faucher, and Nancy Manning Lo rang, M. S., Ko mar, P. D., and Stand ford, J. A., 1993, Lake level regu la tion of the Portland Water Distri ct for their help on the project . and shore line erosion of Flathead Lake, Montana: A re sponse to the re - distri bu tion of annual wave energy : Journal of Coastal Resear ch, v. 9, p. REF ER ENCES 494- 508. Page, C., III, 1996, Beach dy nam ics of Songo Beach, Sebago Lake, Maine: ero- sion and sedim ent transport: Senior Thesis, Bates College, Lewiston, Barnes, F. A., and King, C. A. M., 1955, Beach changes in Lincoln shire since the Maine. 1953 storm surge: East Midland Geography, v. 4, p. 18-28. Smith, C. L., 1997, Bluff reces sion rates and processes at Se bago Lake, Maine: Bar nes, P. W., McCor mick, M., and Guy, D. E., Jr., 1993, Quan tity and fate of Sen ior the sis, Bow doin Col lege, Bruns wick, Maine. sedi ment in Lake Erie coastal ice in Ohio: Geological So ci ety of Amer - Thomp son, W. B., 1976, Surfic ial ge ology of the Gray quadran gle, Maine: ica, Abstra cts with Programs, v. 25, no. 6, p. A- 66-67. Maine Geologi cal Survey, Open- File Map 76- 45, scale 1:62,500. Bloom, A. L., 1959, The Geol ogy of Se bago Lake State Park: Maine Geologi cal Thomp son, W. B., and Borns, H. W., Jr. (editors) , 1985, Surfic ial geo logic map Survey, Bulle tin 11, 24 p. of Maine: Maine Geologi cal Survey, map, scale 1:500,000. Bloom, A. L., 1960, Late Pleisto cene changes in sea level in southwest ern Thomp son, W. B., and Smith, G. W., 1977, Surfic ial ge ology of the Sebago Lake Maine: Maine Geologi cal Survey, 143 p. quadran gle: Maine Geologi cal Survey, Open- File Map 77- 45, scale Bolduc, A. M., Thompson, W. B., and Meglioli, A., 1994, Surfic ial ge ology of 1:62,500. the North Windham quadran gle, Maine: Maine Geologi cal Survey U. S. Army Corps of Engi neer s, 1984, Shore Protec tion Manual, 1984: U. S. Open- File Report 94-2, 7 p., map, scale 1:24,000. Army Corps of Engi neer s, Water ways Experi ment Station, Coastal Engi - Denny, C. S., 1982, Geom orphol ogy of New England: U. S. Geologi cal Survey, neer ing Resear ch Center , Vicksburg, MS, 4th edition. Profes sional Paper 1208, 18 p. Wheeler, R., 1994, The Great Ba sin Dam War: Friends of Se bago Lake, Sebago Lake, Maine, 13 p.

18

Figure 2. Tree roots continue to be exposed at Songo Beach, December 1996 (photo by Robert Johnston).

20

22

Upland

Beachface Wave Length Wave Height

Runnel Ridge

Dry or Subaerial Beach Submerged Beach

Swash Zone

Figure 4. Beach terminology (modified from Dickson and Johnston, 1994). Beach Profile Adjustment to Lake Level Changes

Accretion Rising Lake Level

Landward Profile Erosion Shift

Accretion

a

Erosion Falling Lake Level

Lakeward Profile Shift Accretion

Erosion

b

Figure 5. Equilibrium configurations showing a schematic example of profile change from one "equilibrium" configuration (solid line) to another (dashed line). With a lake level rise (a) the profile adjusts upward and landward, .while with a lake level fall (b) the profile adjusts downward and lakeward (from Dickson and Johnston, 1994).

23

Figure 21. Photograph from spring 1996 of ice carrying sand and ice bulldozing sand on Long Beach (measuring stick is one half meter in length). Photo by Robert Johnston.

39 Figure 22. Photograph of spring 1997 ice ridge, Songo Beach (photo by Robert Johnston).

40

Beach Dynamics of Sebago Lake

Ap pen dix I

Beach Profiles

This appen dix contai ns plots of the profil es re ferred to in the text, where they are arranged by beach and site, in clockwise order around the lake. Within the profil e set for each site, profil es are arranged in chronologi cal order by sea son, with any addi ti onal profil es (com pari son of all fall profil es, for exam ple) at the end. Figure 3 shows loca ti ons of the profil e stati ons.

Prof ile Lines

Location Site Page Frye Island, ...... Site 2 ...... 45 Site 3 ...... 53 Halls Beach, ...... Site 11...... 59 Site 12 ...... 79 Harmon Beach, ...... Banks Residence ...... 101 Bar ton Resi dence ...... 114 Straw #2 Residence ...... 123 Long Point Beach, ...... Sun ning dale Residence ...... 132 Rock wall Beach, ...... Site 7 ...... 143 Site 8 ...... 155 Sand bar Beach, ...... Site 10 ...... 167 Site 12 ...... 178 Songo Beach, ...... Site 3 ...... 189 Site 4 ...... 205 Site 5 ...... 219 Site 7 ...... 235 Standish Boat Launch, ...... Site 1 ...... 250 Site 3 ...... 262

43