KENNEBEC COUNTY, MAINE (ALL JURISDICTIONS) Volume 1 of 2

Transcription

1 KENNEBEC COUNTY, MAINE (ALL JURISDICTIONS) Volume 1 of 2 COMMUNITY NAME COMMUNITY NUMBER ALBION, TOWN OF AUGUSTA, CITY OF BELGRADE, TOWN OF BENTON, TOWN OF CHELSEA, TOWN OF CHINA, TOWN OF CLINTON, TOWN OF FARMINGDALE, TOWN OF FAYETTE, TOWN OF GARDINER, CITY OF HALLOWELL, CITY OF LITCHFIELD, TOWN OF MANCHESTER, TOWN OF MONMOUTH, TOWN OF MT VERNON, TOWN OF OAKLAND, TOWN OF PITTSTON, TOWN OF RANDOLPH, TOWN OF READFIELD, TOWN OF ROME, TOWN OF SIDNEY, TOWN OF UNITY, TOWNSHIP OF VASSALBORO, TOWN OF VIENNA, TOWN OF WATERVILLE, CITY OF WAYNE, TOWN OF WEST GARDINER, TOWN OF WINDSOR, TOWN OF WINSLOW, TOWN OF WINTHROP, TOWN OF Kennebec County Effective Date: June 16, 2011 Federal Emergency Management Agency FLOOD INSURANCE STUDY NUMBER 23011CV001A

2 NOTICE TO FLOOD INSURANCE STUDY USERS Communities participating in the National Flood Insurance Program have established repositories of flood hazard data for floodplain management and flood insurance purposes. This Flood Insurance Study (FIS) may not contain all data available within the repository. It is advisable to contact the community repository for any additional data. Selected Flood Insurance Rate Map panels for the community contain information that was previously shown separately on the corresponding Flood Boundary and Floodway Map panels (e.g., floodways, cross sections). In addition, former flood hazard zone designations have been changed as follows: Old Zone A1 through A30 B C New Zone AE X X Part or all of this Flood Insurance Study may be revised and republished at any time. In addition, part of this Flood Insurance Study may be revised by the Letter of Map Revision process, which does not involve republication or redistribution of the Flood Insurance Study. It is, therefore, the responsibility of the user to consult with community officials and to check the community repository to obtain the most current Flood Insurance Study components. Initial Countywide FIS Effective Date: June 16, 2011

3 TABLE OF CONTENTS Table of Contents Volume 1 Page 1.0 INTRODUCTION Purpose of Study Authority and Acknowledgments Coordination AREA STUDIED Scope of Study Community Description Principal Flood Problems Flood Protection Measures ENGINEERING METHODS Hydrologic Analyses Hydraulic Analyses Vertical Datum FLOODPLAIN MANAGEMENT APPLICATIONS Floodplain Boundaries Floodways INSURANCE APPLICATIONS FLOOD INSURANCE RATE MAP OTHER STUDIES LOCATION OF DATA BIBLIOGRAPHY AND REFERENCES 87 i

4 Table of Contents Volume 1 - continued Page FIGURES Figure 1 Floodway Schematic 60 TABLES Table 1 - CCO Meeting Dates for Precountywide FISs 8 Table 2 - Flooding Sources Studied by Detailed Methods 9 Table 3 - Flooding Sources Studied by Approximate Methods 12 Table 4 Letters of Map Change 16 Table 5 - Population and Total Area by Community 17 Table 6 - Summary of Discharges 33 Table 7 - Summary of Stillwater Elevations 36 Table 8-1-Percent-Annual-Chance Flood Data 50 Table 9 - Manning s n Values 54 Table 10 - Floodway Data 61 Table 11 - Community Map History 85 Table of Contents Volume 2 EXHIBITS Exhibit 1 - Flood Profiles Bond Brook Cobbosseecontee Stream Eastern River Kennebec River Meadow Brook Messalonskee Stream Sebasticook River Togus Stream West Branch Sheepscot River Panels 01P-03P Panels 04P-18P Panels 19P-21P Panels 22P-39P Panel 40P Panels 41P-43P Panels 44P-51P Panels 52P-57P Panels 58P-59P ii

5 Table of Contents Volume 2 continued EXHIBITS - continued Exhibit 2 - Flood Insurance Rate Map Index Flood Insurance Rate Map iii

6 1.0 INTRODUCTION FLOOD INSURANCE STUDY KENNEBEC COUNTY, MAINE [ALL JURISDICTIONS] 1.1 Purpose of Study This Flood Insurance Study (FIS) revises and updates information on the existence and severity of flood hazards in the geographic area of Kennebec County, including the Cities of Augusta, Gardiner, Hallowell, and Waterville; the Towns of Albion, Belgrade, Benton, Chelsea, China, Clinton, Farmingdale, Fayette, Litchfield, Manchester, Monmouth, Mt. Vernon, Oakland, Pittston, Randolph, Readfield, Rome, Sidney, Vassalboro, Vienna, Wayne, West Gardiner, Windsor, Winslow, and Winthrop, and the Township of Unity (referred to collectively herein as Kennebec County), and aids in the administration of the National Flood Insurance Act of 1968 and the Flood Disaster Protection Act of This study has developed flood-risk data for various areas of the community that will be used to establish actuarial flood insurance rates and to assist the community in its efforts to promote sound floodplain management. Minimum floodplain management requirements for participation in the National Flood Insurance Program (NFIP) are set forth in the Code of Federal Regulations at 44 CFR, Also note that there were no previously printed FIS texts for the Towns of Albion, Fayette, Mt. Vernon, Unity, Vassalboro, Vienna, and Windsor. In some States or communities, floodplain management criteria or regulations may exist that are more restrictive or comprehensive than the minimum Federal requirements. In such cases, the more restrictive criteria take precedence, and the State (or other jurisdictional agency) will be able to explain them. 1.2 Authority and Acknowledgments The sources of authority for this FIS report are the National Flood Insurance Act of 1968 and the Flood Disaster Protection Act of This FIS was prepared to incorporate all the communities within Kennebec County in a countywide format. Information on the authority and acknowledgements for each jurisdiction included in this countywide FIS, as compiled from their previously printed FIS reports, is shown below: Augusta, City of: In the October 1, 1980, original FIS, the hydrologic and hydraulic analyses for the Kennebec River and Bond Brook were prepared by the U.S. Geological Survey 1

7 Augusta, City of cont d: (USGS) for the Federal Insurance Administration (FIA), under Inter-Agency Agreement No. IAA-H-9-77, Project Order No. 6. This work was completed in May For the June 15, 1994, revised FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS for Federal Emergency Management Agency (FEMA), under Inter-Agency Agreement No. EMW-90-E-3266, Project Order No. 2. This work was completed in August Belgrade, Town of: Benton, Town of: Chelsea, Town of: The hydrologic and hydraulic analyses for the January 16, 1987, FIS study were prepared by USGS for FEMA, under Inter- Agency Agreement No. EMW-E-1823, Project Order No. 11. This work was completed in August In the May 4, 1988, original FIS, the hydrologic and hydraulic analyses were prepared by the Soil Conservation Service (SCS) for FEMA, under Inter-Agency Agreement No. EMW-84-E-1550, Project Order No. 1. This work was completed in September For the May 7, 2001, revised FIS, the hydrologic and hydraulic analyses for the Sebasticook River were prepared by the USGS for FEMA, under Inter-Agency Agreement No. EMW-97-IA , Project Order No. 1. This work was completed in February In the December 1979, original FIS, the hydrologic and hydraulic analyses for the Kennebec River and Togus Stream were prepared by the USGS for the FIA, under Inter-Agency Agreement No. IAA-H-9-77, Project Order No. 6. This work was completed in January For the June 15, 1994, revised FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS for the FEMA, under Inter-Agency Agreement No. EMW- 2

8 Chelsea, Town of cont d: China, Town of: Clinton, Town of: Farmingdale, Town of: 90-E-3266, Project Order No. 2. This work was completed in August The hydrologic and hydraulic analyses for the June 5, 1989 FIS were prepared by USGS for FEMA, under Inter-Agency Agreement No. EMW-85-E-1823, Project Order No. 20. This work was completed in January The hydrologic and hydraulic analyses for the May 3, 1990 FIS were prepared USGS for FEMA, under Inter-Agency Agreement No. EMW-85-E-1823, Project Order No. 20. This work was completed in March The hydrologic and hydraulic analyses for the Kennebec and Sebasticook Rivers were previously performed by the SCS. In the March 1980, original FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS, Water Resources Division, for FEMA, under Inter-Agency Agreement No. IAA-H-9-77, Project Order No. 6. This work was completed in January For the May 2, 1994, revised FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS for FEMA, under Inter-Agency Agreement No. EMW-90-E-266, Project Order No. 2. This work was completed in August Gardiner, City of: Kennebec River and Cobbosseecontee Stream were prepared by the USGS, Water Resources Division, for the FIA, under Inter-Agency Agreement No.IAA-H-9-77, Project Order No. 6. This work was completed in January For the July 18, 1994, revised FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS for FEMA, under Inter-Agency Agreement No. EMW- 90-E-3266, Project Order No. 2. This work was completed in August

9 Hallowell, City of: Litchfield, Town of: Manchester, Town of: Monmouth, Town of: Oakland, Town of: Pittston, Town of: In the May 1979, original FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS for the FIA, under Inter-Agency Agreement No. IAA-H This work was completed in December For the July 18, 1994, revised FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS for FEMA, under Inter-Agency Agreement No. EMW-90-E- 3266, Project Order No. 2. This work was completed in August The hydrologic and hydraulic analyses for the November 19, 1986 FIS were prepared by the USGS FOR FEMA, under Inter- Agency Agreement No. EMW-85-E This work was completed in May The hydrologic and hydraulic analyses for the February 1980 FIS were performed by the USGS for the FIA under Inter-Agency Agreement No. IAA-H-9-77, Project Order No. 7. This work was completed in May The hydrologic and hydraulic analyses for the March 1980 FIS were performed by the USGS for the FIA, under Inter-Agency Agreement No. IAA-H-9-77, Project Order No. 7. This work was completed in June The hydrologic and hydraulic analyses for the June 15, 1988 FIS were prepared by the USGS during the preparation of the FIS for the Town of Belgrade, for FEMA, under Inter-Agency Agreement No.EMW-E-1823, Project Order No. 11. This work was completed in August For the September 16, 1980, original FIS, and March 16, 1981, FIRM, the hydrologic and hydraulic analyses for the Kennebec River and Togus Stream were prepared by the USGS for FEMA, under Inter-Agency 4

10 Pittston, Town of cont d: Agreement No. IAA-H-9-77, Project Order No. 6. This work was completed in January For the May 3, 1993, revised FIS, the hydrologic and hydraulic analyses for the Eastern River were prepared by the USGS for FEMA, under Inter-Agency Agreement No. EMW-87-E This work was completed in December For the April 6, 1998 revised FIS, the revised hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS for FEMA, under Inter-Agency Agreement No. EMW-92-E-3848, Project Order No. 4. This work was completed in January Randolph, Town of: Readfield, Town of: Rome, Town of: In the March 1979, original FIS, the hydrologic and hydraulic analyses for the Kennebec River and Togus Stream were prepared by the USGS, Water Resources Division for the FIA, under Inter-Agency Agreement No. IAA-H-9-77, Project Order No. 6. This work was completed in January In the July 5, 1994, revised FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS for FEMA, under Inter-Agency Agreement No. EMW-90-E-3266, Project Order No. 2. This work was completed in August The hydrologic and hydraulic analyses for the June 1980 FIS were performed by the USGS for the FIA, under Inter-Agency Agreement No. 1AA-H-9-77, Project Order No. 7. This work was completed in June The hydrologic and hydraulic analyses for the May 17, 1988 FIS were obtained from the FIS for the Town of Belgrade. The Town of Belgrade study was prepared by the USGS for FEMA, under Inter-Agency Agreement No. EMW-E-1823, Project Order No. 11. This work was completed in August

11 Sidney, Town of: Waterville, City of: Wayne, Town of: West Gardiner, Town of: Winslow, Town of: For the November 20, 1998, revised FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS, Augusta District, for FEMA, under Inter-Agency Agreement No. EMW-95-E This work was completed in September Flood hazard information for Messalonskee Lake was based on information obtained from the FIS for the Town of Oakland (Reference 1). For the February 17, 1988, original FIS, the hydrologic and hydraulic analyses were prepared by the SCS for FEMA, under Inter- Agency Agreement No. EMW-84-E-1550, Project Order No. 1. This work was completed in August For the Mary 7, 2001, revised FIS, the hydrologic and hydraulic analyses for the City of Waterville were prepared by the USGS, Maine District Office, for FEMA, under Inter-Agency Agreement No. EMW-97-IA , Project Order No.1. This work was completed in March The hydrologic and hydraulic analyses for the April 3, 1989 FIS were prepared by USGS for FEMA, under Inter-Agency Agreement No. EFW-85-E-1823, Project Order No. 11. This work was completed in October The hydrologic and hydraulic analyses for the September 1979 FIS were performed by USGS for the FIA, under Inter-Agency Agreement No. IAA-H-9-77, Project Order No. 7. This work was completed in May For the September 30, 1987, original FIS, the hydrologic and hydraulic analyses were prepared by the SCS for FEMA, under Inter- Agency Agreement No. EMW-84-E-1 550, Project Order No. 1. This work was completed in July

12 Winslow, Town of cont d: Winthrop, Town of: For the May 7, 2001, revised FIS, the hydrologic and hydraulic analyses for the Kennebec River were prepared by the USGS, Maine District Office, for FEMA, under Inter-Agency Agreement No. EMW- 97-IA-0115, Project Order No.1. This work was completed in August The hydrologic and hydraulic analyses for the February 1980 FIS were performed by the USGS for the FIA, under Inter-Agency Agreement No. IAA-H-9-77, Project Order No. 7. This work was completed in May Base map information shown on this revision was obtained from the Maine Geographic Information System (MeGIS). Base map files were provided in digital format by the Office of Maine GIS ( Orthophoto images, except for the panels listed below, were produced at a scale of 1:2,400 and 1:4,800. The photography was acquired beginning in the spring of 2003 through the spring of The pixel resolution of the ortho imagery used for Kennebec County is both 1- and 2-ft, depending on the panel. Due to the incomplete coverage in orthophoto imagery by MeGIS 2005, an alternate source for base map imagery was used on several FIRM panels. Panels 0068, 0069, 0195, 0335, 0380, and 0544 were created using 1-meter pixel resolution USGS DOQQs acquired between 1991 and The projection used in the preparation for both sources of orthophoto imagery was Universal Transverse Mercator (UTM) Zone 19. The horizontal datum is North American Datum (NAD) of 1983, GRS80 spheroid. 1.3 Coordination The purpose of an initial Consultation Coordination Officer s (CCO) meeting is to discuss the scope of the FIS. A final meeting is held to review the results of the study. The dates of the initial, intermediate and final CCO meetings held for the incorporated communities within Kennebec County are shown in Table 1, CCO Meeting Dates for Precountywide FIS. 7

13 TABLE 1 - CCO MEETING DATES FOR PRECOUNTYWIDE FIS Community Name Initial CCO Date Intermediate CCO Date Final CCO Date Augusta, City of December 18, 1975 May 12, 1978 March 29, 1979 Belgrade, Town of January, 1985 August 13, 1985 January 17, 1986 Benton, Town of June 9, 1997 * March 14, 2000 Chelsea, Town of December 09, 1992 * July 16, 1993 China, Town of January, 1985 February, 1988 May 26, 1988 Clinton, Town of February 1985 * September 28, 1988 Farmingdale, Town of September 22, 1989 * May 14, 1993 Gardiner, City of September 22, 1989 * July 28, 1993 Hallowell, City of December 09, 1992 * July 29, 1993 Litchfield, Town of January, 1985 June 24, 1985 December 10, 1985 Manchester, Town of January 20th, 1976 July 18, 1978 December 13, 1978 Monmouth, Town of February 1976 June 20, 1978 May 10, 1979 Oakland, Town of June 20, 1986 * November 5, 1986 Pittston, Town of April 29, 1992 * * Randolph, Town of September 21, 1989 * May 14, 1993 Readfield, Town of January 1976 June 19, 1978 May 15, 1979 Rome, Town of * * November 18, 1986 Sidney, Town of August 14, 1995 * December 16, 1997 Waterville, City of May 1984 * March 6, 1987 Wayne, Town of January 1985 October 1987 April 4, 1988 West Gardiner, Town of February 1976 May 12, 1978 December 13, 1978 Windsor, Town of * * * Winslow, Town of September 29, 1999 * * Winthrop, Town of September 1976 June 20, 1978 May 09, 1979 *Data not available For this Countywide FIS, the initial Consultation Coordination Officer (CCO) meeting was held on November 7, 2005, and was attended by representatives of FEMA, the Maine Floodplain Management Program, and USGS Maine Water Science Center, and community officials. The results of the study were reviewed at the final CCO meeting held on February 11, 2009 and February 12, 2009, and was attended by representatives from the Towns of Gardiner, Fayette, Readfield, Monmouth, Manchester, Augusta, Hollowell, Vassalboro, Winslow, Pittston, Belgrade, Waterville, and West Gardiner. Representatives also attended from FEMA Region 1, the Maine State Planning Office (SPO), and Camp Dresser & McKee, Inc (CDM). All issues raised at that meeting have been addressed in this study. 8

14 2.0 AREA STUDIED 2.1 Scope of Study This FIS report covers the geographic area of Kennebec County, Maine, including the incorporated communities listed in Section 1.1. The areas studied by detailed methods were selected with priority given to all known flood hazards and areas of project development or proposed construction. All or portions of the flooding sources listed in Table 2, Flooding Sources Studied by Detailed Methods, were studied by detailed methods in the precountywide FISs. Limits of detailed study are indicated on the Flood Profiles (Exhibit 1) and on the FIRM. The areas studied by detailed methods were selected with priority given to all known flood hazards and areas of projected development or proposed construction. TABLE 2 FLOODING SOURCES STUDIED BY DETAILED METHODS Flooding Source Name Androscoggin Lake Annabessacook Lake Belgrade Stream Berry Pond Bond Brook Branch Pond China Lake Cobbosseecontee Lake Description of Study Reaches For its entire shoreline. For its entire shoreline. Including downstream portions of its tributaries in the Town of Belgrade (Hoyt Brook, Sanford Brook and Meadow Brook). In this study, the section of Belgrade Stream above Wings Mill Dam in the Town of Mount Vernon is considered to be a part of Long Pond, and the section below Wings Mill Dam is considered to be a part of Messalonskee Lake. For the entire shoreline. For its entire length within the City of Augusta. For the entire shoreline within the Town of China. For the entire shoreline within the Town of China. For its entire shoreline. 9

15 TABLE 2 FLOODING SOURCES STUDIED BY DETAILED METHODS - continued Flooding Source Name Cobbosseecontee Stream Description of Study Reaches From the outlet of Cobboseecontee Lake in the Town of Manchester through the Town of West Gardner to the West Gardiner-Manchester corporate boundary. From Pleasant Pond in the Town of Litchfield to the upstream corporate limits. Cochnewagon Lake Dexter Pond For the entire shoreline. For the entire shoreline. Eastern River From the downstream Pittston corporate limits to a point approximately 50 feet upstream of State Route 194. Echo Lake Great Pond Kennebec River Little Cobbosseecontee Lake Long Pond Lovejoy Pond Lower Narrows Pond Maranacook Lake Meadow Brook Messalonskee Lake (Snow Pond) For its entire shoreline within the Town of Readfield. For its entire shoreline. For its entire length within Kennebec County. For its entire shoreline. For its entire shoreline within the Towns of Belgrade and Rome. For its entire shoreline within the Towns of Readfield and Wayne. For its entire shoreline. For its entire shoreline. From a point approximately 4,180 feet downstream of Dirigo Road to approximately 50 feet upstream of Dirigo Road in the Town of China. For its entire shoreline. 10

16 TABLE 2 FLOODING SOURCES STUDIED BY DETAILED METHODS - continued Flooding Source Name Messalonskee Stream Pleasant Pond Pocasset Lake Salmon Lake ( McGrath and Ellis Ponds) Sand Pond Description of Study Reaches For its entire length within the City of Waterville. Flooding along the shoreline in the Town of Litchfield. For its entire shoreline. For its entire shoreline. For its entire shoreline within the Town of Monmouth. Sebasticook River Tacoma Ponds Threecornered Pond Threemile Pond Togus Pond For its entire length within the Town of Benton, Clinton and Winslow. Flooding of the shoreline; this includes Woodbury Pond, Sand Pond, Buker Pond, Jimmy Pond, and Little Purgatory Pond in the Town of Litchfield For its entire shoreline. For its entire shoreline within the Town of China. For its entire shoreline. Togus Stream For its entire length. Torsey Lake For its entire shoreline within the Town of Readfield. Tributary A to Threecornered Pond Upper Narrows Pond For its entire shoreline within the City of Augusta For its entire shoreline. Upper Pleasant Pond Flooding along the shoreline. 11

17 TABLE 2 FLOODING SOURCES STUDIED BY DETAILED METHODS - continued Flooding Source Name West Branch Sheepscot River Wilson Pond Woodbury pond Description of Study Reaches From a point approximately 3,200 feet upstream of the Town of China corporate limits to a point approximately 540 feet upstream of Weeks Mills Road Bridge in the Town of China For its entire shoreline. For its entire shoreline within the Town of Monmouth. For this revision, the Kennebec River was redelineated for its entire length within Kennebec County, using more up to date topographic information. The topographic data, in the form of ft contour intervals, GIS layer (shapefile format) was obtained from Maine GIS ( for the redelineation. This data was published by Maine GIS in April 2004 and was created by automation of thematic layer from USGS 7.5 minute 1:24,000 scale quadrangle map. The features were checked for registration, linework quality and the completeness and have been verified following the standard of 90% accuracy by Maine GIS. No new detailed or approximate studies were performed. Approximate analyses were used to study those areas having a low development potential or minimal flood hazards. The scope and methods of study were proposed to, and agreed upon, by FEMA and the individual communities within Kennebec County. All or portions of the flooding sources listed in Table 3, Flooding Sources Studied by Approximate Methods, were studied by approximate methods in the precountywide FISs. TABLE 3 FLOODING SOURCES STUDIED BY APPROXIMATE METHODS Flooding Source Name Community(s) Anderson Pond Apple Valley Lake Beaver Brook Beaver Pond Bellows Stream Black Ash Swamp Bog Brook Bog Brook, upstream portions of Augusta Winthrop Readfield Rome Winslow Litchfield Monmouth Belgrade 12

18 TABLE 3 FLOODING SOURCES STUDIED BY APPROXIMATE METHODS - continued Flooding Source Name Community(s) Bog Stream Bond Brook Bonny Pond Brainard Pond Buzz Brook Carlton Pond Chaffee Brook Chamberlain Pond Chase Meadow Brook Cold Brook, upstream portions of Cold Stream Cold Stream and Tributaries Dam Pond Dead Stream Dennis Brook Dilnow brook Doctor Pond Dutton Pond East Pond Ellis Brook Evans Pond Farber Brook Fifteenmile Stream Fisher Brook Fowler brook Frost Pond Gardiner Brook Given brook Goff Brook Gould Pond Great Sidney Bog Greeley Pond Greeley Pond outlet Grover Brook and Unnamed Tributary Hales Brook Hamilton Pond Hewitt Brook Hog Brook Holland Brook Hoyt Brook Hoyt Brook, upstream portions of Hunter Brook Hutchinson Pond Readfield Augusta Monmouth Readfield China Readfield, Winthrop Winslow Belgrade Chelsea Belgrade West Gardiner Farmingdale, West Gardiner Augusta Readfield Litchfield Monmouth Sidney China Oakland Sidney China Winslow Benton Augusta Benton Monmouth Readfield, Winthrop Pittston Sidney Sidney Augusta Augusta Augusta West Gardiner Wayne Belgrade China Manchester Waterville Readfield, Winthrop Belgrade China Farmingdale, Manchester 13

19 TABLE 3 FLOODING SOURCES STUDIED BY APPROXIMATE METHODS - continued Flooding Source Name Community(s) Inlet and Outlet Brooks to Hutchinson Pond Inlet Brook Jimmie Pond Jimmie Pond and Inlet Brook Jock Stream Joe Pond Jones Brook Jones Brook Joys Pond and Tributaries Kezar Brook Kezar Pond Kimball Brook and Tributaries Lilly Lake Lily Pond Lily Pond Brook Little Pond Loon Pond Lower Silver Lake Magotty Meadow Brook Meadow Brook Meadow Brook and Tributaries Meadow Brook, upstream portions of Mears Brook Messalonskee Stream Mill Brook Mill Brook and Tributaries Mill Pond Mill Stream Morton Brook and Tributaries Mud Brook Mud Mills Stream Mud Pond Muddy Pond Numerous Tributaries Outlet brook of Carlton Pond Outlet Brook of Lower Narrows Pond Outlet Stream Pattee Pond Pattee Pond Brook Penn Pond Penny Pond Manchester Hallowell Farmingdale Manchester Monmouth Sidney Wayne Winthrop Pittston Winthrop Winthrop Pittston Manchester West Gardiner, Augusta Manchester Rome Litchfield Manchester Litchfield China Farmingdale Belgrade Manchester, Readfield, Winthrop Oakland Sidney Manchester Readfield Readfield, Winthrop Pittston Monmouth Monmouth China, Monmouth, Winslow Wayne Benton Winthrop Winthrop Winslow Winslow Winslow Sidney Belgrade 14

20 TABLE 3 FLOODING SOURCES STUDIED BY APPROXIMATE METHODS - continued Flooding Source Name 15 Community(s) Ponding areas along Cobbosseecontee Stream West Gardiner Potters Brook Litchfield Riggs Brook Augusta Robbins Mill Stream Rome Rolling Dam Brook Gardiner Rome Trout Brook Rome Round Pond Rome Sanford Brook and tributaries Manchester Sanford Brook, upstream portions of Belgrade Shed Pond Manchester, Readfield Sidney Bog Brook Augusta Silver Lake Sidney Spectacle Pond Augusta Spring Brook Augusta Spring Lake Manchester Stickney Brook Augusta Stone Brook Augusta Stoney Brook Rome Stony Meadow Brook Chelsea, Pittston Stuart Pond Belgrade Tannery Brook Monmouth, Winthrop Tanning Brook and tributary below intersection of Worthing and Readfield Roads Manchester Tingley Brook Readfield Togus Brook Randolph Togus Stream Tributary 1 Chelsea Togus Stream Tributary 2 Chelsea Tolman Pond Augusta Town Farm Brook Sidney Trafton Road Brook Waterville Tributary A to Bond Brook Augusta Tributary A to Maranacook Lake Winthrop Tributary B to Bond Brook Augusta Tributary No. 1 to Vaughan Brook Hallowell Tributary No. 2 to Vaughan Brook Hallowell Tributary No. 4 to Vaughan Brook Hallowell Tributary to Berry Pond Winthrop Tributary to Little Cobbosseecontee Lake Winthrop Tributary to Riggs Brook Augusta Tributary to Wilson Stream Winthrop Tyler Pond Manchester Unnamed Brooks (3X) Farmingdale

21 TABLE 3 FLOODING SOURCES STUDIED BY APPROXIMATE METHODS - continued Flooding Source Name Community(s) Unnamed ponding areas Unnamed Ponds Unnamed Streams Unnamed Tributaries (4X) Unnamed Tributaries (7X) Unnamed tributary to the Dead River Upper Silver Lake Vaughan Brook Vaughan Brook and Tributaries Vaughan Pond Ward Pond Watson Road Wellman Pond West Branch Sheepscot River Western Brook and Tributary Whitney Brook Whittier Brook Whittier Pond Wilson Brook Wilson Stream Wayne Oakland, Belgrade Oakland, Belgrade, Wayne Litchfield Sidney Monmouth Manchester Hallowell Farmingdale Hallowell Sidney Rome Augusta, Belgrade China Manchester Augusta Rome Rome Winslow Monmouth Detailed-studied streams that were not re-studied as part of this revision may include a profile baseline on the FIRM. The profile baselines for these streams were based on the best available data at the time of their study and are depicted as they were on the previous FIRMs. In some cases the transferred profile baseline may deviate significantly from the channel or may be outside of the floodplain. This FIS also incorporates the determinations of letters issued by FEMA resulting in map changes (Letter of Map Revision [LOMR]), as shown in Table 4, Letters of Map Change. TABLE 4 LETTERS OF MAP CHANGE Community Case Number Flooding Source Letter Date Readfield, Town of P Torsey Lake-Quiet Harbor 08/06/2004 Peninsula 2.2 Community Description Kennebec County is located in the central southern portion of Maine. In Kennebec County, there are 25 towns, 4 cities, and 1 township. The City of 16

22 Waterville, the Towns of Albion, Benton, China, Clinton, Oakland, Winslow, and the Township of Unity are in the northeastern part of the county. The City of Augusta and the Towns of Belgrade, Manchester, Readfield, Sidney, and Vassalboro are in the central part of the county. The Towns of Monmouth, Wayne and Winthrop are in the western portion of the county. The southern portion of the county is comprised of the Cities of Gardiner and Hallowell, and the Towns of Chelsea, Farmingdale, Litchfield, Pittston, Randolph, and West Gardiner. The Towns of Fayette, Mt. Vernon, Rome, and Vienna are in the northwestern corner of the county, and the Town of Windsor is in the southeast part of the state. Kennebec County is bordered on the north by Somerset County, on the east and southeast by Waldo and Lincoln Counties, on the south by Sagahadoc County, on the west by Androscoggin County, and on the northwest by Franklin County. According to census records, the population of Kennebec County was 117,114 in 2000 (Reference 2). The total area in Kennebec County consists of 951 mi 2, including 84 mi 2 of water area. All communities in Kennebec County, along with their population and total area, are listed in Table 5, Population and Total Area by Community. TABLE 5 POPULATION AND TOTAL AREA BY COMMUNITY Community Population 1 Total Area (sq. mi) 1 Albion, Town of 1, Augusta, City of 18, Belgrade, Town of 2, Benton, Town of 2, Chelsea, Town of 2, China, Town of 4, Clinton, Town of 3, Farmingdale, Town of 2, Fayette, Town of 1, Gardiner, City of 6, Hallowell, City of 2, Litchfield, Town of 3, Manchester, Town of 2, Monmouth, Town of 3, Mt. Vernon, Town of 1, Oakland, Town of 5, Pittston, Town of 2, Randolph, Town of 1, Readfield, Town of 2, Data obtained from U.S Census Bureau (Reference 2) 17

23 TABLE 5 POPULATION AND TOTAL AREA BY COMMUNITY - continued Community Population 1 Total Area (sq. mi) 1 Rome, Town of Sidney, Town of 3, Unity, Township of Vassalboro, Town of 4, Vienna, Town of Waterville, City of 15, Wayne, Town of 1, West Gardiner, Town of 2, Windsor, Town of 2, Winslow, Town of 7, Winthrop, Town of 6, Data obtained from U.S Census Bureau (Reference 2) The bedrock in Kennebec County area is primarily made up of metamorphosed calcareous sand stone, and siltstone, with occasional outcrops of granite bedrock (Reference 3). Surficial geology is classified as till, which is a poorly sorted mixture of clay, silt, sand, gravel, cobbles, and boulders (Reference 4). Surficial deposits in low-lying areas (such as along the Kennebec River and Togus Stream) tend to consist of marine clay, silt, and sand. Topography is generally characterized by low, rolling hills with gentle to moderate slopes. Soils are primarily fine, sandy loams and very stony, fine sandy loams. The Belgrade Lakes in northern Kennebec County are mainly valley lakes, created by glacial deposits that blocked normal drainage and became dams. Many of the lakes have been further dammed by human activities. These dams have deepened and enlarged the lakes; however, their basic pattern was established as a result of the last ice-age glaciation. The last glaciation in Maine ended approximately 12,000 years ago; thus, the lakes are considered "young" in geological terms. Since the old valleys and the movement of the glacier tended to be north to south in orientation, the Belgrade Lakes tend to be elongated and run in a general north to south direction. The quality of the water in the lakes helps make the area good for recreational activities (Reference 5). The Kennebec River basin is located in west-central Maine and drains about onefifth of the state. The Kennebec River originates at the outlet of Moosehead Lake and flows south for approximately 145 miles to Abagadasset Point in Merrymeeting Bay, where the Kennebec River is joined by the Androscoggin River and four smaller rivers before it flows for an additional 20.5 miles to the Atlantic Ocean. 18

24 The headwater lakes and streams that contribute to the Kennebec River flow through a mountainous region with peak elevations of up to 3,000 feet. The central portion of the Kennebec River Valley is characterized as an upland area of rounded hills having local relief of up to 1,000 feet. Below the City of Waterville, the Kennebec River valley widens, and the land elevations are generally less than 100 feet. Between Moosehead Lake and mean tide level at the City of Augusta, the Kennebec River falls 1,026 feet, an average gradient of 8.5 feet per mile. In the Town of Randolph, the Kennebec River is a wide, flowing estuary with a mean tidal range of approximately 5 feet. Land use within the county is primarily open forest and agricultural, with scattered residential development. The City of Augusta, the capital of Maine, is the primary center of employment in Kennebec County. 2.3 Principal Flood Problems In Kennebec County, flooding generally occurs in the winter or early spring as a result of heavy rainfall on snow covered or frozen ground. Ice jams in the rivers can compound flood problems. Flooding occurs in the spring months due to rapid runoff caused by heavy rains combined with snowmelt. Flooding in the summer months is most often associated with thunderstorms, although tropical hurricanes occasionally generate prolonged heavy rainfall in the area. Trees, brush, and other vegetation growing along stream banks impede flood flows during high waters, creating backwater and increasing flood heights. Furthermore, trees, ice, and other debris may be washed away and carried downstream to collect on bridges and other obstructions. As the flood flow increases, significant amounts of this debris often break loose and a wall of water and debris surges downstream until another obstruction is encountered. Debris may collect against a bridge or culvert until the load exceeds the structural capacity, causing its destruction. It is difficult to predict the degree to which, or the location where, debris may accumulate. Therefore, in the development of the flood profiles it has been necessary to assume no accumulation of debris or obstruction of flow. The flood problems for the communities within Kennebec County have been compiled and are described below. Flooding along the Kennebec River usually occurs in the spring from rapid runoff as a result of unusually heavy rains combined with snowmelt. Flooding also occurs as a result of heavy rainfall associated with hurricanes. The most notable floods on the Kennebec River occurred in May 1832, December 1901, March 1936, and April The 1987 flood had a peak discharge of 232,000 cubic feet per second (cfs) at the USGS gaging station in Sidney (station No ) and a recurrence interval of over 100 years (Reference 6). This flood was closely approached by the 1901 event that had a discharge of 157,000 cfs and has an 19

25 estimated recurrence interval of nearly 100 years (Reference 7). The March 1936 event had a peak discharge of 154,000 cfs and has an estimated recurrence interval of 90 years. Discharge data is not available at the gage for the 1832, 1901, and 1936 floods. Other notable floods occurred in the following years: 1854, 1869, 1870, 1896, 1923, 1953, 1969, 1973, 1979, 1983, and All of the floods caused extensive damage within the river basin. Historical peak elevation data indicates that the 1936 flood was higher than the 1987 flood for virtually all of the Cities of Augusta, Gardiner, Hallowell, and the Towns of Benton, Clinton, Chelsea, Farmingdale, Pittston, Randolph, and Sidney (References 6 and 8). The flood of record on the Kennebec River occurred in 1987 in Waterville and Winslow. Several other floods resulting from ice jams have occurred on the Kennebec River and the potential for ice-jam flooding is an annual concern (References 9 and 10). Flood damage from the 1936 flood was exacerbated by the presence of ice jams on the lower part of the Kennebec River, which increased flood elevations upstream as far as the City of Augusta (Reference 8). The Maine Central Railroad tracks were submerged, and ice cakes that were left by the flood obstructed the tracks. U.S. Route 201, which runs parallel to the tracks, has also been subject to flooding. Flooding in the Cities of Augusta and Gardiner occur principally in the commercial area along the banks of the Kennebec River. Gardiner has a downtown business district along Water Street, as well as a shopping center located along the bank of the Kennebec River near the mouth of Cobbosseecontee Stream, which would suffer extensive damage in a major flood. Flooding problems in Benton, Clinton, and Winslow occur mainly along the Kennebec and Sebasticook Rivers. Here, single and multi-family homes, public buildings, businesses, utilities, roads, bridges, and a park experience frequent flood problems. The April 2, 1987 flood on the Sebasticook River had a peak discharge of 17,600 cfs at USGS gaging station near the Town of Pittsfield, and a 1- percent-annual-chance recurrence. Prior to the 1987 flood, the flood of record was recorded in 1936 at the gage near Pittsfield with a discharge of 14,400 cfs (Reference 11). The 1936 flood has a recurrence interval of just less than1- percent-annual-chance. Floods that occur when the river is frozen, such as the 1936 event, cause considerable damage due to floating ice and backwater from ice jams. Extensive damage to property, streets, and structures occurred during the 1936 flood. The 1987 flood on the Sebasticook River breached the dam in Hartland, causing washouts to highways and bridges and extensive water damage to homes in the Hartland and Pittsfield areas. The 1987 flood marks could not be used in the hydraulic analyses of Benton because of significant hydraulic changes caused by the dam at Benton Falls. In the Town of China, in a verbal report from the owner of the dam, the flood of 20

26 record on Branch Pond occurred in the spring of 1940 when the dam owner allowed all gates on the dam to remain fully shut. Also, floods that occur when the West Branch Sheepscot River is frozen may cause considerable damage due to backwater from ice jams and floating ice. Flooding problems in the Towns of Litchfield, Monmouth, Manchester, West Gardiner, and Winthrop occur mainly in the area along the shores of Cobbosseecontee Lake. Litchfield and Monmouth also have flood problems along the shores of the Sand Pond and Woodbury Pond, and Litchfield has flooding problems along the shores of Buker Pond, Jimmy Pond, Pleasant Pond, and Little Purgatory Pond. Additionally, Monmouth has flooding along the shore of Cochnewagon Lake. Flooding problems in Readfield and Winthrop occur principally in the area along the shores of Maranacook Lake, with additional flooding along the shores of Torsey Lake, Echo Lake, and Lovejoy Pond in Readfield, and in the area along the shores of Annabessacook Lake, Wilson, Dexter, and Berry Ponds, Little Cobbosseecontee Lake, Upper Narrows Pond, and Lower Narrows Pond in Winthrop. Two of the most notable events in this region occurred in March 1936 and December During the March 1936 event, heavy rains combined with snowmelt generated an elevation of (170.4 feet NAVD) (0.3-percent-annual-chance recurrence interval) on Cobbosseecontee Lake (Reference 8). In December 1973, similar conditions brought the elevation of Cobbosseecontee Lake up to (169.7 feet NAVD) (0.6-percent-annual-chance). The same storm produced a peak of (244.2 feet NAVD) (1.5-percent-annualchance recurrence interval) on Wilson Pond, and a peak of (271.8 feet NAVD) (1- percent-annual-chance recurrence interval) on Cochnewagon Lake. These 1973 high water marks were noted by local residents; and USGS personnel ran levels to several of these observed marks to establish flood levels for this event. USGS gaging stations are located on Cobbosseecontee Lake in East Winthrop, Maine, and on Cobbosseecontee Stream in Gardiner, Maine. Peak elevations for the December 1973 storm are available for several of the other detailed study lakes in Readfield and Winthrop. On Upper and Lower Narrows Ponds, the peak elevation was (176.0 feet NAVD) with a 1-percent-annual-chance recurrence interval; on Maranacook Lake, the peak was (214.9 feet NAVD) with a recurrence interval of 400 years; on Torsey Lake, during this same storm event, an elevation of (264.9 feet NAVD) with a 0.25-percent-annual-chance recurrence interval; and on Wilson Pond, the peak was (244.2 feet NAVD) with a 1.5- percent-annual-chance recurrence interval. Manchester and West Gardiner are also affected by flooding along Cobbosseecontee Stream. Flow estimates made downstream in Gardiner found the December 1973 peak flow to be in excess of 6,000 cfs. This flood was a 0.5- percent-annual-chance event. Town officials of Gardiner have requested that the U.S. Coast Guard icebreaker open up the river when snow and ice accumulation are considerable. The Collins Mill Dam, located near the Manchester-West Gardiner corporate limits affects the Cobbosseecontee Stream stages for approximately 4,000 feet in the lowest reach of the stream in Manchester. 21

27 Flooding problems in Oakland occur principally along the shores of Salmon and Messalonskee Lakes. The Town of Sidney also has flooding problems along Messalonskee Lake. Flooding problems in the Town of Rome occur principally along the shores of Long and Great Ponds. The history of floods in Waterville has indicated that most flooding occurs in the winter or early spring months as a result of heavy rainfall on snow-covered or frozen ground. The major flood damage in Waterville is to single and multifamily residences, businesses, a factory complex, two parks, utilities, roads, and bridges. Properties on the Kennebec River, near the Lockwood Dam and on Messalonskee Stream from the Union Gas Project Dam to the North Street bridge experience the most frequent flood problems. Historic data shows that floods have occurred on Messalonskee Stream in 1936, 1954, 1973, 1984, and Discharge and elevation data for these floods is very limited in Waterville. The 1936, 1954, and 1973 floods appear to have recurrence intervals in excess of 2-percent-annualchance, while the 1984 flood is estimated to have a recurrence interval of 10- percent-annual-chance. Most floods in the Androscoggin River basin occur in the spring when heavy snowmelt and rains combine to cause significant runoff. However, the fall season can also bring floods as a result of hurricanes and storms. Flooding in Wayne is limited to Pocasset and Androscoggin Lakes; and Berry, Dexter, Wilson, and Love Joy Ponds. Significant flooding in Wayne has occurred in March 1936, December 1973, and most recently in April Following the April 1987 flood, all sources of flooding were visited by personnel from the USGS Office in Augusta. During the 1987 flood, the high level of Pocasset Lake was due to heavy runoff and the high elevation of Androscoggin Lake that affected the getaway head at the outlet of the lake. High levels in Androscoggin Lake were due to a reversal of flow in its outlet stream, the Dead River, which backed into the lake. During flood peaks, elevations in the Androscoggin River (288.3 feet NAVD) at the mouth of the Dead River exceed those in Androscoggin Lake (284.7 feet NAVD). When this occurs the Dead River reverses direction and flows upstream back into Androscoggin Lake. This situation apparently has occurred frequently enough in the past for the silt-laden Androscoggin River to form a visible delta in the lake (Reference 12). The 1-percent-annual-chance flood elevation, taken from the FIS for the Town of Livermore, on the Androscoggin River was (289.4 feet NAVD) (Reference 13). The April 1987 flood was slightly higher than a 1.3-percent-annual-chance recurrence interval event in the Androscoggin River near Wayne. Flood elevations on Pocasset Lake for the 1987 flood were (288.5 feet NAVD). During the 1987 flood, Berry Pond was at an elevation of approximately 2 feet above a 1-percent-annual-chance flood elevation due to the bridge at the outlet of the lake failing. Dexter Pond was at approximately a 1-percent-annual-chance flood level and Wilson Pond was lower 22

28 than a 1-percent-annual-chance level due to foresight on the part of the operators of the dam who took steps to make space available in the pond for such a flood. Flood elevations during the 1936 flood on the Androscoggin River, Androscoggin Lake, and Pocasset Lake were (292.6, 288.7, and feet NAVD), respectively. Three flooding events were identified within the Kennebec River Basin after 1987 (Reference 14). The Kennebec River and Cobbosseecontee Stream flooded in March and December of 2003 as a result of ice jams. Both flooding events were estimated by the USGS to be below the 50-percent-annual-chance flood in the City of Gardiner, ME. Flooding of the Kennebec and Cobbosseecontee Rivers in April 2005 was estimated by USGS to be between the 20, and 10-percent-annualchance flood. The April 2005 flooding event impacted the Cities of Augusta, Gardiner and Hallowell. Flooding along the Kennebec River in April 2008 (19.14 feet NAVD) reached similar gage heights in the City of Augusta as the April 2005 flooding event (19.49 feet NAVD) (Reference 15). 2.4 Flood Protection Measures Flood protection measures for Kennebec County have been compiled and are summarized below. The natural stream flow of the Kennebec River, which flows through parts of Augusta, Chelsea, Clinton, Farmingdale, Gardiner, Hallowell, Pittston, Randolph, Sidney, Waterville, and Winslow, is altered by several hydroelectric plant reservoirs located upstream from Kennebec County. The structure that has the most pronounced effect on normal flow of the Kennebec River is the dam at Solon, 55 miles upstream from Waterville. It is a low-headed dam operated to maintain a discharge of 3,600 cfs at Madison, 14 miles downstream. When inflow at the dam is more than 4,000 cfs, it no longer controls river discharges. The most downstream dam on the Kennebec River, Edwards Dam, was removed in 1999, at the head of the tidal effect in Augusta. This low-head dam is operated for power generation. The Lockwood Dam and Scott Paper Company Dam are also located on the Kennebec River. The major storage lakes and their capacities in billions of cubic feet are as follows: Moosehead Lake , Indian Pond - 3.2, Flagstaff Lake , and Wyman Pond These reservoirs have a dampening influence on peak flows downstream. Large amounts of potential flood control exist in the upper part of the Kennebec River basin, particularly in Moosehead and Flagstaff Lakes. During the 1987 flood, the reservoirs in the basin above the USGS gaging station at Bingham (station No , drainage area 2,715 square miles), which represent 50 percent of the watershed above Augusta, contributed only about 25 percent of the peak flow at Augusta (Reference 7). Because these reservoirs are regulated primarily for power generation, the potential for major flooding exists at a time when they are at or near capacity and could offer little appreciable flood control. However, under normal operation the reservoirs are lowered before peak spring runoff. The most downstream dam on the Kennebec River is the Cushnoc Dam at 23

29 the head of the tidal effect in Augusta. This low-head dam is operated for power generation and to process water for mills. When the discharge at Augusta is more than 113,000 cfs, the dam is ineffective in controlling flooding downstream. In the past, the Coast Guard has often been requested to break open the channel in the Kennebec River when ice cover is thick and flood potential is high. These conditions are most likely to occur in the month of March. The numerous lakes and ponds in the Belgrade chain, although not designed as flood protection reservoirs, do exert an impact in reducing flood peaks. The Salmon Lake outlet dam is generally operated to lower lake levels to prevent flooding during periods of anticipated high water. Prior to 1977, Central Maine Power Company monitored water-surface elevations of major lakes in the basin of which the Towns of Belgrade, Oakland, Rome, and Sidney are a part. After that date, the company sold their water rights on these lakes, except for Messalonskee Lake, to a private enterprise. In 1984, the Belgrade Lakes Association bought the water rights on Salmon Lake, and Great and Long Ponds and currently manages the lake levels. The dam between Long Pond and Great Pond are generally operated to lower lake levels to prevent flooding during periods of anticipated high water. The Cobbossecontee Stream basin includes over 20 lakes and ponds. The largest of these lakes (in downstream order) and their capacities in million cubic feet are as follows: Torsey Lake, 441; Maranacook Lake, 2,000; Annabessacook Lake, 1,070; and Cobbosseecontee Lake, 6,810. Although these lakes are not used as flood control reservoirs, they do have a great dampening influence in peak flows downstream. The Town of Manchester controls release from the dam at the outlet of Cobbosseecontee Lake and when high water periods are anticipated, the operator tries to lower lake levels to prevent flooding. This dam will reduce the 1- percent-annual-chance flood peak discharge. The dam at the outlet of Woodbury Pond controls the water-surface elevation of the Tacoma Ponds. The Cobbossee Watershed District monitors water-surface elevations of major lakes in the Cobbosseecontee basin of which the Towns of Manchester, Monmouth, Readfield, West Gardiner, and Winthrop are a part. This agency has been granted the authority by the State of Maine to manage water levels within the Cobbosseecontee basin. The City of Augusta has adopted a shoreland zoning ordinance which covers most of the land within 250 feet of the normal high water of the major bodies of water in the City. This ordinance was enacted to restrict development and land use in the floodplain. There are seven dams located on the Sebasticook River in the Towns of Benton, Burnham, Hartland, Pittsfield, and Winslow. The original Benton Falls Dam was a hydro-mechanical facility breached in the flood of Construction of the new Benton Falls Dam hydroelectric facility, located 4.6 miles upstream of the confluence of the Sebasticook and Kennebec Rivers, began in The Benton 24

30 Falls Dam is designed to operate in run of river mode. Three hydraulically operated deep fixed-wheel sluice gates are designed to automatically maintain a pond elevation between (84.4 feet NAVD) and (86.4 feet NAVD). The dam's three sluice gates are designed to discharge a flow up to the 1-percent-annualchance event while maintaining a pond elevation of (86.4 feet NAVD) (Federal Energy Regulatory Commission, 1987). The natural flow of the Sebasticook River is altered very little at times of major flooding in the Town of Clinton. The Towns of Chelsea and Farmingdale have adopted the Minimum Shoreland Zoning Ordinance as required by the State of Maine Shoreland Zoning Act. The shorelands located within 250 feet, horizontal distance, of the normal high water marks of the Kennebec River, Togus Stream, Tinkham Pond, Jimmie Pond and Hutchinson Pond have been zoned resource protection. There will be no new residential, commercial, or industrial structures allowed in the Resource Protection Zone. Since 1969, flooding has caused considerable damage in the City of Gardiner. The three dams located on Cobbosseecontee Stream in Gardiner are not able to retain large volumes of water. The Gardiner Water District Dam is the only one of the three to offer any significant flood protection. This dam creates a pond that holds the ice, keeping it from breaking and flowing downstream and possibly causing jams. The City of Gardiner has also adopted the Minimum Shoreland Zoning Ordinance (Reference 16). The land along the Kennebec River from the Richmond-Gardiner corporate limits to a point approximately 900 feet south of the intersection of Kingsbury Street and State Route 24, is zoned Resource Protection. This zone generally extends from the normal high-water mark to the Maine Central Railroad tracks, but may go west of the tracks or may not reach as far as the tracks in some locations. The shoreland of Rolling Dam Brook is also zoned Resource Protection for a horizontal distance of 250 feet from the normal high-water mark for the entire length of the stream in Gardiner. The land along Cobbosseecontee Stream, from the Gardiner-Richmond corporate limits to the bridge at Cobbossee Avenue is zoned as the Shoreland District for a horizontal distance of 250 feet (in most places) from the normal high-water line, downstream of the Gardiner Water District pumping station. This 250-foot Shoreland District extends only along the west side of the stream to the Maine Central Railroad bridge near Summer Street. Any residential development in this zone is by special exception permit only, must be set back 100 feet from the normal high-water elevation, and must not encroach the regional floodplain. The City of Hallowell has zoned the land along the Kennebec River that is below an elevation of 21 feet as shoreland. No future building, except for boat-landing facilities, is permitted in a 25-foot wide strip along the bank of the Kennebec River. The Town of Manchester has adopted a flood plain zone which includes all areas 25

31 subject to flooding by the 1-percent-annual-chance flood. This floodplain zone will supersede and supplement other zoning designations. The Town of Pittston has also adopted the Minimum Shoreland Zoning Ordinance. The shorelands located within 250 feet, horizontal distance, of the normal high-water marks of the Kennebec River, the Eastern River, Togus Stream, Joys Pond, and Nehumkeag Pond have been zoned to restrict development (Reference 17). Most of the Kennebec River shoreline in the town is zoned as resource protection. There is a small strip of shoreland near Ripley Road, which has been designated as Limited Residential and Recreation Districts. The Eastern River from the Dresden-Pittston corporate limits to Kelly Road, and all of the Togus Stream shorelands are zoned as resource protection. The shorelines of Nehumkeag and Joys Ponds are zoned as Resource Protection Zone or Limited Residential and Recreation Districts. There will be no new residential, commercial, or industrial structures allowed in the Resource Protection Zone, and such structures will be permitted in the Limited Residential and Recreation Zone by permit only. In the Town of Randolph, the shorelands located within 250 feet (horizontal distance) of the normal high-water marks of the Kennebec River, and within 75 feet of the normal high-water marks of Togus Stream and Togus Brook have been zoned to restrict development as well, according to the Minimum Shoreland Zoning Ordinance. The land between the Kennebec River and State Route 9, extending from Belmont Avenue north to the corporate limits, is zoned Resource Protection. Parts of the area along the river from the mouth of Togus Brook south to Togus Stream and west of State Route 27, as well as part of the area bordering Togus Stream, is also zoned Resource Protection. The Town of West Gardiner has adopted a shoreland zoning ordinance which covers most of the land area within 250 feet, horizontal distance, of the normal high water of Cobbosseecontee Stream and Cobbosseecontee Lake (Reference 16). Additionally, the Towns of Belgrade, China, Clinton, Litchfield, Monmouth, Oakland, Readfield, Rome, Wayne, and Winthrop have adopted the minimum shoreland zoning ordinance as required by the State of Maine (Reference 16). There are no authorized or proposed flood control measures in the Town of Wayne. The only existing structure built for some degree of flood control on Androscoggin Lake is the Dead River Dam located approximately 0.5 mile downstream of the State Route 219 highway bridge. This structure was completed in 1933 and is a concrete retaining wall 171 feet long with an elevation on top of the wall of 274 feet. The original purpose of the dam was to prevent debris and chemical wastes in the Androscoggin River from being carried into Androscoggin Lake during periods of spring runoff or minor flooding. The dam was built with one-way wooden flap gates to allow normal downstream flow and to prevent reverse flow from the Androscoggin River from entering the lake. The dam was 26

32 originally constructed with 4 foot flashboards. These flashboards have been destroyed and the existing dam will not even prevent normal spring runoff on the Androscoggin River from entering the lake (Reference 12). Although there are no flood protection measures existing at this time which affect flooding along the Sebasticook River in the Town of Winslow, the Fort Halifax Dam was removed in There are no major flood control structures in the Town of China, but dams at the outlets of China Lake and Branch Pond offer limited control of minor floods. Also, there are no flood protection measures that affect flooding along Messalonskee Stream in the City of Waterville. 3.0 ENGINEERING METHODS For the flooding sources studied by detailed methods in the community, standard hydrologic and hydraulic study methods were used to determine the flood-hazard data required for this study. Flood events of a magnitude that is expected to be equaled or exceeded once on the average during any 10-, 50-, 100-, or 500-year period (recurrence interval) have been selected as having special significance for floodplain management and for flood insurance rates. These events, commonly termed the 10-, 50-, 100-, and 500-year floods, have a 10-, 2-, 1-, and 0.2-percent chance, respectively, of being equaled or exceeded during any year. Although the recurrence interval represents the long-term, average period between floods of a specific magnitude, rare floods could occur at short intervals or even within the same year. The risk of experiencing a rare flood increases when periods greater than 1 year are considered. For example, the risk of having a flood that equals or exceeds the 1-percent-annual-chance flood in any 50-year period is approximately 40 percent (4 in 10); for any 90-year period, the risk increases to approximately 60 percent (6 in 10). The analyses reported herein reflect flooding potentials based on conditions existing in the community at the time of completion of this study. Maps and flood elevations will be amended periodically to reflect future changes. 3.1 Hydrologic Analyses Hydrologic analyses were carried out to establish peak discharge-frequency relationships for floods of the selected recurrence intervals for each flooding source studied by detailed methods affecting the county. For each community within Kennebec County that has a previously printed FIS report, the hydrologic analyses described in those reports have been compiled and are summarized below. Precountywide Analyses Because of the potential for ice-jam flooding along the Kennebec River, hydrologic analyses were performed for both free-flow and ice- jam events. For free-flow events on the Kennebec River, discharges were determined using daily mean peak flow records from the Scott Paper Company Dam at Waterville. 27

33 These records were published by the USGS from 1892 to 1935 at gaging station No Flow records from 1936 to 1977 were obtained from the Scott Paper Company (records past 1977 could not be located). A log-pearson Type III analysis of these peak flows was performed to compute preliminary flood discharges (Reference 18). Final free-flow discharges were computed by adjusting the preliminary discharges for the difference between daily mean and instantaneous peaks and the difference in drainage area between Waterville and the other communities on the Kennebec River, including Augusta, Chelsea, Farmingdale, Gardiner, Hallowell, Pittston, Randolph, Sidney, and Winslow. The daily mean-to-instantaneous peak adjustment was developed based on the relationship between daily mean and peak flood discharges at the USGS gage on the Kennebec River in North Sidney. The adjustment varied from 9.9 percent for the 10-percent-annual-chance flood to 17.8 percent for the 0.2-percent-annualchance flood. The adjustment for drainage area was computed using the formula, Q s = Q g (A s /A g ) a where Qs is the discharge at the site, Q, is the discharge at the gage, A s is the drainage area of the site, A g is the drainage area of the gage, and "a" is an exponent. For the Kennebec River from the Scott Paper Company Dam in Winslow to Augusta, the value of "a" varies as follows: 10-percent-annual-chance recurrence interval, "a" equals 0.52; 2-percent-annual-chance recurrence interval, "a" equals 0.44; 1-percent-annual-chance recurrence interval, "a" equals 0.41 ; and 0.2-percent-annual-chance recurrence interval, "a" equals For the Kennebec River in Winslow to Bingham, the value varies as follows: 10-percentannual-chance recurrence interval, "a" equals 1.83; 2-percent-annual-chance recurrence interval, "a" equals 2.11; 1-percent-annual-chance recurrence interval, "a" equals 2.23; and 0.2-percent-annual-chance recurrence interval, "a" equals In Sidney, the peak flow from the 1987 flood was used in the calibration run. The 1987 flood produced a flow of 232,000 cfs at the USGS gage in North Sidney, and was classified as greater than a 1-percent-annual-chance flood event (Reference 6). For ice-jam events, discharges were determined for the Kennebec River using the general method described above for free-flow events. The difference between the two analyses is the peak daily discharges used in the log-pearson Type III analysis. For the ice-jam hydrologic computations, peak daily discharges were tabulated for the ice-jam season only. The ice-jam season was assumed to extend from December 20 to April 15. Peak daily discharges at Waterville during the icejam season were only available for the years 1892 to For the May 4, 1988, Town of Benton FIS, flood discharges on the Kennebec River were computed utilizing the following data: a log-pearson Type III analysis of former USGS gage No at Waterville (described above; Reference 18); and a log-pearson Type III analysis of USGS gage No in Bingham with a period of record from 1907 to 1910 and from 1930 to the present 28

34 (Reference 19). Discharges at sites in Benton and Winslow were computed using the drainage adjustment equation described above for Kennebec River. For the Sebasticook River in Benton and Winslow, the value of a (from the drainage area adjustment equation explained above) equals 0.80 for all recurrence intervals. Peak discharges were determined using river discharge records from the USGS gaging station near Pittsfield published from 1928 to 1996 as station number A log-pearson Type III analysis of these peak flows was done to compute flood discharges. Benton flood discharges were computed by adjusting the USGS gaging station discharges at Pittsfield for the difference in drainage area between Benton and the USGS gage. In Augusta, the values of the desired magnitude flow for Bond Brook at the 10-, 2-, and 1-percent-annual-chance intervals were obtained using equations developed by R. A. Morrill, whose methods are outlined in the USGS Open-File Report (Reference 20). These equations relate flood flows to the basin characteristics: drainage areas, main channel slope, and storage area in the basin. This method was also adapted for calculating flow for the 0.2-percent-annualchance interval. The Town of Clinton is located directly across the Kennebec River from the Town of Fairfield and results determined for the Town of Fairfield (Somerset County) were used. The SCS calculated 1-percent-annual-chance flood discharges for the Kennebec River as part of the precountywide FIS for the Town of Fairfield (Somerset County) (Reference 21). The peak flow calculations were based on peak flow data at the Winslow Dam provided by Scott Paper Company and flow data from the USGS stream gage, Kennebec River at Bingham (station No ). The SCS, during the preparation of the precountywide FIS for the Town of Benton, determined the 1-percent-annual-chance flood discharge for the Sebasticook River at the Clinton-Benton corporate limits (Reference 22). The SCS based the determination on a log-pearson Type III analysis of peak flow data from the upstream USGS gage near Pittsfield (Station No , drainage area 572 square miles) for the period (Reference 18). For the precountywide Clinton FIS, the log- Pearson Type III analysis for the Pittsfield gage was updated to include the 1987 flood peak. Results calculated by the SCS fell within the 90 percent confidence interval for the updated analysis and therefore, SCS 1-percent-annual-chance flood determinations at the Clinton- Benton corporate limits were used in the precountywide Town of Clinton FIS. Peak discharges for the Sebasticook River, upstream of the Clinton- Benton corporate limits was calculated using the drainage area equation described above for the Kennebec River. The exponent a was determined to be 0.51 by substituting calculated 1-percent- annual-chance flood discharges at the Clinton- Benton corporate limits and the Pittsfield gage and their respective drainage areas into the equation and solving. In the Town of China, the 1-percent-annual-chance flow for the West Branch Sheepscot River, Meadow Brook, and the outlet of Branch Pond were determined 29

35 by applying a USGS regression equation (Reference 20). The flow of the West Branch Sheepscot River was changed just upstream of the first major tributary from the downstream limit of detailed study. This adjustment was made on the basis of the ratio of the change in drainage area. The primary source of data for China Lake was 17 annual maximum lake elevations from records furnished by the Kennebec Water District. The Cobbosseecontee Stream Basin is composed of over 20 lakes and ponds. Many of the communities in Kennebec County are within this basin, including Gardiner, Litchfield, Manchester, Monmouth, West Gardiner, and Winthrop. The largest of these lakes (in downstream order) and their capacities in million cubic feet are as follows: Torsey Lake--441, Maranacook Lake--2,000; Annabessacook Lake--1,070, and Cobbosseecontee Lake--6,810. Although these lakes are not regulated as storage reservoirs, they do have a considerable dampening influence on peak flows downstream. The analysis of Cobbosseecontee Lake in Litchfield, Manchester, Monmouth, West Gardiner, and Winthrop was based on a normal-pearson Type III distribution of annual peak elevation data (Reference 23). The principal source of data for Cobbosseecontee Lake was the record of lake elevations maintained by the Gardiner Water District for the period In September 1975, the USGS established a gaging station (no ) on Cobbosseecontee Lake at East Winthrop, Maine. These records and historic recordings of the 1936 peak elevation also were included in the analysis. The principal sources of data for Cobbosseecontee Stream in Gardiner, Litchfield, Manchester, and West Gardiner are records published by the USGS. These records were published for USGS gaging station No on Cobbosseecontee Stream at Gardiner for the period from 1890 to 1964 (74 years of record). These data were collected at the Gardiner Water District dam operated by the Gardiner Water District. In 1976, the USGS re-established the gaging operation on Cobbosseecontee Stream. In the 1977 edition of Water Resources Data for Maine, these records were again published for USGS gaging station No (Reference 19). The value of the 1-percent-annual-chance peak discharge at the gaging station was obtained from a log-pearson Type III distribution of annual peak flow data, according to the instruction in Water Resources Council Bulletin No. 17 (Reference 23). A significant flood event occurred during 1973 when no gaging station was in operation. The peak flow for this event was approximated using the flow over dam method (Reference 24). The peak was then included in the log-pearson Type III distribution. Due to the difference in the drainage area of Cobbosseecontee Stream from Gardiner (217 square miles) to the Litchfield-West Gardiner corporate limits (140 square miles) and to the Manchester-West Gardiner Town line (139 square miles), decreases in discharge were required to make the peak flow representative of Litchfield, Manchester, and West Gardiner. These decreases were based on drainage area adjustments and information contained in USGS Open-File Report (Reference 20). 30

36 Frequency data for Rolling Dam Brook in Gardiner were determined by using regional regression equations (Reference 20). The 1-percent-annual-chance flood elevation for Pleasant Pond, including Upper Pleasant Pond, was obtained from the hydraulic analyses for Cobbosseecontee Stream. Flood discharges in Cobbosseecontee Stream, which drains the outlet of Pleasant Pond, control flood elevations in the pond. Hydrologic analyses were carried out using the log-pearson Type III method to establish the maximum mean daily discharge-frequency relationships for floods of the selected recurrence intervals for Togus Stream in the Towns of Pittston and Randolph (Reference 23). The results from this analysis were increased by 10 percent to simulate instantaneous peak flows because the flood-flow data for Waterville were expressed as maximum daily discharges. The 10 percent factor was found to be the average amount by which instantaneous peak flows exceeded daily flows at Waterville, based on a comparison of 10 determinations of instantaneous peak flow over a dam with the corresponding daily flow. The values of the 10-, 2-, 1-, and 0.2-percent-annual-chance peak discharges for Togus Stream in Pittston and Randolph were obtained using equations developed by R.A. Morrill (Reference 20). Discharge-frequency data for the Eastern River, also in Pittston, were determined from equations based on multiple-regression analyses of data from USGS gaged sites in Maine and adjacent areas of bordering states (Reference 20). The 1- percent-annual-chance flood elevations along the Eastern River from the downstream corporate limits to a point approximately 2.5 miles upstream are controlled by backwater from the Kennebec River. For Togus and Threecornered Ponds in Augusta, Annabessacook Lake, Cochnewagon Lake and Wilson Pond in Monmouth, the Tacoma Ponds, including Woodbury Pond, Sand Pond, Buker Pond, Jimmy Pond, and Little Purgatory Pond in Litchfield and Monmouth, the ponds in Readfield, and Upper Narrows Pond, Lower Narrows Ponds, Maranacook Lake, Annabessacook Lake, Berry Pond, Dexter Pond, and Wilson Pond in Winthrop, the values of the desired magnitude flows at the two outlet dams for the 10-,2-, and 1-percent-annualchance intervals were obtained using the methods outlined in the USGS Open- File Report (Reference 20). The flows for the 0.2-percent-annual-chance flood interval were obtained by a formula developed by the Augusta, Maine, USGS office as follows: P 500 = 75.7 A S ST P 500 is discharge for a 0.2-percent-annual-chance recurrence interval; A is the drainage area in square miles; S is main channel slope in feet per mile; and ST is the area of lakes and ponds expressed as the percentage of drainage area plus one percent. 31

37 Discharges on Messalonksee Stream in Waterville were generated from the SCS Technical Release No. 20 hydrologic evaluation model (Reference 25). This information was checked against historic data provided by the USGS, the Maine Department of Transportation, and the Central Maine Power Company. In the Town of Wayne, flood elevations for Berry, Dexter, and Wilson Ponds were taken from the precountywide FIS for the Town of Winthrop (Reference 26). In the Winthrop study, flood discharges were related to elevations at the Wilson Pond outlet dam using USGS regression techniques (References 20 and 27). It was assumed that the dam on Wilson Pond controlled flood elevations for both Dexter and Berry Ponds as well. Field analyses made during the April 1987 flood proved this was not correct and 1-percent-annual-chance flood elevations for Dexter and Berry Ponds needed to be recomputed. Resultant 1-percent-annualchance flood discharges are 510 feet per cubic second for Berry Pond, 570 feet per cubic second for Dexter Pond, and 930 feet per cubic second for Wilson Pond. The 1-percent-annual-chance flood elevation for Wilson Pond determined in the Winthrop study is still valid. This elevation was used in this study as the 1- percent-annual-chance flood elevation for Wilson Pond and the starting-water elevation in Dexter and Berry Ponds (Reference 28). Flood discharges were routed from Wilson Pond upstream through the bridges that separate both Dexter and Berry Ponds. The bridges and appropriate cross-sections were surveyed by the USGS. The 1-percent-annual-chance flood of Lovejoy Pond in the Town of Wayne was taken from the precountywide FIS for the Town of Readfield (Reference 29). In the Readfield study, flood discharges were computed using regression equations developed by the USGS (Reference 20). The resultant 1-percent-annual-chance flood discharge of 1,400 feet per cubic second was routed over the outlet dam to determine the flood elevation (Reference 27). Countywide Analyses No new hydrologic analyses were performed in Kennebec County. Peak discharge-drainage area relationships for flooding sources in Kennebec County are shown in Table 6, Summary of Discharges. For the Kennebec River, only the drainage area-peak discharge relationships for free-flow events are shown. 32

38 TABLE 6 SUMMARY OF DISCHARGES FLOODING SOURCE AND LOCATION DRAINAGE AREA (SQUARE MILES) PEAK DISCHARGES (CUBIC FEET PER SECOND) PERCENT PERCENT PERCENT PERCENT ANNUAL ANNUAL ANNUAL ANNUAL CHANCE CHANCE CHANCE CHANCE BOND BROOK At Confluence with Kennebec River ,390 2,310 2,800 4,210 At first crossing of Bond Brook Road ,110 1,830 2,230 3,360 Upstream of Confluence of Stone Brook ,500 1,820 2,760 COBBOSSEECONTEE STREAM At Confluence with Kennebec River 219 3,850 5,290 5,910 7,380 At inlet to Pleasant Pond near the Sagadahoc-Kennebec County limits 213 3,850 5,290 5,910 7,380 Gardiner- West Gardiner corporate limits 212 3,850 5,290 5,910 7,380 At Pond Road 186 3,400 4,660 5,180 6,500 At the Litchfield-West Gardiner upstream corporate limits 140 * * 4,140 * At Collins Mill Road 140 2,710 3,740 4,140 5,200 At West Gardiner-Manchester corporate limits 139 2,710 3,740 4,140 5,200 At Confluence with Cobbosseecontee Lake 131 2,570 3,540 3,920 4,930 EASTERN RIVER Downstream from Confluence of Kimball Brook 19.2 * * 2,250 * Upstream from Confluence of Kimball Brook 16.5 * * 1,960 * KENNEBEC RIVER At the Lincoln-Kennebec County limits 5,822 * * 233,000 * *Data not computed 33

39 TABLE 6 SUMMARY OF DISCHARGES (continued) FLOODING SOURCE AND LOCATION KENNEBEC RIVER - continued DRAINAGE AREA (SQUARE MILES) PEAK DISCHARGES (CUBIC FEET PER SECOND) PERCENT PERCENT PERCENT PERCENT ANNUAL ANNUAL ANNUAL ANNUAL CHANCE CHANCE CHANCE CHANCE At Sagadahoc-Kennebec County limits 5, , , , ,000 At Randolph-Pittston corporate limits 5, , , , ,000 Upstream of the Confluence of Cobbosseecontee Stream 5, , , , ,000 At Farmingdale-Gardiner corporate limits 5, , , , ,000 At Chelsea-Randolph corporate limits 5, , , , ,000 At Farmingdale-Hallowell corporate limits 5, , , , ,000 At Chelsea-Augusta corporate limits 5, , , , ,000 At Augusta-Hallowell corporate limits 5, , , , ,000 At Cushnoc Dam 5, , , , ,000 At North Sidney Gage (No ) 5, , , , ,000 At downstream Waterville-Winslow corporate limits 5, , , , ,300 Downstream of Confluence of Sebasticook River 5, , , , ,900 At Scott Paper Company Dam 4,226 98, , , ,000 At Bridge Street 4,220 98, , , ,300 At Shawmut Dam 4,212 97, , , ,400 At Hinckley Bridge 4,146 * * 153,300 * MEADOW BROOK Approximately 4,180 feet downstream of Dirigo Road 0.97 * * 168 * Approximately 100 feet upstream of Dirgo Road 0.90 * * 168 * *Data not computed 34

40 TABLE 6 SUMMARY OF DISCHARGES (continued) FLOODING SOURCE AND LOCATION DRAINAGE AREA (SQUARE MILES) PEAK DISCHARGES (CUBIC FEET PER SECOND) PERCENT PERCENT PERCENT PERCENT ANNUAL ANNUAL ANNUAL ANNUAL CHANCE CHANCE CHANCE CHANCE MESSALONSKEE STREAM At Confluence with Kennebec River 207 2,680 4,270 4,890 6,200 At County Road 189 2,110 2,740 2,990 3,440 SEBASTICOOK RIVER At Confluence with Kennebec River ,900 19,000 20,700 24,600 At upstream Winslow corporate limits ,800 17,600 19,200 22,900 At State Route ,300 18,700 20,700 25,300 Above Confluence of Fowler Brook ,100 18,500 20,400 24,900 At Clinton-Benton corporate limits 850 * * 19,000 * Above mouth of Fifteen Mile Stream 757 * * 17,900 * TOGUS STREAM At Confluence with Kennebec River ,170 1,790 2,110 2,990 Upstream of Confluence of Stony Meadow Brook ,200 1,410 2,000 At Chelsea Pittston corporate limits ,200 1,410 2,000 Upstream of Culvert under Windsor Road ,010 1,100 1,660 Upstream of Confluence with Chase Meadow Brook WEST BRANCH SHEEPSCOT RIVER Approximately 3,200 feet above downstream corporate limits (China) 20.2 * * 1,790 * *Data not computed 35

41 TABLE 6 SUMMARY OF DISCHARGES (continued) FLOODING SOURCE AND LOCATION DRAINAGE AREA (SQUARE MILES) PEAK DISCHARGES (CUBIC FEET PER SECOND) PERCENT PERCENT PERCENT PERCENT ANNUAL ANNUAL ANNUAL ANNUAL CHANCE CHANCE CHANCE CHANCE WEST BRANCH SHEEPSCOT RIVER - continued Approximately 1,490 Feet downstream of highway bridge in Village of Weeks Mills 18.6 * * 1,650 * *Data not computed A summary of peak elevation-frequency relationships is shown in Table 7 Summary of Stillwater Elevations. TABLE 7 SUMMARY OF STILLWATER ELEVATIONS FLOODING SOURCE AND LOCATION 10-PERCENT ANNUAL CHANCE ELEVATION (FEET NAVD) 2-PERCENT 1-PERCENT ANNUAL ANNUAL CHANCE CHANCE 0.2-PERCENT ANNUAL CHANCE ANDROSCOGGIN LAKE Entire shoreline within the Town of Wayne * * * ANNABESSACOOK LAKE Entire shoreline within the Town of Monmouth Town of Winthrop BELGRADE STREAM Above Wings Mills Dam in the Town of Mount Vernon * * * Below Wings Mills Dam in the Town of Mount Vernon * * * *Data not computed 36

42 TABLE 7 SUMMARY OF STILLWATER ELEVATIONS - (continued) FLOODING SOURCE AND LOCATION 10-PERCENT ANNUAL CHANCE ELEVATION (FEET NAVD) 2-PERCENT 1-PERCENT ANNUAL ANNUAL CHANCE CHANCE 0.2-PERCENT ANNUAL CHANCE BERRY POND For its entire length within the Town of Wayne * * * For its entire length within the Town of Winthrop * * * BRANCH POND For its entire shoreline within the Town of China * * 350 * BUKER POND For its entire shoreline within the Town of Litchfield * * * CHINA LAKE For its entire shoreline within the Town of China * * * COBBOSSEECONTEE LAKE Town of Manchester Town of Monmouth Town of West Gardiner Town of Winthrop Town of Litchfield * * * COCHNEWAGON LAKE Town of Monmouth DEXTER POND Entire shoreline within the Town of Wayne * * * Entire shoreline within the Town of Winthrop * * * *Data not computed 37

43 TABLE 7 SUMMARY OF STILLWATER ELEVATIONS (continued) FLOODING SOURCE AND LOCATION 10-PERCENT ANNUAL CHANCE ELEVATION (FEET NAVD) 2-PERCENT 1-PERCENT ANNUAL ANNUAL CHANCE CHANCE 0.2-PERCENT ANNUAL CHANCE ECHO LAKE Town of Readfield GREAT POND Entire shoreline within the Town of Belgrade * * * Entire shoreline within the Town of Rome * * * JIMMY POND Town of Litchfield * * * LITTLE COBBOSSEECONTEE LAKE Town of Winthrop LITTLE PURGATORY POND Town of Litchfield * * * LONG POND Entire shoreline within the Town of Belgrade * * * Entire shoreline within the Town of Rome * * * LOVEJOY POND Town of Readfield Entire shoreline within the Town of Wayne * * * LOWER TOGUS POND Town of Chelsea *Data not computed 38

44 TABLE 7 SUMMARY OF STILLWATER ELEVATIONS (continued) FLOODING SOURCE AND LOCATION 10-PERCENT ANNUAL CHANCE ELEVATION (FEET NAVD) 2-PERCENT 1-PERCENT ANNUAL ANNUAL CHANCE CHANCE 0.2-PERCENT ANNUAL CHANCE MARANACOOK LAKE Town of Readfield Town of Winthrop MESSALONSKEE LAKE Entire shoreline within the Town of Belgrade * * * Entire shoreline within the Town of Oakland * * * Entire shoreline within the Town of Sidney * * * PLEASANT POND Town of Litchfield * * * POCASSETT LAKE Entire shoreline within the Town of Wayne * * * SALMON LAKE Entire shoreline within the Town of Belgrade * * * Entire shoreline within the Town of Oakland * * * SAND POND Town of Litchfield * * * Town of Monmouth THREE CORNERED POND City of Augusta *Data not computed 39

45 TABLE 7 SUMMARY OF STILLWATER ELEVATIONS (continued) FLOODING SOURCE AND LOCATION 10-PERCENT ANNUAL CHANCE ELEVATION (FEET NAVD) 2-PERCENT 1-PERCENT ANNUAL ANNUAL CHANCE CHANCE 0.2-PERCENT ANNUAL CHANCE THREEMILE POND Entire shoreline within the Town of China * * * TOGUS POND City of Augusta TORSEY LAKE Town of Readfield UPPER AND LOWER NARROWS POND Town of Winthrop WILSON POND Town of Monmouth Entire shoreline within the Town of Wayne * * * Town of Winthrop WOODBURY POND Town of Litchfield * * * Town of Monmouth *Data not computed 3.2 Hydraulic Analyses Analyses of the hydraulic characteristics of flooding from the sources studied were carried out to provide estimates of the elevations of floods of the selected recurrence intervals. Flood profiles were drawn showing the computed watersurface elevations for floods of the selected recurrence intervals. Users should be aware that some flood elevations shown on the FIRM represent rounded whole-foot elevations and may not exactly reflect the elevations shown on the Flood Profiles or in the Floodway Data tables in the FIS report. For construction and/or floodplain management purposes, users are encouraged to use the flood elevation data presented in this FIS in conjunction with the data shown on the FIRM. 40

46 Cross section data for the below-water sections were obtained from field surveys. Cross sections were located at close intervals above and below bridges, culverts, and dams in order to compute the significant backwater effects of these structures. In addition, cross sections were taken between hydraulic controls whenever warranted by topographic changes. Locations of selected cross sections used in the hydraulic analyses are shown on the Flood Profiles (Exhibit 1). For stream segments for which a floodway was computed (Section 4.2), selected cross-section locations are also shown on the FIRM (Exhibit 2). The hydraulic analyses for this study were based on unobstructed flow. The flood elevations shown on the Flood Profiles (Exhibit 1) are thus considered valid only if hydraulic structures remain unobstructed, operate properly, and do not fail. All flood profiles were drawn showing computed water-surface elevations for floods of the selected recurrence intervals. For each community within Kennebec County that has a previously printed FIS report, the hydraulic analyses described in those reports have been compiled and are summarized below. Precountywide Analyses All cross sections, bridges and culverts were surveyed to obtain elevation data, structural data, and structural geometry. Cross sections were selected immediately below changes in stream configuration. Roughness coefficients (Manning s n ) were determined by field inspection at each cross section using a step-by-step procedure. Where feasible, transposed cross sections were used to reduce the number of surveyed cross sections. Transposed cross sections are surveyed sections which can be transferred either upstream or downstream to represent a location which is similar in valley shape. All elevations from the precountywide analyses were referenced from National Geodetic Vertical Datum of 1929 (NGVD), but were converted to North American Vertical Datum of 1988 for this countywide FIS. Refer to Section 3.4 Vertical Datum for more information about the conversion factor. For the 17.1 mile length of the Kennebec River that stretches across the Towns of Farmingdale, Chelsea, Clinton, Pittston, and Randolph, and the Cities of Augusta, Gardiner, and Hallowell, water-surface elevations of floods of the selected recurrence intervals were computed using the USGS E431 and J635 stepbackwater computer programs (References 30 and 31). The starting water-surface elevation for the Kennebec River was determined by using the 1-percent-annualchance coastal flood elevation on the Lower Kennebec River at Merrymeeting Bay, which was taken from the precountywide FIS for the City of Bath (Reference 32). That elevation was inserted into the WSPRO models used to prepare the precountywide FISs for the downstream communities of Bowdoinham, Dresden, Pittston, and Richmond; the models were calibrated to the 41

47 1936 and 1987 floods (References 33, 34, 35, and 36). The resulting computed water-surface elevation at the Richmond-Gardiner corporate limits from the calibrated WSPRO models was used as the starting water-surface elevation for the study of the Kennebec River (Reference 37). In the hydraulic analyses for the Kennebec River, it was necessary to combine the probability of flooding due to free flow events with the probability of flooding due to ice jams. This was done at each cross section using the equation: P(s) = P(si) + P(sq) P(si) x P(sq) where: P(s) = probability of a given stage being equaled or exceeded from either an ice-jam event or a free-flow event P(si) = probability of that stage being equaled or exceeded from an ice-jam event P(sq) = probability of that stage being equaled or exceeded from a free-flow event Free-flow stage-frequency curves were developed at each cross section using the USGS J635 step-backwater model calibrated using profile and discharge information available from the April 1987 flood (References 6 and 31). Flood discharges for the model were taken from the free-flow hydrologic computations. Ice-jam stage-frequency curves were developed at each cross section using the U.S. Army Corps Engineers (USACE) HEC-2 step-backwater model, calibrated using profile and discharge information available from the March 1936 flood (References 8 and 38). The 1936 flood was the worst ice-affected flood to occur on the Kennebec River in recorded history. In the calibration run, an ice jam blocking 71 percent of the channel was simulated at the Richmond-Dresden bridge at the head of Swann Island. Upstream of the jam, an ice thickness of three feet was assumed. A smaller ice jam was modeled upstream of Nehumkeag Island with an ice thickness of 2.5 feet. Flood discharges for the model were taken from the ice-jam hydrologic computations, described in Section 3.1 of this report. A necessary component in combining the probability of ice-jam floods with freeflow floods is the percentage of annual peak stages attributable to each type of flooding. Based on historic information it was assumed that 34 percent of annual peak stages on the Kennebec River are caused by ice jams (References 9 and 10). The combination of the stage-frequency curves for ice-jam and free-flow events resulted in a composite stage-frequency curve at each cross section within the community. Final flood elevations for each recurrence interval were obtained at each cross section from the curves. The flooding effect of ice jams on the Kennebec River was analyzed and determined not to extend upstream beyond the corporate limits of Augusta. Therefore, ice-jam flooding is not significant in the Towns of Sidney, Waterville, 42

48 and Winslow. The USGS operated temporary water-stage recorders in the Cities of Gardiner and Augusta from April to December The data obtained indicate that there would be less than a 0.1 foot tide effect for a discharge of 148,000 cfs in Augusta below the Cushnoc Dam, and in Chelsea. In Augusta, this flow has a recurrence interval of 40 years or an exceedance probability of The 10-percentannual-chance flood profile below the Cushnoc Dam could be as much as 0.5 foot higher in elevation in order to reflect the influence of normal high tide. In Chelsea, Gardiner, Pittston, and Randolph this flow has a recurrence interval of 15 years (approximately 7-percent-annual-chance), or an exceedance probability of The 10-year flood profile could be approximately 0.2 foot higher in elevation in order to reflect the influence of normal high tide. Cross-section data for the Kennebec River in Augusta, Chelsea, Farmingdale, Gardiner, Hallowell, Pittston, and Randolph, Bond Brook in Augusta, Cobboseecontee Stream in Litchfield, Manchester, and West Gardiner, and the Togus Stream in Chelsea, Pittston, and Randolph were obtained by photogrammetric methods using aerial photographs taken on November 12, 1976 (References 39 and 40). Below water data were obtained from field surveys. In Sidney, cross sections for the Kennebec River were obtained from the precountywide FISs for the Cities of Waterville and Augusta, Edwards Manufacturing Company (owners of the former Edwards Dam), and field surveys during the 1995 field season. Water-surface elevations of floods of the selected recurrence intervals were computed using the WSPRO step-backwater computer program (References 28 and 53). Starting water-surface elevations were taken from the June 1994 precountywide FIS for the City of Augusta. In Winslow, cross sections for the Kennebec River were obtained from the City of Waterville and Augusta FISs, and field surveys during the 1997 field season. Water-surface elevations of floods of the selected recurrence intervals for the Kennebec River were computed using RiverCAD, a step-backwater computer program (References 42, 56, and 57). Starting water-surface elevations for the Kennebec River were taken from the Town of Sidney Flood Insurance Study (Reference 55). The water-surface elevations for Bond Brook in Augusta were taken from profiles developed using the USGS step-backwater computer program model (Reference 41). The computed backwater elevations of the Kennebec River at the mouth of Bond Brook were the source of the starting water-surface elevations for Bond Brook. These elevations are in agreement with the historical flood marks. For Togus Stream in Chelsea, Pittston, and Randolph water-surface elevation of floods of the selected recurrence intervals were computed using the USGS E431 step-backwater computer program (Reference 30). The elevations for Togus Stream in Chelsea were determined from its mouth near Gardiner to Togus Pond 43

49 in Augusta. In the December, 1979, precountywide Town of Chelsea FIS, the computations were performed as part of the analyses of 7.2 miles of Togus Stream, which are part of the precountywide FISs for three Togus Stream communities, including Chelsea, Pittston and Randolph. The starting water-surface elevations for Togus Stream in Pittston were determined from normal depth calculations. These profiles are in agreement with historical flood marks. Roughness factors (Manning's n ) for Togus Stream were chosen on the basis of aerial photographs and field inspection at each cross section (Reference 52) Cross section data for the West Branch Sheepscot and Meadow Brook in China were obtained from USGS topographic maps (Reference 43). The water-surface elevations for the West Branch Sheepscot and Meadow Brook were taken from profiles developed using a USGS step-backwater computer program (Reference 28). The starting water-surface elevations used for the profile determinations were based upon the slope of each stream. The 100-year recurrence interval elevation of China Lake was determined by analyzing the 17 maximum annual lake elevations. The analysis was performed using the log-pearson Type III method (Reference 18). The 1-percent-annual-chance flood elevation of Threemile Pond in China was determined by using a technique for determining depths for T-year discharges in rigid boundary channels. This method is described in USGS Water Resource Investigations (Reference 44). The data necessary to make the computation were field surveyed. The 100-year elevation of Branch Pond was determined by passing the estimated 1-percent-annual-chance flow of 888 cfs through the dam and over the spillway (all gates wide open). The corresponding head was determined. The flood elevations for Branch Pond was based on the elevation of State Route 3 at the intersection of Parmeter Hill Road or established by the USGS for this study. This road elevation was then calibrated to known spot elevations in the Town of Palermo. Cross section data for the Sebasticook River in Clinton were obtained from USGS topographic maps at a scale of 1:24,000 with a contour interval of 10 feet (Reference 45). The below-water portions of the cross sections were obtained from field surveys. Bridge evaluations and geometry were surveyed in the field, and checked against construction plans of the Maine Department of Transportation, Bridge Division. For the Sebasticook River in Benton, water-surface elevations of floods of the selected recurrence intervals were computed using the USACE HEC-RAS computer program (Reference 42). Starting water-surface elevations were taken from the precountywide Town of Winslow FIS, described below. In Clinton, the starting water surface elevation for the Sebasticook River at the downstream corporate limits was determined using the FIS for the Town of 44

50 Benton (described above). Water-surface elevations of floods of the selected recurrence interval for the Sebasticook River were taken from profiles developed using a USGS step-backwater computer program model (Reference 28). The stepbackwater model for the Sebasticook River was checked using water level information obtained during the 1987 flood. Flood elevations observed during the 1987 flood are shown on the Sebasticook River profile at critical points throughout the entire length of the river within the Town of Clinton. The 1987 flood was the most severe since systematic collection of data was begun in October 1928 at the USGS stream gage near Pittsfield (Reference 46). Cross sections for the Sebasticook River in Winslow were obtained from aerial photographs flown in November 1984 at a scale of 1:14,400 (Reference 58). The below-water sections were obtained from field surveys. Water-surface elevations of floods of the selected recurrence intervals for the Sebasticook River were computed using the SCS WSP-2 computer program (Reference 59). Starting water-surface elevations for the Sebasticook River were started from a given slope at its confluence with the Kennebec River. The water-surface elevations for Cobbosseecontee Stream in Gardiner, Litchfield, Manchester, and West Gardiner were taken from profiles developed using the USGS E431 step-backwater computer program (References 30, 31, 41, 47, 48 and 49). The computations for the Manchester, Gardiner, and West Gardiner studies were performed as part of the analyses of 17.6 miles of Cobbosseecontee Stream. The computed backwater elevations of the Kennebec River at the mouth of Cobbosseecontee Stream were the source of the starting water-surface elevations used for the profiles on Cobbosseecontee Stream in Gardiner. These profiles were compared with historical flood marks. In Litchfield, the starting water-surface elevations for Cobbosseecontee Stream were taken from historical flood information and a rating curve developed by applying the flow over dam method at the Gardiner Water District Dam located 4.7 miles downstream from the Litchfield-West Gardiner corporate limits. The same was done in West Gardiner, where the Gardiner Water District Dam is located 0.8 mile downstream from the West Gardiner-Gardiner town boundary (Reference 24). The profiles were calibrated using historical flood information available from local landowners. In Manchester, the profiles were determined for Cobbosseecontee Stream from its mouth to the outlet of Cobbosseecontee Lake. The starting elevations used for the Manchester profile determinations were computed using historical flood information and a stage-discharge relation developed for Collins Mill Dam, located in West Gardiner, 1.1 miles downstream of the Manchester corporate limits. To determine the stage-discharge relationship for this dam, USGS personnel surveyed the dam and took current-meter measurements. This rating was extended on the basis of the stand flow-over-dam formulae, specifically the Francis Formula (Q = CLH 3/2), where Q is the discharge being studied, C is the coefficient of discharge, L is the length of the dam perpendicular to the direction of the flow, and H is the head on the dam (Reference 50). The coefficient of discharge (C) was also determined using the tables and graphs in U.S. Geological Survey Techniques of Water-Resources Investigations Book 3 which lists "C" 45

51 values for various types of dams (Reference 24). The actual "C" value used in the Francis Formula to compute the flood elevations at Collins Mill Dam was based on both of these "C" determinations. For Togus and Threecornered Ponds in Augusta, Annabessacook Lake, Cochnewagon Lake and Wilson Pond in Monmouth, the Tacoma Ponds, including Woodbury Pond, Sand Pond, Buker Pond, Jimmy Pond, and Little Purgatory Pond in Litchfield and Monmouth, the ponds in Readfield, and Upper Narrows Pond and Lower Narrows Ponds, Maranacook Lake, Annabessacook Lake, Berry Pond, Dexter Pond, and Wilson Pond in Winthrop, the elevations for the 10-, 2-, 1-, and 0.2-percent-annual-chance floods were derived as explained below. Elevations for the Tacoma Ponds are controlled by the dam at the outlet of Woodbury Pond. During reconnaissance, all available information was collected for historical highwater elevations on the lakes, including accounts from long-term residents. The construction, operation, and maintenance of the outlet dams was also investigated in order to rate the dams. This investigation led to the conclusion that a sophisticated study of high-water elevations in the lakes such as routing flows through the lake would be misleading. The reasons are that the dams were in poor physical shape and leaking, the fish ladders and trash racks are seldom cleaned, and a log of gate changes rarely exists. The original dams at the outlets of both the ponds have been altered and dates of these changes were not recorded. Accurate elevation records have not been kept at either of the dams in Augusta. However, since the December 1973 flood the dam operations have been recorded. To determine stage-discharge relations for each of the outlet dams controlling the ponds, the USGS made direct readings of pond elevations, surveyed the dams, and recorded their physical dimensions. Reference points were set upstream of the dams so the head on the dams could be computed for observed and measured flows. Current-meter measurements were made at each of the outlet dams, and a relationship between stages and discharges was determined for each dam. These ratings were extended on the basis of the Francis Formula (described above for Manchester). The measurements of flow to obtain values of "C" did not differentiate the flow for leakage or flow through deep gates. The coefficient of discharge was also determined using the tables and graphs in USGS Techniques of Water Resources Investigations, Book 3, Chapter 5, which list "C values for various dam types (Reference 24). The actual "C" value used in the Francis Formula to compute the flood elevation was based on both of these "C" determinations. The dams were rated as open when the "C values were determined. From the stage-discharge ratings drawn for the outlet dams, the elevations for the 10-, 2-, 1-, and 0.2-percent-annual-chance floods were derived. When high water is anticipated, the dam operators release as much water as possible. The Annabessacook Lake outlet dam has no provisions for flash boards. There is one deep gate in the dam which can be opened, and for all discharges where a coefficient of discharge C was computed, that gate was at a maximum 46

52 opening. The concrete dam at Cochnewagon Lake outlet has a spillway section and a small, deep gate. Woodbury and Wilson Ponds have deep gates that can be opened to pass peak flows. Maranacook Lake outlet dam in Winthrop has provisions for flash boards but USGS personnel did not observe any in place during the two-year study. There is one deep gate in the dam which can be opened and for all discharges where a coefficient of discharge was computed, that gate was at a maximum opening. Upper Narrows Pond and Lower Narrows Pond are not divided; their elevation is controlled by a small dam located on Lower Narrows Pond outlet brook. This is a small rock dam about 1 foot high which is not maintained or operated. This dam has very little effect, and the lake level is controlled more by the natural topography than by the dam. Berry Pond, Dexter Pond, and Wilson Pond are a chain of ponds with the same elevations, controlled by the Wilson Pond Dam in North Monmouth. This dam, owned and operated by the Globe-Albany Manufacturing plant, is approximately 14 feet high and has 3 deep gates. At least 2 of the gates are operable, and there are no flash boards. The Torsey Lake Outlet Dam near Old Kents Hill Road in Readfield has had no gate changes in the past two years and does not have flash boards. The crest of the dam is a drop inlet which accounts for such a small change in stage for increased flows. The Echo Lake Outlet Dam in Fayette and Mill Stream Outlet Dam near State Route 17 at Fayette Village are owned and operated by the Echo Lake Association, Inc. The Echo Lake Dam has two 5-foot sections of removable boards, and the operators try to maintain the pond at spillway level, approximate elevation (317.4 feet NAVD). Their present policy is to keep the water level high in the winter to minimize ice damage to the dam. High water would be allowed to flood the spillway and the Mill Pond (Fayette Pond) downstream from it. This pond is controlled by the Mill Stream Dam which also has two 5-foot sections of removable boards. The dam operators do open up the boards during periods of high water and release water to Lovejoy Pond. Lovejoy Pond Outlet Dam at North Wayne is privately owned as of 1978 and has not been operated or maintained for the past two years. This dam has a small, deep gate and no flashboards. Elevations computed for these ponds were substantiated by interviews with lifelong residents of the area and highway maintenance workers, visual checks for washouts and ice scratch lines, and historical news releases. The analysis reported herein reflects the stillwater elevation but does not include the contributions from wave action effects such as wave-crest height and wave runup. However, the additional hazard due to wave action effects should be considered in planning of future development. Elevations of the selected recurrence intervals for Great Pond and Long Pond in Belgrade and Rome, Messalonskee Lake in Belgrade, Oakland, and Sidney, and Salmon Lake in Belgrade and Oakland were based on a log-pearson Type III distribution of annual peak elevation data (Reference 18). The principal source of data for the water bodies was records of elevations maintained by Central Maine Power Company for the period from 1932 t o Flood elevations for Belgrade Stream above Wings Mills Dam are the same as those for Long Pond. Flood 47

53 elevations for Belgrade Stream below Wings Mills Dam are the same as those for Messalonskee Lake. Levels were run at selected locations along the stream to verify this fact. The normal pond elevations for Great Pond, Long Pond, Messalonskee Lake, and Salmon Lake are, respectively, (247.1 feet NAVD), (237.5 feet NAVD), (234.8 feet NAVD), and (277.4 feet NAVD). Cross section extensions and basin characteristics for the Eastern River in Pittston were based on information contained on topographic maps (Reference 46). Watersurface elevations for the Eastern River were computed using the WSPRO stepbackwater computer model (References 28 and 53). The starting water-surface elevation for the Eastern River was determined to be the backwater elevation for the 1-percent-annual-chance flood from the Kennebec River. Water-surface elevations of floods of the selected recurrence interval in Pittston were computed using the USGS E431 step-backwater computer program (Reference 53). The starting water-surface elevation was taken from the precountywide FIS for Dresden, Maine, which was performed in conjunction with the Pittston precountywide FIS (Reference 34). Cross sections for the Kennebec River and Messalonskee Stream for Waterville were obtained from the earlier FISs for the City of Waterville (Reference 54) and Town of Sidney (Reference 55), and field surveys during the 1997 field season. Water-surface elevations of floods of the selected recurrence intervals were computed using RiverCAD, a step-backwater computer program (References 42, 56, and 57). Starting water-surface elevations for the Kennebec River was taken from the precountywide Town of Sidney FIS (Reference 55). Starting watersurface elevations for Messalonskee Stream were taken at its confluence with the Kennebec River. Flood profiles were drawn showing computed water-surface elevations for floods of the selected recurrence intervals. The 1-percent-annual-chance flood elevations on Pocasset and Androscoggin Lakes in Wayne were determined by analyzing historical records at the lakes and for the Androscoggin River. During flood peaks, the Androscoggin River flows up the Dead River and into Androscoggin Lake. Flood elevations in the lake are determined by the flood elevation in the Androscoggin River at the mouth of the Dead River and the natural constrictions and bridges along the Dead River that restrict upstream movement of floodwaters and result in head losses in the upstream direction. From historical flood marks (taken from the precountywide FIS for the Town of Livermore) of the 1936 and 1987 floods, 200-year (0.5- percent-annual-chance) and 75-year (approximately 1.3-percent-annual-chance) floods, respectively, on the Androscoggin River near Wayne, direct interpolation was made to determine the 1-percent-annual-chance flood elevation of Pocasset and Androscoggin Lakes (Reference 13). In Monmouth and Winthrop, the approximate 1-percent-annual-chance flood boundaries were determined by a method developed by the USGS office in Augusta, Maine, in A regional analysis was made between the stages of the 48

54 1-percent-annual-chance flood and 50 percent duration discharge, at sites where the stream flow data were available. The results show that, generally, the 1- percent-annual-chance flood elevations are 10 feet higher than the mapped stream elevations (Reference 47). For the streams studied by approximate methods in the Cities of Augusta and Hallowell, and the Towns of Manchester, Monmouth, Pittston, Randolph, West Gardiner, and Winthrop, the 1-percent-annual-chance flood elevations were plotted using the following method which was developed by USGS hydrologists at the Augusta, Maine office. They have determined a regional stage-frequency relationship and estimate a 10-foot rise over the mapped stream elevation as the inundation limit of the 1-percent-annual-chance flood (Reference 51). In Manchester, USGS topographic maps at a scale of 1:24,000 with a contour interval of 10 feet were used to determine this elevation (Reference 45). Countywide Analyses No new hydraulic analyses were performed in Kennebec County. The results of the water-surface computations for Meadow Brook, Sebasticook River, and West Branch Sheepscot River are tabulated for selected cross sections in the table below, Table 8, 1-Percent-Annual-Chance Flood Data. 49

55 CROSS SECTION FLOODING SOURCE DISTANCE 1 WIDTH (FEET) SECTION AREA (SQARE FEET) RIVER CHANNEL MEAN VELOCITY (FEET PER SECOND) STREAM-BED ELEVATION (FEET NAVD) BASE FLOOD ELEVATION (FEET NAVD) A B C D E F G FEET FROM LIMIT OF STUDY NOTE: LIMIT OF STUDY IS LOCATED APPROXIMATELY 0.8 MILES DOWNSTREAM OF DIRIGO ROAD TABLE 8 FEDERAL EMERGENCY MANAGEMENT AGENCY KENNEBEC COUNTY, ME (ALL JURISDICTIONS) 1% ANNUAL CHANCE FLOOD DATA MEADOW BROOK

56 CROSS SECTION FLOODING SOURCE DISTANCE 1 WIDTH (FEET) SECTION AREA (SQARE FEET) RIVER CHANNEL MEAN VELOCITY (FEET PER SECOND) STREAM-BED ELEVATION (FEET NAVD) BASE FLOOD ELEVATION (FEET NAVD) X Y Z AA AB AC AD AE AF AG AH AI AJ AK AL AM AN AO AP AQ AR AS AT AU AV AW AX FEET ABOVE CONFLUENCE WITH KENNEBEC RIVER NOTE: Cross Sections A through W can be found in Table 10, Floodway Data TABLE 8 FEDERAL EMERGENCY MANAGEMENT AGENCY KENNEBEC COUNTY, ME (ALL JURISDICTIONS) 1% ANNUAL CHANCE FLOOD DATA SEBASTICOOK RIVER

57 CROSS SECTION FLOODING SOURCE DISTANCE 1 WIDTH (FEET) SECTION AREA (SQARE FEET) RIVER CHANNEL MEAN VELOCITY (FEET PER SECOND) STREAM-BED ELEVATION (FEET NAVD) BASE FLOOD ELEVATION (FEET NAVD) AY AZ BA BB BC BD BE FEET ABOVE CONFLUENCE WITH KENNEBEC RIVER TABLE 8 FEDERAL EMERGENCY MANAGEMENT AGENCY KENNEBEC COUNTY, ME (ALL JURISDICTIONS) 1% ANNUAL CHANCE FLOOD DATA SEBASTICOOK RIVER

58 CROSS SECTION FLOODING SOURCE DISTANCE 1 WIDTH (FEET) SECTION AREA (SQARE FEET) RIVER CHANNEL MEAN VELOCITY (FEET PER SECOND) STREAM-BED ELEVATION (FEET NAVD) BASE FLOOD ELEVATION (FEET NAVD) A B C D E F G FEET ABOVE THE TOWN OF WINDSOR / TOWN OF CHINA MUNICIPAL BOUNDARY TABLE 8 FEDERAL EMERGENCY MANAGEMENT AGENCY KENNEBEC COUNTY, ME (ALL JURISDICTIONS) 1% ANNUAL CHANCE FLOOD DATA WEST BRANCH OF SHEEPSCOT RIVER

59 Roughness factors (Manning s n values) used in the hydraulic computations are shown in Table 9. Table 9, Manning s n values, shows the channel and overbank n values for the streams studied by detailed methods: TABLE 9 MANNING S n VALUES Flooding Source Channel "N" Overbanks Bond Brook (Augusta) Cobbosseecontee Stream (Gardiner) Cobbosseecontee Stream (Litchfield, West Gardiner) Cobbosseecontee Stream (Manchester) Eastern River (Pittston) Kennebec River (Augusta, Chelsea, Farmingdale, Gardiner, Hallowell) Kennebec River (Benton) Kennebec River (Pittston) Kennebec River (Randolph) Kennebec River (Sidney) Kennebec River (Waterville) Kennebec River (Winslow) Meadow Brook (China) Messalonskee Stream (Waterville) Sebasticook River (Benton) Sebasticook River (Clinton) Sebasticook River (Winslow) Togus Stream (Chelsea) Togus Stream (Pittston) West Branch Sheepscot River (China) Vertical Datum All FIS reports and FIRMs are referenced to a specific vertical datum. The vertical datum provides a starting point against which flood, ground, and structure elevations can be referenced and compared. Until recently, the standard vertical datum used for newly created or revised FIS reports and FIRMs was the National Geodetic Vertical Datum of 1929 (NGVD). With the completion of the North American Vertical Datum of 1988 (NAVD), many FIS reports and FIRMs are now prepared using NAVD as the referenced vertical datum. 54

60 Flood elevations shown in this FIS report and on the FIRM are referenced to the NAVD 88. These flood elevations must be compared to structure and ground elevations referenced to the same vertical datum. This can be done by applying a standard conversion factor. The Flood Profiles, and Base (1-percent-annualchance) Flood Elevations (BFEs) in the precountywide FIS reports, are in NGVD. These were converted to NAVD by applying the conversion factor of -0.6 feet to each detailed study stream in the effective FIS reports (NGVD 0.6 ft. = NAVD). It is important to note that adjacent communities may be referenced to NGVD 29. This may result in differences in base flood elevations across the corporate limits between the communities. For information regarding conversion between the NGVD 29 and NAVD 88, visit the National Geodetic Survey website at or contact the National Geodetic Survey at the following address: NGS Information Services NOAA, N/NGS12 National Geodetic Survey SSMC-3, # East-West Highway Silver Spring, Maryland (301) Temporary vertical monuments are often established during the preparation of a flood hazard analysis for the purpose of establishing local vertical control. Although these monuments are not shown on the FIRM, they may be found in the Technical Support Data Notebook associated with the FIS report and FIRM for this community. Interested individuals may contact FEMA to access these data. The BFEs shown on the FIRM represent whole-foot rounded values. For example, a BFE of will appear as 102 on the FIRM and will appear as 103. Therefore, users that wish to convert the elevations in this FIS to NGVD 29 should apply the stated conversion factor to elevations shown on the Flood Profiles and supporting data tables in the FIS report, which are shown at a minimum to the nearest 0.1 foot. To obtain current elevation, description, and/or location information for benchmarks shown on this map, please contact the Information Services Branch of the NGS at (301) , or visit their website at FLOODPLAIN MANAGEMENT APPLICATIONS The NFIP encourages State and local governments to adopt sound floodplain management programs. To assist in this endeavor, each FIS report provides 1-percent-annual-chance floodplain data, which may include a combination of the following: 10-, 2-, 1-, and 0.2-percent-annual-chance flood elevations; delineations of the 1- and 0.2-percent-annual-chance floodplains; and a 1-percent-annual-chance floodway. This information is presented on the FIRM and in many components of the FIS report, including Flood Profiles, Floodway Data tables, and Summary of Stillwater Elevation 55

61 tables. Users should reference the data presented in the FIS report as well as additional information that may be available at the local community map repository before making flood elevation and/or floodplain boundary determinations. 4.1 Floodplain Boundaries In order to provide a national standard without regional discrimination, the 1-percent-annual-chance flood has been adopted by FEMA as the base flood for floodplain management purposes. The 0.2-percent-annual-chance flood is employed to indicate additional areas of flood risk in the community. For unrevised streams in Kennebec County, data was taken from previously printed FISs for each individual community and are compiled below. For each stream studied by detailed methods, the 1- and 0.2-percent-annual-chance floodplain boundaries have been delineated using the flood elevations determined at each cross section. Between cross sections, the boundaries were interpolated using topographic maps at a scale of 1:62,500, with a contour interval of 20 feet (Reference 62), at a scale of 1:24,000, with a contour interval of 20 feet (Reference 43), at a scale of 1:24,00 with a contour interval of 10 feet (Reference 45), at a scale of 1:4,800, with a contour interval of 4 feet (References 61 and 63), and at a scale of 1:2,400, with a contour interval of 5 feet (Reference 60). For the Towns of Farmingdale, Manchester, Monmouth, Readfield, West Gardiner, and Winthrop, the boundaries were interpolated between cross sections using photogrammetric maps at scales of 1:4,800 and 1:6,600 (References 39, 40, 52, and 64). For the City of Augusta, the boundaries of Bond Brook were interpolated between cross sections using photogrammetric maps at a scale of 1:4,800 (Reference 40). For Cobbosseecontee Stream in the City of Gardiner, the boundaries between cross sections were interpolated photogrammetrically using aerial photographs at a scale of 1:6,600 (Reference 39). For the Kennebec River in the Cities of Gardiner and Hallowell, and the Town of Randolph, the boundaries were interpolated between cross sections using photogrammetric maps at a scale of 1:4,800 (Reference 40). For Togus Stream in the Town of Randolph, the boundaries were interpolated between cross sections, using photogrammetric maps at a scale of 1:6,600 (Reference 39). In the Town of Pittston, in addition to using topographic maps, the boundaries were interpolated between cross sections using photogrammetric maps at scales of 1:4,800 and 1:6,600 (References 40 and 52). 56

62 For flooding sources studied by approximate methods, the boundaries of the 1- percent-annual-chance floodplain were delineated using the Flood Hazard Boundary Maps for the Town of Belgrade (Reference 65), Town of Benton (Reference 66), Town of China (Reference 67), Town of Litchfield (Reference 68), Town of Oakland (Reference 69), Town of Rome (Reference 70), the Town of Wayne (Reference 71), and the City of Waterville (Reference 72). For the streams studied by approximate methods in the Cities of Augusta and Gardiner, and the Towns of Farmingdale, Monmouth, Pittston, Randolph, West Gardiner, and Winthrop the 1-percent-annual-chance floodplain boundaries were delineated using USGS topographic quadrangle maps enlarged to a scale of 1:12,000 (References 45 and 73). For the City of Gardiner and the Towns of Monmouth, West Gardiner, and Winthrop, the boundaries were then transferred to town maps at a scale of 1:12,000 (References 74, 75, 76, and 77). For the streams studied by approximate methods in the Towns of Chelsea and Readfield, 1-percent-annual-chance floodplain boundaries were determined by a procedure developed in the Maine Office of the USGS in A regional analysis was made between stages of 1-percent-annual-chance and 50 percent duration flows at sites where stream flow data were available. The results of this analysis show that the 1-percent-annual-chance flood elevations are generally 10 feet higher than the stream elevations determined from USGS topographic maps (Reference 51). The 1-percent-annual-chance flood boundaries in Chelsea were then delineated on the USGS topographic maps (Reference 62). 1-percent-annualchance flood boundaries in Readfield and sections of Stony Meadow Brook in Chelsea were delineated on topographic maps at a scale of 1:24,000, with a contour interval of 10 feet (Reference 45). The approximate 1-percent-annualchance floodplain boundaries in Chelsea were drawn on a town base map at a scale of 1:12,000 (Reference 78). In the City of Hallowell, the approximate 1-percent-annual-chance floodplain boundaries for Vaughan Brook were delineated on USGS topographic maps at a scale of 1:24,000, with a contour interval of 10 feet, and were compared to the Flood Hazard Boundary Map for Hallowell, Maine (References 45 and 79). The approximate 1-percent-annual-chance floodplain flood boundaries for the Town of Manchester were delineated on 7.5-Minute USGS topographic maps at a scale of 1:24,000 with a contour interval of 10 feet (Reference 45). For the flooding sources studied by approximate methods in Sidney, the 1- percent-annual-chance floodplain boundaries were taken from the March 18, 1987, Town of Sidney FIS (Reference 80). For the streams studied by approximate methods in West Gardiner, the 1-percentannual-chance floodplain boundaries were delineated using the FIRM for the Town of Winslow (Reference 81). 57

63 There were no streams revised for this countywide FIS. The 1- and 0.2-percent-annual-chance floodplain boundaries are shown on the FIRM. On this map, the 1-percent-annual-chance floodplain boundary corresponds to the boundary of the areas of special flood hazards (Zones A and AE), and the 0.2-percent-annual-chance floodplain boundary corresponds to the boundary of areas of moderate flood hazards. In cases where the 1- and 0.2-percent-annual-chance floodplain boundaries are close together, only the 1-percent-annual-chance floodplain boundary has been shown. Small areas within the floodplain boundaries may lie above the flood elevations, but cannot be shown due to limitations of the map scale and/or lack of detailed topographic data. For streams studied by approximate methods, only the 1-percent-annual-chance floodplain boundary is shown on the FIRM. 4.2 Floodways Encroachment on floodplains, such as structures and fill, reduces flood-carrying capacity, increases flood heights and velocities, and increases flood hazards in areas beyond the encroachment itself. One aspect of floodplain management involves balancing the economic gain from floodplain development against the resulting increase in flood hazard. For purposes of the NFIP, a floodway is used as a tool to assist local communities in this aspect of floodplain management. Under this concept, the area of the 1-percent-annual-chance floodplain is divided into a floodway and a floodway fringe. The floodway is the channel of a stream, plus any adjacent floodplain areas, that must be kept free of encroachment so that the base flood can be carried without substantial increases in flood heights. Minimum Federal standards limit such increases to 1 foot, provided that hazardous velocities are not produced. The floodways in this study are presented to local agencies as minimum standards that can be adopted directly or that can be used as a basis for additional floodway studies. Floodway widths were computed at cross sections. Between cross sections, the floodway boundaries were interpolated. The results of the floodway computations are tabulated for selected cross sections (see Table 10, Floodway Data ). In cases where the floodway and 1-percent-annual-chance floodplain boundaries are either close together or collinear, only the floodway boundary is shown. The concept of a floodway does not apply in areas of lacustrine flooding. Therefore, no floodway was computed or delineated within the Towns of Belgrade, Oakland, and Rome. The floodway concept is generally not applicable for bodies of water with 58

64 significant impoundment effects. Because Cobbosseecontee Lake does provide storage, no floodway was computed along its boundaries in West Gardiner. No floodways were calculated for the Sebasticook River in the Town of Clinton, the Eastern River in the Town of Pittston, or the Towns of China, Monmouth, Readfield, Sidney, and Wayne. Portions of the computed floodway widths extend beyond the corporate limits for the Kennebec River in Benton, Chelsea, Clinton, Farmingdale, Gardiner, Hallowell, Pittston, Randolph, Waterville, and Winslow; the Cobbosseecontee Stream in Gardiner, Litchfield, Manchester, and West Gardiner; and the Togus Stream in Pittston and Randolph. Encroachment into areas subject to inundation by floodwaters having hazardous velocities aggravates the risk of flood damage, and heightens potential flood hazards by further increasing velocities. A listing of stream velocities at selected cross sections is provided in Table 10, Floodway Data. To reduce the risk of property damage in areas where the stream velocities are high, the community may wish to restrict development in areas outside the floodway. Near the mouths of streams studied in detail, floodway computations are made without regard to flood elevations on the receiving water body. Therefore, Without Floodway elevations presented in Table 10, for certain downstream cross sections of Bond Brook in Augusta, the Sebasticook River in Benton and Winslow, the Cobbosseecontee Stream in Gardiner, the Togus Stream in Pittston and Randolph, and Messalonskee Stream in Waterville, are lower than the regulatory flood elevations in those areas, which must take into account the 1- percent-annual-chance flooding due to backwater from other sources. The area between the floodway and 1-percent-annual-chance floodplain boundaries is termed the floodway fringe. The floodway fringe encompasses the portion of the floodplain that could be completely obstructed without increasing the water-surface elevation (WSEL) of the base flood more than 1 foot at any point. Typical relationships between the floodway and the floodway fringe and their significance to floodplain development are shown in Figure 1, Floodway Schematic. Eastern River in Pitttson, cross section are shown on the profile and effective firm, a floodway data table was not located in the previous FIS report. For this countywide FIS, no new floodways were computed. Under the State of Maine Revised Statutes Annotated (M.R.S.A.) Title A, 7C where the floodway is not designated on the Flood Insurance Rate Map, the floodway is considered to be the channel of a river or other water course and the adjacent land areas to a distance of one-half the width of the floodplain, as measured from the normal high water mark to the upland limit of the floodplain, unless a technical evaluation certified by a registered professional engineer is provided demonstrating the actual floodway based upon approved FEMA modeling methods. 59

65 1-PERCENT-ANNUAL-CHANCE FLOODPLAIN FIGURE 1 SCHEMATIC 60