National Assessment Of Shoreline Change Part 3: Historical Shoreline Change and Associated Coastal Land Loss Along Sandy Shorelines of the California Coast Cheryl J. Hapke(1), David Reid(2), Bruce M. Richmond(2), Peter Ruggiero(3), and Jeff List(4) 2006 U.S. Geological Survey Open-file Report 2006-1219. U.S. Department of the Interior U.S. Geological Survey (1) U.S. Geological Survey, Coastal Field Station, Department of Geosciences, University of Rhode Island, Kingston, RI 02881 (2) U.S. Geological Survey, Pacific Science Center, Santa Cruz, CA 95060 (3) Dept. of Geosciences, Oregon State University, Corvallis, OR 97331 (4) U.S. Geological Survey, Woods Hole Science Center, Woods Hole, MA 02543 Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Contents EXECUTIVE SUMMARY....1 INTRODUCTION....1 U.S. Geological Survey National Assessment of Shoreline Change....1 Disclaimer...2 Acknowledgments....2 THE ROLE OF STATE AND FEDERAL GOVERNMENTS....4 PRIOR NATIONAL AND CALIFORNIA COAST SHORELINE ASSESSMENTS....4 METHODS OF ANALYZING SHORELINE CHANGE....5 Compilation of Historical Shorelines....5 Delineation of a Modern (Lidar-derived) Shoreline....5 Geographic Information System (GIS) Procedures....6 Calculation and Presentation of Rates of Change....7 Beach Alterations and Shoreline Definitions that Influence Rates of Change....8 Human Activities....8 Shoreline Definitions....9 Estimating the proxy-datum bias and the bias uncertainty....10 Uncertainties and Errors....11 End-point (short-term) shoreline change error ....13 Linear regression (long-term) shoreline change error....14 GEOLOGIC HISTORY AND SETTING....15 General Geology of the California Coast....15 Climate ....19 Coastal Processes....20 Waves ....20 Tides ....21 Winds ....22 Currents....22 Littoral cells and transport directions....22 Sand Sources....23 GEOMORPHOLOGY OF THE CALIFORNIA COAST....24 Cliffs ....24 Beaches....24 Coastal Dune Complexes....26 Estuaries and Lagoons....26 General Characteristics of the California Coast Sections....27 Northern Section: Oregon Border to Tomales Bay....27 Central Section: Tomales Bay to Point Conception....30 Southern Section: Point Conception to the Mexican Border....30 HISTORY OF INFRASTRUCTURE DEVELOPMENT....30 HISTORICAL SHORELINE CHANGE ANALYSIS....33 Northern California....33 Klamath Region....33 Eureka Region....34 Navarro Region....36 Russian River Region....40 Central California....42 San Francisco North Region....43 San Francisco South Region....43 Monterey Bay Region....47 Big Sur Region....50 Morro Bay Region....50 Santa Barbara North....54 Southern California....56 Santa Barbara South Region....57 Santa Monica Region ....57 San Pedro Region....61 Oceanside Region....63 San Diego Region....65 DISCUSSION AND FURTHER CONSIDERATIONS....67 Summary of Shoreline Changes....67 Planned Updates and Related Research....68 Influence of Human Activities....68 REFERENCES....68 Figures Figure 1. Index map of California showing the fifteen regions discussed in the text....3 Figure 2. Example of a lidar profile from April 10, 1998 at Santa Cruz, California....7 Figure 3. Examples of common conditions where transects are eliminated in the absence of four shoreline intersections....9 Figure 4. Major tectonic elements and geomorphic provinces of California ....16 Figure 5. Index map showing locations of place names discussed in Geologic History and Geomorphology section....17 Figure 6. Simplified geologic map of California....18 Figure 7. General wave directions for central California based on offshore wave buoy data....20 Figure 8. Monthly mean sea level for selected tide gauges along the California coast....21 Figure 9. Map showing major littoral cell boundaries, coastal watersheds, and conceptual net longshore drift directions for the California coast....23 Figure 10. Average annual sand and gravel discharge from major rivers in California....25 Figure 11. Sketch of common coastal landforms of California....26 Figure 12. Beach profiles from Cowells Beach in Santa Cruz illustrating beach erosion over an El Ni_o winter and the subsequent recovery the following summer....27 Figure 13. Map showing the major coastal dune complexes of California....28 Figure 14A. Index map showing the four analysis regions in Northern California and various locations....29 Figure 14B. Index map showing the six analysis regions in Central California and various locations ....31 Figure 14C. Index map showing the five analysis regions in Southern California and various locations....32 Figure 15. Shoreline change rates for the Klamath region....35 Figure 17. Shoreline change rates for the Eureka region....37 Figure 18. Spit and jetty on north side of Humboldt Bay Harbor in photograph taken in June 1987....38 Figure 19. Shoreline change rates for the Navarro region....39 Figure 20. Photographs from 1972 (top) and 2002 (bottom) show the widening of the spit beach at the Ten Mile River mouth....40 Figure 21. Shoreline change rates for the Russian River region....41 Figure 22. Many of the large rivers in Northern California have well-developed sandy spits such as this one extending across the mouth of the Russian River....42 Figure 23. Shoreline change rates for the San Francisco North region....44 Figure 24. Houses on the spit at Stinson Beach are partially buried by sand during an El Ni_o winter storm in 1983....45 Figure 25. Shoreline change rates for the San Francisco South region....46 Figure 26. A riprap revetment protects an apartment building that overlooks a narrow beach near Pillar Point Harbor in a photograph taken in February, 2002....47 Figure 27. Shoreline change rates for the Monterey Bay region....48 Figure 28. Aerial photographs of the Santa Cruz Yacht Harbor show the impoundment of sand against the constructed jetties....49 Figure 29. Shoreline change rates for the Big Sur region....51 Figure 30. JP Burns waterfall shows the beach that formed as a result of the sediment added to the system from the 1983 landslide....52 Figure 31. Shoreline change rates for the Morro Bay region....53 Figure 32. Development of a tombolo at Morro Rock after a causeway was built in the 1930s....54 Figure 33. Shoreline change rates for the Santa Barbara North region....55 Figure 34. Photographs from January 1989 shows the well- developed dune field at Guadalupe Dunes....56 Figure 35. Shoreline change rates for the Santa Barbara South region....58 Figure 36. Dredging of spit at Santa Barbara Harbor....59 Figure 37. Shoreline change rates for the Santa Monica region....60 Figure 38. Oblique aerial photograph of a breakwater along the highly engineered Santa Monica coastline....61 Figure 39. Shoreline change rates for the San Pedro region....62 Figure 40. Oblique aerial photograph of the Los Angeles Harbor....63 Figure 41. Shoreline change rates for the Oceanside region....64 Figure 42. Narrow beach fronting tall coastal bluffs at Torrey Pines City Beach....65 Figure 43. Shoreline change rates for the San Diego region....66 Figure 44. Oblique aerial photograph of Zuniga Point and breakwater at the entrance to the San Diego Harbor....67 Tables Table 1. Providers and original sources of historical shorelines for each California region....5 Table 2. Dates of compiled shorelines in different regions for selected periods....6 Table 3. List of tide gauge measurements used to calculate mean high water elevation....8 Table 4. Absolute horizontal and vertical differences between high water and mean high water shorelines....10 Table 5. Maximum estimated measurement errors for California shorelines....11 Table 6A. Average shoreline change rates for Northern California....12 Table 6B. Average shoreline change rates for Central California....12 Table 6C. Average shoreline change rates for Southern California....12 Table 7. Coastal cliff rock type along the California coast (from Runyan and Griggs, 2002)....19 Table 8. Estimated annual littoral drift rates and directions along the California Coast....24 Table 9A. Maximum and minimum shoreline change rates: Northern California....34 Table 9B. Maximum and minimum shoreline change rates: Central California....45 Table 9C. Maximum and minimum shoreline change rates: Southern California....59 EXECUTIVE SUMMARY Beach erosion is a chronic problem along many open-ocean shores of the United States. As coastal populations con_tinue to grow and community infrastructures are threatened by erosion, there is increased demand for accurate informa_tion regarding past and present trends and rates of shore_line movement. There is also a need for a comprehensive analysis of shoreline movement that is consistent from one coastal region to another. To meet these national needs, the U.S. Geological Survey is conducting an analysis of histori_cal shoreline changes along open-ocean sandy shores of the conterminous United States and parts of Hawaii and Alaska. One purpose of this work is to develop standard repeatable methods for mapping and analyzing shoreline movement so that periodic updates regarding coastal erosion and land loss can be made nationally that are systematic and internally con_sistent. In the case of this study, the shoreline being measured is the boundary between the ocean water surface and the sandy beach. This report on the California Coast represents the first of two reports on long-term sandy shoreline change for the western U.S., the second of which will include the coast of the Pacific NW, including Oregon and Washington. A report for the Gulf of Mexico shoreline was completed in 2004 and is available at: http://pubs.usgs.gov/of/2004/1043/. This report summarizes the methods of analysis, interprets the results, provides explanations regarding long-term and short-term trends and rates of change, and describes how different coastal communities are responding to coastal erosion. Shore_line change evaluations are based on comparing three histori_cal shorelines digitized from maps, with a recent shoreline derived from lidar (Light Detection and Ranging) topographic surveys. The historical shorelines generally represent the following periods: 1800s, 1920s- 1930s, and 1950s-1970s, whereas the lidar shoreline is from 1998-2002. Long-term rates of change are calculated using all four shorelines (1800s to lidar shoreline), whereas short-term rates of change are calculated for only the most recent period (1950s-1970s to lidar shoreline). The rates of change presented in this report represent past conditions and therefore are not intended for predicting future shoreline positions or rates of change. Due to the geomorphology of the California Coast (rocky coastline instead of beach) as well as to data gaps in some areas, this report presents beach erosion rates for 45% of California's 1100 km of coast. The average rate of long-term shoreline change for the State of California was 0.2_0.1 m/yr, an accretional trend. This is based on shoreline change rates averaged from 14,562 individual transects, of which 40% were eroding. Of the transects on which the shoreline was eroding, the long- term erosion rates were generally lowest in Southern California where coastal engineering projects have greatly altered the natural shoreline movement. On a regional scale, long-term accretion rates were either equal to (Central California) or greater than (Northern and Southern California) the long-term erosion rates, yielding the net accretional trend for the entire state. This accretional trend is most likely due to changes in the large volumes of sediment that are added to the system from large rivers and to the impact from coastal engineering and beach nourishment projects. The average rate of short-term shoreline change for the state was erosional. The net short-term rate as averaged along 16,142 transects was -0.2_0.4 m/yr. Of the transects used to measure short-term change, 66% had erosional trends. In addition erosion rates were higher in the short-term period, possibly related to the localized artificial nourishment that occurred over much of the 20th century but that has recently slowed or stopped (Flick, 1993; Wiegel, 1994). Short-term accretion rates were highest in Northern California where the overall magnitudes of shoreline change are systematically higher than in Central and Southern California. The most stable (low erosion and accretion rates) California beaches were most commonly found in Central California. Seawalls and/or riprap revetments have been constructed in all three sections of California, although many of these structures were built to protect houses and infrastructures from the erosion of coastal cliffs and bluffs rather than to protect against long-term beach erosion. California permits shoreline stabilization structures where homes, buildings or other community infrastructure are imminently threatened by erosion. A second California report that is following this publica_tion will include analyses and reports on long-term coastal cliff erosion, as this hazard is of equal or greater concern to coastal communities in many areas along the California Coast. INTRODUCTION U.S. Geological Survey National Assessment of Shoreline Change Sandy ocean beaches represent some of the most popular tourist and recreational destinations in the United States, and also constitute some of the most valuable real estate in the country. These changing and ephemeral interfaces between water and land are the sites of intense residential and commercial development even though they are frequently subjected to a range of natural hazards that can include flood_ing, storm impacts, coastal erosion and tsunami inundation. Because population and economic trends have made the coasts so valuable, the U.S. Geological Survey (USGS) is conducting a National Assessment of Coastal Change Haz_ards. One component of this effort, the National Assessment of Shoreline Change, documents changes in shoreline posi_tion as a proxy for coastal change. Shoreline position is one of the most commonly monitored indicators of environmen_tal change (Morton, 1996), and it is easily understood by those who are interested in historical movement of beaches. A principal purpose of the USGS shoreline change research is to develop a repeatable surveying methodology so that shorelines for the continental U.S., and portions of Hawaii and Alaska, can be periodically and systematically updated in an internally consistent manner. In addition, new methods for developing datum-based shorelines and assess_ing coastal change can provide the opportunity to achieve more comprehensive assessments of error in the future. The primary objectives of this effort are: (1) to develop and implement improved methods of assessing and monitoring shoreline movement, (2) to obtain a better understanding of the processes controlling shoreline movement, and (3) to enter into partnerships to facilitate data dissemination. Achieving these ongoing objectives requires research that (1) examines the original sources of shoreline data (maps, air photos, global positioning system (GPS), lidar), (2) evaluates the utility of different shoreline proxies (geomorphic feature, water mark, tidal datum, elevation contour) including the errors associated with each method, (3) investigates the bias and potential errors associated with integrating different shoreline proxies from different sources, (4) develops standard uniform methods of shoreline change analysis, (5) determines the effects of human activi_ties on shoreline movement and rates of change, and (6) investigates alternative mathematical methods for calculat_ing historical rates of change and forecasting future rates of change. This report summarizes historical changes in the California sandy shoreline, both accretion and erosion, but emphasizes the erosion hazard because of its impacts on natural resources and the economy. The descriptions of coastal land loss for each region (Figure 1) within the state provide a more comprehensive view of coastal processes and key references that can be used to learn more about coastal change in a regional context. Disclaimer Results of the National Assessment of Shoreline Change are organized by coastal regions. This report for California is part of a series of reports that will include text summarizing methods, results, and implications of the results in addition to maps, via Internet Map Server (IMS), illustrating rates of shoreline change. Rates of shoreline change are being published for the purpose of regional characterization. The shoreline change results and products prepared by the USGS are not intended for comprehensive detailed site specific analysis of shoreline movement, nor are they intended to replace any official sources of shoreline change information identified by local or state government agencies, or other federal entities that are used for regulatory purposes. Rates of shoreline change presented herein may differ from other published rates, and differences do not necessarily indicate that the other rates are inaccurate. Some discrepancies are expected, considering the many possible ways of determining shoreline positions and rates of change, and the inherent uncertainty in calculating these rates. Rates of shoreline change presented in this report represent shore_line movement under past conditions. The results are not intended for predicting future shoreline positions or future rates of shoreline change. This publication was prepared by an agency of the United States Government. Neither the United States Gov_ernment nor any agency thereof, nor any of their employ_ees, makes any warranty, expressed or implied, or assumes any legal responsibility for the accuracy, completeness, or usefulness of any information apparatus, product, or process disclosed in the report, or represents that its use would not infringe privately owned rights. Reference to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. Any views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. Acknowledgments This report was made possible by the hard work and generous cooperation of many individuals. We owe a debt of gratitude to Mike Rink (NOAA) and Paul Frascione (Information Manufacturing Corporation) for providing digital historical shorelines and scans of selected T-sheets, and David Doyle (NOAA) for providing datum corrections so that T-sheets could be rectified before they were digitized by the USGS. An enthusiastic and untiring USGS team was responsible for developing the methods and computer codes for calculating operational mean high water elevations and extracting the lidar shorelines. The lidar research and development team included Abby Sallenger, Jeff List, Karen Morgan, Eric Nelson, Hillary Stockdon, and Kathy Webber. Tara Miller and Kathy Webber provided invaluable help in data processing and analysis. Rob Thieler worked closely with TPMC Environmental Services to develop and improve the Digital Shoreline Analysis System (DSAS) code for shoreline change measurement. Brian Spear (UCSC) and Evan Lindenbach (UCSC) provided assistance by digitizing shorelines and cliff edges, georectifiying maps and prepar_ing metadata. Bruce Rogers provided valuable assistance in the drafting of figures, and Ben Melosh (UCSC) helped with background data compilation. Ann Gibbs provided excellent advice and input regarding data presentation. Comments from and discussions with Ed Thornton, Bob Guza, Dick Seymour, Bob Wiegel, Mark Crowell, Lesley Ewing, Bob Morton and Sam Johnson were also very helpful. Finally, a detailed review by Gary Griggs was invaluable and greatly improved the report. THE ROLE OF STATE AND FEDERAL GOVERNMENTS One reason for conducting this National Assessment of Shoreline Change is that there is no widely accepted standardized method of analyzing shoreline changes. Each state or region has its own data needs and coastal zone man_agement responsibilities (e.g. construction set-back lines), and therefore different techniques and standards are used to compile shorelines and to calculate rates of shoreline move_ment. Consequently, existing calculated rates of shoreline change and projected shoreline positions are inconsistent from state to state and even within states, such as in Califor_nia, and cannot be compared directly. These inconsisten_cies were clearly demonstrated by the Federal Emergency Management Agency (FEMA) sponsored erosion studies (Crowell and Leatherman, 1999) that were used as the basis for evaluating erosion hazards (The Heinz Center, 2000). Within California, the FEMA sponsored erosion studies only addressed coastal cliff erosion, not sandy shoreline ero_sion. The USGS National Assessment of Shoreline Change represents the first time that shorelines from original data sources have been compiled and rates of shoreline change have been calculated on a national scale using internally consistent methods. The results of this analysis allow direct comparison of rates of change from one coastal segment to another and form the basis for future comparison of shore_line position. Several federal agencies (USGS, FEMA, NOAA, U.S. Army Corps of Engineers (USACE) have regulatory or administrative responsibilities pertaining to shorelines. Yet these responsibilities are quite different, requiring differ_ent approaches and offering substantial opportunities for cooperation. For example, the USACE is authorized and funded by Congress to report on the economic and envi_ronmental implications of shoreline change and the costs of erosion mitigation. Their National Shoreline Management Study (Stauble and Brumbaugh, 2003) is being conducted using existing shoreline data. The USGS will share data and information, such as the lidar-derived shoreline and rates of change, in support of their effort. NOAA has the mandate to establish the official shoreline boundary for the nation using tidal datums. Their emphasis is on safe navigation and using the shoreline to generate nautical charts. NOAA also has a developing program (V datum), which will greatly assist other agencies in establishing alternative shorelines for a variety of purposes where the official shoreline is inappropriate. FEMA is authorized and partially funded by Congress to map coastal (and riverine) flood hazard areas. These maps and associated information are used for flood risk assessment, floodplain management, and setting insur_ance rates through the National Flood Insurance Program (NFIP). Because of perceived deficiencies in the way the NFIP considers coastal erosion, Congress authorized FEMA to report on the economic impact of erosion hazards on coastal communities, and on the NFIP. To accomplish this, FEMA contracted state agencies and academic researchers to conduct a pilot study of erosion hazards that included shoreline change data for limited geographic areas. The USGS is responsible for conducting research pertain_ing to coastal change hazards including shoreline change, understanding the processes that cause coastal change, and developing models to predict future change. The USGS is the only government agency that has a dedicated program to monitor coastal change into the future using consistent methods nationwide. Such a program is critically important to assess national issues, such as the coastal impacts of sea level rise. PRIOR NATIONAL AND CALIFORNIA COAST SHORELINE ASSESSMENTS There are very few studies of regional sandy shoreline erosion for California. The USACE (1971) conducted the first national assessment of coastal erosion that included California. That study identified areas of critical and non-critical erosion on the basis of economic development and potential for property loss, but rates of shoreline movement were not quantified. Numerous analyses have been con_ducted for specific sites by private consultants or contrac_tors, or cities and counties where erosion rates have been required for regulatory or management purposes. Some of these analyses were incorporated into Dolan and others (1985), and Griggs and Savoy (1985), where rates of change were presented on maps, and the long-term trends of erosion and accretion were summarized in an accompanying text. The Griggs and Savoy (1985) compilation has recently been updated (Griggs and others, 2005a), although most of the erosion hazards addressed therein pertain to coastal cliff and bluff erosion, with the exception of Southern California. Since these earlier works, methods of obtaining, analyzing, displaying, and storing shoreline data have improved substantially, and coastal change has continued. Furthermore, coastal scientists have not agreed on standard methods for analyzing and reporting shoreline changes, nor have they identified rigorous mathematical tests that are widely accepted for quantifying the change and associ_ated errors. Consequently, there are critical needs for (1) a nationwide compilation of reliable shoreline data including the most recent shoreline position, and (2) a standardization of methods for obtaining and comparing shoreline positions and mathematically analyzing the trends. METHODS OF ANALYZING SHORELINE CHANGE Compilation of Historical Shorelines Coastal scientists in U.S. universities and government agencies have been quantifying rates of shoreline movement and studying coastal change for decades. Before GPS and lidar technologies were developed, the most commonly used sources of historical shoreline position were NOAA Topo_graphic Sheets (T-sheets, see Shalowitz, 1964) and aerial photographs. Ideally, extraction of shoreline position from these data sources involves geo-referencing and removing distortions from maps or aerial photographs, followed by digitizing shoreline position. Depending on coastal loca_tion, data source, and scientific preference, different proxies for shoreline position are used to document coastal change, including the high water line (for discussion of the high water line (HWL) see Shalowitz, 1964), wet-dry line, veg_etation line, dune toe or crest, toe or berm of the beach, cliff base or top, and the line of mean high water (MHW). The USGS National Assessment of Shoreline Change analysis for California incorporates shoreline positions from 4 time periods and three unique data sources. To main_tain consistency at a national scale, these four periods are mid- to late1800s, 1920s-1930s, 1950s-1970s, and post-1997. Several organizations have provided the USGS with digital maps and/or shoreline data (Table 1). The historical shorelines from the 1850-1890s and 1920s-1930s were digi_tized from scanned and georeferenced historical T- sheets. In addition, shorelines were digitized from USGS Digital Raster Graphic (DRG) maps where data gaps in the T- sheets existed, as these were the only source of shoreline data that could be located for the data gap areas. These occurred for the third period (1950s-1970s) for most of Central and Northern California. The modern (post-1997) shoreline represents a MHW elevation derived from lidar data. Shorelines were compiled for the state following the guidelines established for selected periods (mid- to late 1800s, 1920s-1930s, 1950s-1970s, and post-1997) as closely as possible. Table 2 lists the final range of years (and months where known) for shorelines compiled for each period by region. Delineation of a Modern (Lidar-derived) Shoreline The most recent shoreline used in this National Assess_ment (post-1997) was derived from lidar (Light Detec_tion and Ranging) data. The USGS, in collaboration with NASA, has been using the NASA Airborne Topographic Mapper (ATM) to map coastal areas since 1997 (Krabill and others, 2000; Sallenger and others, 2003). The ATM surveys ground elevation using an elliptically rotating blue-green laser. GPS (global positioning system) positions and inertial navigation systems are used to correct for aircraft pitch, roll, and heading, providing ground elevations with accuracies of about "15 cm (Sallenger and others, 2003). The lidar surveys used to extract shorelines for this report were conducted either in 1998 or 2002 (Table 2). To compare with historical shorelines, an operational MHW shoreline was extracted from the lidar surveys using a method developed by Stockdon and others (2002) (Fig_ure 2). Shorelines were extracted from cross-shore pro_files which consist of bands of lidar data 10 m wide in the alongshore direction and spaced every 20 m along the coast. A least- squares linear regression line is passed through the 2-D cluster of data that encompasses the operational MHW datum (Table 3) and is limited to the seaward-sloping foreshore. The regression equation is then used to derive the horizontal intersection of the operational MHW datum with the profile, giving the shoreline position for that profile. Repeating this procedure at successive profiles 20 m apart generates points that can be connected to create a continu_ous shoreline. To determine the operational MHW elevation, Cali_fornia was divided into 3 sections (Northern, Central and Southern California). For each section, the operational MHW elevation represents an average of MHW elevations from individual open-ocean or near open-ocean tide gauges (Weber and others, 2005). A list of tide gauges and MHW elevations used in each section is presented in Table 3. The lidar- extracted MHW shoreline is not necessarily the same as a MHW shoreline surveyed by a licensed land surveyor. This is because the operational MHW elevation used for the lidar shoreline is an average of the MHW elevations at sev_eral tide gauges. Furthermore, the lidar-extracted shoreline is intended only as a reference feature for measuring shore_line change. It is not intended to establish legal boundaries. Because inland bays generally are not suitable sites for extraction of a lidar shoreline and because this report focuses on the open-ocean coasts, extensive bay areas such as San Francisco and Tomales Bay shorelines were not included in the shoreline change analysis. Also, lidar data were not available for all sandy beaches in California; gaps exist in Northern California along the sandy shorelines near Arcata and Eureka, as well as along a stretch of coast around San Simeon and Cambria in Central California. When lidar data are available for these gaps, the shoreline change analy_ses will be conducted and provided as on-line updates and in future reports. Geographic Information System (GIS) Procedures NOAA T-sheet indexes were used to determine T-sheet availability for shorelines that were not already available as ESRI ArcGIS shapefiles. T-sheets were then requested from NOAA and received as scanned TIF images. Existing digital shorelines for each period were compiled and a quality assessment was performed. Scanned TIF image T-sheets were rectified using Erdas Imagine geographic imaging software by placing at least 6 well-spaced ground control points (GCPs) on selected T- sheet graticules in geographic coordinates. Some T-sheets produced before 1930 required additional coordinate transformation information from NOAA to convert from the United States Standard Datum (USSD) to the North American Datum of 1927 (NAD27). The datum transforma_tion was applied to T-sheet graticule coordinates prior to rectification. Total Root Mean Square (RMS) error for the rectification process was maintained below 1 pixel, which is approximately 4 m at a scale of 1:20,000 and approximately 1.5 m at a scale of 1:10,000. Typically the resulting RMS was much lower than one pixel. Newly geo- referenced T-sheets were loaded in ArcGIS and shorelines were digitized. All shoreline vectors were converted to the Universal Trans_verse Mercator (UTM) projection with the North American Datum of 1983 (NAD83). Shorelines from all sources were merged to produce a single shoreline for each of the 4 time periods by sec_tion of California (Northern, Central and Southern). Final shorelines were coded with 6 attribute fields (ID, Type, Date, Description, Source, and Accuracy) to prepare for calculating shoreline change rates with the Digital Shoreline Analysis System (DSAS; Thieler and others, 2003). Calculation and Presentation of Rates of Change Rates of long-term shoreline change were generated in a GIS with the Digital Shoreline Analysis System (DSAS), an ArcGIS tool developed by the USGS in cooperation with TPMC Environmental Services. The tool is designed to efficiently lead a user through the major steps of shoreline change analysis. This ArcGIS tool contains three main com_ponents that define a baseline, generate orthogonal transects at a user-defined separation along the coast, and calculate rates of change (linear regression, endpoint rate, average of rates, average of endpoints, jackknife). The extension utilizes Visual Basic scripts to develop transects and rates, and uses the Visual Basic programming environment to automate and customize the user interface. Baselines were constructed seaward of, and parallel to, the general trend of the four shorelines. The coastline of California, and hence the baselines, are curvilinear. Using DSAS, transects were spaced 50 m apart. Transects were manually eliminated to prevent calculation of rates in areas where less than four shorelines were intersected. Fewer than four shorelines can result from one or more of the following conditions (Figure 3): 1) the position of a river mouth has changed or migrated, 2) shoreline segments were missing (data gaps), 3) a harbor or other coastal structure eliminated one or more of the shorelines, and 4) no lidar shoreline is available for rocky coasts. Long-term rates of shoreline change were calculated at each transect using a linear regression applied to all four shoreline positions from the earliest (1800s) to the most recent (derived from lidar). Linear regression was selected because it has been shown to be the most statistically robust quantitative method when a limited number of shorelines are available (Crowell and others, 1997). It is also the most commonly applied statistical technique for expressing shore_line movement and estimating rates of change (Crowell and Leatherman, 1999). Short-term rates of shoreline change were calculated using the endpoint method comparing the 1970s and most recent (lidar-derived) shoreline positions. Long-term rates and short-term rates of shoreline change, as defined here, are used throughout the report. Beach Alterations and Shoreline Definitions that Influence Rates of Change Human Activities Attempts to stabilize the shore can greatly influence rates of shoreline change. Activities such as beach nour_ishment or emplacement of shoreline stabilization struc_tures tend to alter coastal processes, sediment transport, and shoreline position. For example, beach nourishment artificially causes rapid, temporary shoreline accretion. Depending on the frequency of beach nourishment, the placement of large volumes of sand on the beach will bias the rates of observed shoreline change toward accretion or stability, even though the natural beach, in the absence of nourishment, may have an erosional trend. In addition, the emplacement of shoreline protection structures such as seawalls, bulkheads and revetments can result in both active and passive erosion of the beach. In the case of passive erosion, the back beach area is fixed by the structure, and the beach in front gradually narrows. Eventually erosion will cease (until the structure fails), thus indicating a stable shoreline in the shoreline change record. Active erosion associated with shoreline protection structures refers to the acceleration of shoreline erosion in front of a structure caused by the alteration of wave, tide and current patterns. Clayton (1991), Flick (1993) and Wiegel (1994) pro_vide a summary of identifiable beach nourishment projects on the U.S. West Coast that had been conducted up to the late 1980s. These records were used to identify shoreline segments that had been influenced by beach nourishment. Only projects that pre-date the lidar shoreline were included. There is no distinction made between large volume, continu_ous projects and small volume, finite projects. According to Flick (1993), Wiegel (1994), and many others, beaches along the coast of Southern California were extensively nourished from the early part of the 20th- century through the mid-1970s. Nourishment programs became far less frequent in the post-1970s era. Differentiating between natural rates of erosion and the influences of beach nourishment is difficult because studies have not been conducted to specifically address this issue. In addition, available data may be inadequate to address this question because there are not enough shoreline positions immediately before, after, and between nourishment proj_ects. Human responses to shoreline erosion, including beach nourishment and emplacement of structures, are included in the discussion of the results of the shoreline change analysis. Shoreline Definitions Inclusion of a lidar-derived shoreline represents a new approach to the investigation of shoreline change. The three pre-lidar historical shorelines come from topographic maps that use the HWL as the shoreline proxy. For more than 150 years, the HWL has served as the most commonly used shoreline because it could be visually identified in the field. With advanced technologies, such as GPS and lidar, it is now possible to objectively define the shoreline on the basis of an elevation or a tidal datum, such as MHW. Changing the shoreline definition from a proxy-based physical feature that is uncontrolled in terms of an elevation datum (HWL) to a datum-based shoreline defined by an elevation contour (MHW) has important implications with regard to inferred changes in shoreline position and calculated rates of change. Morton and others (2004) first compiled published and unpublished data to evaluate the horizontal and verti_cal differences in HWL determined from beach profiles, aerial photographs, or GPS surveys, and the MHW derived from beach profiles, GPS surveys, or lidar surveys. We have updated this to include the most recent available analyses (Table 4). The HWL and MHW positions were established at the same time, or within a few weeks of one another at multiple sites around the U.S. where the beach and wave characteristics are diverse. Comparing these HWL and MHW positions assumes that the observed proxy-datum offsets are entirely artifacts of shoreline definition and are not related to actual changes in the beach profile due to sediment transport (erosion or accretion processes) between the survey dates. This is a relatively safe assump_tion considering the short intervals between surveys or the knowledge that a particular shoreline segment is relatively stable. Moore and others (2006) avoided the need for this assumption by deriving HWL and MHW shorelines from aerial photography and lidar data collected during the same tidal cycle. Table 4 shows that average absolute horizontal and vertical offsets between the HWL and MHW range from a few meters to more than 50 m, and vertical offsets can be as much as 2 m. Most of the horizontal offsets are less than 20 m, and most of the vertical offsets are less than 1 m. Offsets are typically greatest on relatively flat beaches where high waves produce high wave runup (i.e. southwest Washing_ton). Conversely, offsets are least where beaches are rela_tively steep and wave runup is low. For the data analyzed by Morton and others (2004), the MHW was seaward of the HWL on virtually all of the transects (Table 4). This nearly systematic horizontal offset between the HWL and the MHW causes shoreline positions and calculated rates of change to imply slower erosion, a change from erosion to accretion, or faster accretion than actual shoreline move_ment can account for. The recent study by Moore and others (2006) illus_trated that overall, the importance of incorporating a proxy-datum offset into shoreline change analysis depends on several factors including the magnitude of the offset, the length of time over which rates are being measured and the statistical significance of the shoreline change rates. This proxy-datum offset is particularly important when averaging shoreline change rates alongshore. Since the proxy-datum offset is a bias, virtually always acting in the same direction, the error associated with the apparent shoreline change rate shift does not cancel during averaging and it is important to quantify the bias in order to account for the rate shift. The shoreline change rates presented in this report have been calculated by accounting for the proxy-datum bias using the methodology described below. Estimating the proxy-datum bias and the bias uncertainty Comparison of HWL shorelines and a MHW datum-based shoreline for a single-day survey on Assateague Island (Moore and others, 2006) revealed an average hori_zontal offset between shoreline indicators of 18.8 m (Table 4). Vertical offsets were also substantial and were strongly correlated with foreshore beach slope. A simple total water level model, that combines the effects of tidal variations and wave runup, (Ruggiero and others, 1996; Ruggiero and others, 2001; Ruggiero and others, 2003) successfully reproduced these vertical offsets suggesting that the proxy-datum offset is primarily governed by wave runup. In order to estimate the proxy-datum bias for the State of California we use the approach outlined in Moore and others (2006) with the improvement of including the wave runup formula_tion of Stockdon and others (2006). The horizontal offset between HWL and MWH shorelines can be estimated by: Equation (1) where ZT is the tide level, tan b is the beach slope, Ho is the deepwater significant wave height, and Lo is the deep- water wave length given by linear theory as gT2/2p, where g is the acceleration due to gravity and T is the peak wave period. In order to calculate the bias, as well as the bias uncer_tainty, for this regional shoreline change analysis, long-term best estimates and measures of uncertainty are derived for beach slope, wave height, wave length, and tide level. The best estimate for beach slope was derived by averaging individual lidar transect slope estimates within 1-km blocks along the coast. We take the long-term mean wave height and length to be the best estimate to use in the bias calcu_lation. The long-term mean wave height is derived from USACE Wave Information Studies (WIS) hindcasts while the long-term mean wave length is derived from long- term buoy records (NDBC and CDIP) along the California Coast. Finally, the best estimate of the tide level responsible for generating HWL shorelines is taken as the elevation of MHW (see Table 3) (Weber and List, 2005). The mea_sures of uncertainty for the beach slope, wave height, and wave length are estimated as the difference between the 95% exceedance statistic and the 50% exceedance statistic of the cumulative distributions. This gives a 90% confi_dence interval on each of the cumulative distributions. The uncertainty of assuming that the tide responsible for leaving HWL-type shorelines was at MHW is calculated simply by MHHW-MHW The proxy-datum bias, and the associated uncertainty, is calculated at each of the 1-km blocks in which the aver_age beach slope has been calculated. The nearest WIS station, wave buoy, and tide gage to each individual 1-km block were used in the application of Equation 1. Once the bias was calculated, it was incorporated into DSAS and applied on a transect-by-transect basis, so that the estimated bias is removed from the final long- and short-term shore_line change rates. The bias, averaged over 815 1-km sec_tions of the California coast, was approximately 18 m with an average uncertainty of approximately 8.7m. Uncertainties and Errors Documented trends and calculated rates of shoreline change are only as reliable as: (1) measurement errors that determine the accuracy of each shoreline position, (2) sampling errors that account for the variability of shoreline position, and (3) statistical errors associated with compil_ing and comparing shoreline positions. Anders and Byrnes (1991), Crowell and others (1991), Thieler and Danforth (1994), and Moore (2000), provided general estimates of the typical measurement errors associated with mapping methods and materials for historical shorelines, registry of shoreline position relative to geographic coordinates, and shoreline digitizing. For this analysis we report estimates of individual shoreline position uncertainty (Table 5) and shoreline change uncertainty for the regional averages presented in Table 6A-C. Uncertainties associated with shoreline change on individual transects can be calculated using similar meth_ods as are used for the regional averages as discussed below for both long- and short-term analyses. The largest shoreline position errors were errors of _10 m, which were attributed to scales and inaccuracies in the original surveys (T-sheets) and typical positioning errors associated with DRGs (_15 m). Stockdon and others (2002) provided estimates of GPS positioning errors ("1 m) and regression errors ("1.5 m) associated with deriva_tion of the MHW elevation from the lidar data. A previ_ously unreported error term in shoreline change analyses is the uncertainty in HWL shorelines due to variations in water levels. Our estimates of the uncertainty associated with the proxy-datum bias in effect quantify this error term for the first time. Following the methodology of Taylor (1997), Equation 1 was used to translate the estimates of the uncertainty of each parameter into an estimate of the uncertainty of the proxy-datum bias. Equation 1 can also be used to demonstrate that the uncertainty of the proxy- datum bias is equivalent to the uncertainty in any individual HWL shoreline. Sampling errors relating to the local short-term vari_ability of true shoreline positions (Morton, 1991; Douglas and Crowell, 2000) are less well known. Along the Califor_nia coast, as in many locales worldwide, there is pronounced cyclical erosion and accretion of the shoreline. This variability is driven by variations in wave conditions from summer to winter, years with severe versus average storms or swells, and episodic events like El Ni_os or hurricanes. In addition, the seasonal shoreline variability also has a high spatial variability, depending on the orientation of the coast with respect to the wave direction and effects of refraction or reflection from headlands, offshore islands and manmade structures. As a result, an uncertainty term quantifying sea_sonal shoreline variability for regionally averaged shore_line change rates is a rough estimate at best. Site specific, temporally dense data are required to evaluate short-term shoreline variability. The lack of reliable high frequency data regarding short-term variability (of true shoreline position) at most coastal sites limits the ability to quantify this uncertainty into the overall shoreline position uncer_tainty. Studies that do exist for California (Shepard, 1950; Bascom, 1954; Johnson, 1971) focus on changes in beach width or profile. An estimate of the variability in the posi_tion of the MSL intersect on the beach from eleven profile envelopes from La Jolla (Shepard, 1950) is approximately 9 m. In Monterey Bay, similar qualitative estimates from 9 profiles, surveyed over 15 years (Dingler and Reiss, 2002), suggests an average variability envelope of the MHW of approximately 40 m. However, these data include both the 1982-83 and 1997- 98 El Ni¤os and thus incorporate extreme conditions. One of the most extensive records to date is the 20-year record of beach profiles surveyed by the Army Corps of Engineers at Duck, N.C. Using 460 shoreline posi_tions from the Duck profile data, Barton and others (2003) showed that the envelope of shoreline positions around a relatively stable shoreline was about _20 m. Due to the lack of accurate, systematic data regarding the seasonal variation of the shoreline along the California coast, the error values reported here do not include an uncer_tainty term for the seasonal shoreline position variability in the quantification of errors associated with the regionally averaged shoreline change rates. For the long-term shoreline change analysis we assume that the seasonal variability in each shoreline is random and uncorrelated to the others and that the regression error will account for this uncertainty (see Linear regression (long-term) shoreline change error below). For the short- term analysis it is likely, at least in Northern and Central California, that due to the length of time over which our rates are calculated, the seasonal variability uncertainty term has a negligible impact on the total error value. In Northern and Central California, the shorelines from the most recent historical shoreline DRGs are based on aerial photographs from the 1940s and 1950s. Table 2 includes not only the years of the data sources, but also the months of the photography on which the T-sheet maps were based for the most recent historical shoreline. In Southern California nearly all of the source data are from the winter months, as is the lidar data. Therefore, the sea_sonal variability term is again assumed not to have a mea_surable impact on the error term for the regionally averaged rates presented in this report. Independent comparisons of our shoreline change results with published rates in South_ern Monterey Bay (Thornton and others, 2006) are in close agreement, even in the short-term. However, we recommend that anyone using the data associated with this report for a more site-specific analysis incorporate an error term to account for seasonal shoreline position variability. Estimates of the maximum positional errors for this study are provided in Table 5 to show how each error con_tributes to inaccuracy in the shoreline position. The annual_ized error for short-term shoreline change is calculated and subsequently incorporated into the shoreline change rate calculations as outlined below. The uncertainty on the short-term (end-point) rates, using a best estimate for California shorelines is "0.4 m/yr (Table 5 and Table 6A- C). End-point (short-term) shoreline change error The total error for the end-point shoreline change rate, (Esp)(Equation 2), is calculated by taking the square root of the sum of the squares (or adding in quadrature) of: georef_erencing error (Eg), digitizing error (Ed), T-sheet survey or DRG error (Et), and shoreline position error (Ep), similar to the methods outlined by Crowell and others (1993). Values for each of the error terms are given in Table 5. The georeferencing error represents the elected maximum acceptable RMS error for T-sheets at a scale of 1:20,000 in this study. The georeferencing error is applied to the historical shorelines that are derived from T-sheets only. The digitizing error reflects the maximum error specified in past studies (Anders and Byrnes, 1991; Crowell and others, 1991; Moore, 2000), and is applied to the historical shore_lines only. The maximum T-sheet survey error, determined by Shalowitz (1964), incorporates all of the errors associ_ated with the mapping process including distance to rodded points, plane table position, and identification of the HWL. The maximum DRG error is the stated accuracy of USGS Topographic Quadrangle maps from which the DRGs are derived; National Map Accuracy Standards give _15 m as the maximum acceptable error. The T-sheet survey error is applied to all historical shorelines; however, note that the error associated with the 1950s- 1970s era T-sheets is considerably lower than the older T- sheets; this difference is based on findings by Ruggiero and others (2003), as well as the fact that more recent shorelines are derived from aerial photos or other sources. Shoreline position error is the maximum error associated with the derivation of the lidar shoreline (Stockdon and others, 2002) for lidar data and the average bias uncertainty (8.7 m) for the historical shore_lines. Thus, the total shoreline position error as shown in Table 5 for each shoreline is expressed by: Equation (2) A separate Esp is calculated for each period. For the short-term shoreline change rates only these values can be com_bined and annualized to provide an error estimation for the shoreline change rate at any given transect. The annualized error (Ea) is expressed by: Equation (3) For determining short-term uncertainty error at a specific location (at an individual transect) we can add the uncertainty terms from Table 5 to get a total uncertainty of the shoreline change rate at a given location. Dividing this total by the time between shoreline dates provides the error on the short-term change rate at that location. Linear regression (long-term) shoreline change error Linear regression is the most commonly applied statistical technique for expressing shoreline movement and estimating rates of change (Crowell and Leatherman, 1999) where there are a statistically valid number of samples. Because linear regression fails to recognize the potential for temporal differences in trend (trend reversals) and accelera_tions or decelerations (Morton, 1991; 1996), average trends and rates of shoreline change in this study were calculated for both long-term (entire period) and short-term (most recent) time scales. Long-term rates of shoreline change were determined at each transect by taking the slope of the regression line applied to all four shoreline positions. The resulting rates are reported in units of m/yr (Table 6A-C). Uncertainties for the average shoreline change rates are reported as the " values in Table 6A-C. Two uncertainty terms arise in the calculation of the long- term shoreline change rates. The first term is the 90% confidence interval of the linear regression shoreline change rate for each transect. The second term arises from the uncertainty in our best estimates of the proxy-datum offsets. We calculate linear regression slopes using shoreline data that has been adjusted based on our best estimate of the proxy-datum bias as well as data that has been adjusted according to our best estimate of the " the bias uncertainty. From this analysis we get a best estimate of the shoreline change rate and an uncertainty of the rate due to the bias uncertainty. At each transect we can add the regression error and the proxy-datum bias uncertainty error to get a total uncertainty of the shoreline change rate at a given loca_tion. However, in terms of calculating regionally-aver_aged shoreline change rate uncertainties the two terms discussed above need to be treated differently. Because the 90% confidence interval on the linear regression of each transect is assumed to be random and independent, when averaged over many transects the resulting average uncertainty associated with this term can be quite small; the greater the number of transects over which the uncertainty is averaged, the smaller the uncertainty on the average rate. However, for the second term we need to account for the fact that the proxy-datum offset is a bias and always acts in one direction. Therefore, the regionally averaged shoreline change rate uncertainty associated with the proxy-datum bias is simply the average value of the error resulting from the uncertainty of the proxy-datum bias. The regionally averaged total shoreline change uncertainty terms can be expressed by: Equation (4) where U is the alongshore averaged shoreline change rate uncertainty, C is the linear regression 90% confidence inter_val, B is the shoreline change rate uncertainty associated with the proxy-datum bias, and n is the number of transects included in average. Field observations and prior studies of shoreline move_ment within each analysis region in California suggest that the trends and relative rates of change presented in this study are as accurate as the methodology allows. Reliability of the mapped results increases as both the persistence of the trend and the magnitude of the rates increase. Stated another way, confidence in the analytical results is greatest where the rates of shoreline erosion or accretion are high and the trend has persisted for decades. On the other hand, confidence in the absolute results decreases where the shore_line is relatively stable and the rates of change are low. This is because minor differences in historical or lidar shoreline positions can substantially alter the regression line and the calculated results. Data confidence also decreases in areas where frequent trends reversals occur. Advanced technology such as GPS and lidar can better constrain shoreline positions, reduce the methodological errors, and improve the accuracy (reduce the error) of future shorelines. Establishing a datum-based shoreline (lidar derived MHW) as the standard for comparison provides, for the first time, the ability to perform an error analysis that is both quantitative and meaningful, in terms of its application. In the future, each electronic MHW shoreline could be pre_sented with an accompanying error bar that would define the alongshore envelope of confidence. Subsequent shorelines and associated confidence envelopes would provide a more precise basis for determining the statistical significance of observed shoreline change. Unfortunately, the use of lidar or any other shoreline mapping technology will still require distinguishing between short-term variability in shoreline position and the long-term trend of shoreline change. GEOLOGIC HISTORY AND SETTING California straddles the boundary between the Pacific and North American tectonic plates (Figure 4). The diverse landscape and complex geology of the California coast is largely a result of the interactions between these plates. Lateral movement between the two plates occurs along the San Andreas Fault Zone (SAFZ), which extends nearly 1300 kilometers from the Gulf of California to Shelter Cove near Point Delgada in Humboldt County (Figure 5). Lateral movement on the fault zone averages 2.5-3 cm/yr (Harden, 1998) with a total accumulated displacement from slip dur_ing earthquakes and aseismic creep of at least 560 kilome_ters since lateral movement began about 15-25 million years ago. Inman and Nordstrom (1971) recognized the impor_tance of tectonic setting to the development of coasts worldwide. They developed a coastal classification scheme based upon the position of a given coast relative to plate movement. Within this classification scheme they recog_nized three first-order classes: collision coasts, trailing-edge coasts, and marginal sea coasts. Collision coasts (i.e. active margins) are characterized as being relatively straight and mountainous, having the presence of coastal cliffs and raised marine terraces, and bordered by narrow continental shelves. Coastal watersheds are typically steep and undergo high rates of erosion. Although the California coast south of Cape Mendocino presently borders a transform margin dominated by lateral movement, and is therefore not strictly collision controlled, it maintains many of the characteris_tics of a collision coast. Between Cape Mendocino and the Oregon border, the coast is a collision coast. Movement along the SAFZ has created three broad geomorphic provinces (Figure 4) along the coast: Coast Ranges, Transverse Ranges, and Peninsular Ranges (Cali_fornia Geological Survey, 2002a). The northwest trending Coast Ranges run roughly parallel to the SAFZ and extend from the Oregon border in the north to the east-west trend_ing Transverse Ranges in the south. The Coast Ranges are composed mostly of uplifted Mesozoic and Cenozoic marine sedimentary rocks that typically form a terraced and wave- cut coastline. The range is broken by the depression forming San Francisco Bay. The Transverse Ranges trend roughly east-west and lie oblique to the general northwest trend of the California coast. They extend from the San Bernardino Mountains to the offshore islands of San Miguel, Santa Rosa, and Santa Cruz. Intense north-south compression from SAFZ move_ment has resulted in very high uplift rates in this region (California Geological Survey, 2002a). Tertiary sedimen_tary rocks are the dominant rock type of the Transverse Ranges along the coast. The Peninsular Ranges extend from the Transverse Ranges to the Mexican border, and, like the Coast Ranges, trend in a northwest direction. Along the coast they are composed mostly of Tertiary sedimentary rocks and further inland are characterized by Sierra Nevada-type rocks includ_ing granitics and older metamorphic rocks (Figure 6). The Los Angeles basin and offshore islands of Santa Catalina, Santa Barbara, San Clemente, and San Nicolas are consid_ered part of the province. To summarize, the features of the California coast have formed over millions of years of interaction between two large tectonic plates that continues today. Proxim_ity to an active tectonic margin results in features such as: diverse rock types in close juxtaposition, high relief, high erosion rates of the land surface, and high rates of sediment supply to the coast. Block uplift and subsidence between high-angle reverse faults occur within the broader transform margin scenario. Segmentation of the coast through these types of vertical crustal movements, as well as horizontal (strike-slip) displacement forms important structural founda_tions for coastal sedimentary environments. General Geology of the California Coast The diverse morphology of the California coast is primarily a result of the local geology where lithology, geologic structure, and vertical tectonic movement play a prominent role in the configuration of the coast. Figure 6 is a generalized geologic map showing the major rock types of California (California Geologic Survey, 2002b). Tertiary and Mesozoic sedimentary rocks are clearly the dominant coastal rock type; they are mostly marine in origin and represent sediment deposition, lithification, and uplift along the Pacific-American plate subduction / transform bound_ary. The Mesozoic rocks, which include the Franciscan Complex, are typically sandstone and shale from oceanic crust and deeper marine settings. The Tertiary rocks such as those of the Monterey Formation tend to be sandstone, shale, and conglomerate from more shallow marine environ_ments closer to the continental margin. Crystalline rocks are also present along the coast and are most common near San Francisco and Monterey. The strength of the rocks exposed along the coast is a critical parameter in determining the erodability of the coast (Benumoff and others, 2000; Hapke, 2005). Stronger rocks form prominent headlands that resist erosion and often form natural boundaries to littoral and aeolian transport. Weaker rocks erode more quickly and form embayments where coastal sediment may accumulate. Monterey Bay is an excellent example of an embayed coast where resistant rocks in Santa Cruz and Monterey (Figure 5) form head_lands and the interior of the bay is backed by easily eroded Quaternary shallow marine, aeolian, and fluvial deposits. Coastal cliffs tend to be either high, steeply- dipping coastal mountains that plunge directly into the sea, or, broad near- planar marine terraces. Marine terraces are prominent features for much of the California coast and are best developed where uplifted marine clastic rocks form the bedrock. Terrace preservation varies from moderate to poor in the other rock types that form coastal slopes including metamorphic, granitic, and ophiolitic terranes. Marine terraces form when coastal cliff retreat generates wave-cut platforms, most notably during sea-level highstands, and are preserved as a slightly sea_ward-sloping planar surface during tectonic uplift (Anderson and others, 1999). Local uplift rates, duration of marine pla_nation, and terrace composition determine the width and ele_vation of the terraces - they are typically 10s of meters high and 100s of meters wide. The terrace surface often contains beach, dune, or alluvial deposits and when combined with terrace erosional material they can provide an important component of sediment contribution to the coast. Weaker rock types with an abundant sand component may contrib_ute a significant amount of sediment to the beach system (up to ~10-30%; Hearon and Willis, 2002; Inman and Mas_ters, 1991; Runyan and Griggs, 2002). Table 7 shows the approximate amount of different rock types for the cliffed portion of the California coast (~72% of the 1760 km long coast). Cliff retreat rates vary dramatically from very low in granitic terranes to several meters per year in cliffs formed in poorly-consolidated sediment. In addition to providing sediment to the coast, marine terraces are important features because of their low surface relief and proximity to the ocean they are the sites of numerous developments along the California coast (Griggs and others, 2005c). Climate The climate of California is strongly influenced by a persistent zone of high pressure in the north Pacific, a southerly flowing cold water ocean current, and the Sierra Nevada mountains, which block the continental air from affecting the coastal climate. During the summer months the northward migration of the semi-permanent North Pacific High diverts most storm tracks to the north. California seldom receives rain from Pacific storms during the summer but coastal fog is widespread. Cold upwelling waters at the surface come into contact with the relatively warm moist air from the Pacific causing massive fog banks to form. Dur_ing the winter the North Pacific High migrates southward directing storms towards California. Occasionally storms will arrive from the southwest and are accompanied by relatively warm temperatures and heavy rains (often referred to as the pineapple express). Average annual precipitation varies dramatically from north to south with 80 inches and above in the north and only about 10 inches reaching the San Diego area. The seasonal weather patterns are modified during El Ni_o and La Ni_a events. During El Ni_os California's climate is typically characterized by above normal rain_fall, warmer sea-surface temperatures, and large waves from Pacific-generated storms often resulting in increased beach erosion. The 1997-98 ENSO (El Ni¤o - Southern Oscillation) was a significant climatic event responsible for widespread coastal flooding and beach loss (USGS/UCSC/NASA/NOAA Collaborative Research Group, 1998). La Ni_as are generally accompanied by colder ocean temperatures, drier conditions, and less severe storms. El Ni_o's and La Ni_a's generally last for 6 to 18 months and their occurrence and intensity are related to longer term atmospheric variations termed the Pacific Decadal Oscil_lation (PDO). The PDO is an ENSO-like phenomenon that lasts for 20 to 30 years and consists of warm and cool phases (Zhang and others, 1997). The cool phase, which is likened to an extended La Ni_a, is characterized by a cool wedge of lower than normal sea-surface heights and ocean temperatures in the eastern equatorial Pacific resulting in cooler temperatures and lower rainfall in California. Condi_tions during the warm phase are reversed and are similar to extended El Ni¤o conditions. Because phases tend to last between 20-30 years, with the last warm phase starting in 1977, some believe that we have entered a cool phase marked by the inception of the 1998/99 La Ni_a event as discussed in Hare and Mantua (2002). Coastal Processes Waves Waves and currents are the primary forces that move sediment in the littoral zone and annual wave height varia_tions are responsible for seasonal erosion and accretion patterns. Wave characteristics along the California coast depend on weather patterns, geographical effects such as offshore islands, storm climatology, coastline orientation and local bathymetry. The offshore wave climate of Cali_fornia is characterized by four regimes (Figure 7): Northern Pacific swell, Southern Hemisphere swell, northwest wind waves, and by locally driven seas (Storlazzi and Wingfield, 2005). North Pacific swell is generated by extra tropical storms, mid-latitude low-pressure systems, and cold fronts that originate in the North Pacific. Southern swell is domi_nant in the summertime and generated by winter storms in the Southern Hemisphere. Northwest wind waves gener_ated by daily sea breeze conditions are more common in the northern part of the state and are strongest in the spring and early summer months. Local seas are driven by wind and storms along the coast. Given the variety of local and sea_sonal variations in wave climate, the predominant direction of nearshore sediment transport along the California coast is from north-to-south (Hearon and Willis, 2002). Wave climate varies along the California coast and can region_ally summarized as follows (after Storlazzi and Wingfield, 2005): In northern California the average wave height is greatest from November to February and averages about 3 m, with approximately 20% of the time wave heights are greater than 4 m. Summer wave heights are smaller with mean values around 1.8 m with waves higher than 4 m being extremely rare. Early winter is the most common time for waves to exceed 6 m. During El Ni¤o winter months mean annual wave heights are 0.3 m - 1.2 m greater than normal winter months. El Ni¤o driven storms typically approach from the west or southwest and may cause local littoral drift to the north - counter to the predominant southerly drift. La Ni_a winter months have slightly higher than average wave height values of 0.1 m - 0.4 m whereas during the summers wave heights are smaller than average. Central California is a transition zone between harsh stormy waves of the northeast Pacific and milder conditions of Southern California. North Pacific Swell is the largest swell to impact the region with heights between 2m - 10m and periods ranging between 10 - 25 seconds. These waves, which are generated by storms in the North Pacific, occur most commonly between October and May. Northwest wind waves are generated from sea breeze and are dominant from April to October. The height of these waves typi_cally varies between 1 and 4 meters with a period of 3 to 10 seconds. Southern Swell occurs between April and October and typical wave heights range from 0.3 - 3 m with a period of 10 - 25 seconds. Local wind-driven waves are generated by storms passing through central California. They gener_ally occur between October and April with typical heights 1 and 4 meters and periods of 3 to 10 seconds. In southern California peak wave heights are greatest from November to February and average about 2.4 m during this time. In the summer wave heights are smaller with mean values around 1.8 m. Waves greater than four meters occur about 11% of the time at Point Concep_tion (Figure 5) and are most common during the month of March. Waves that damage the Southern California coast originate from extra- tropical storms in the northeast Pacific or Southern Hemisphere, although the second case is rare (Newberger, 1982). During El Ni_o winter months, wave heights at Point Conception are 0.7 m above average. In general, the southern region of the West Coast experiences more storms and higher wave energy during ENSO events (Seymour, 1998). Wave conditions along the southern Cali_fornia coast are extremely variable due to coastal configura_tion, bathymetry, orientation of coastline and the presence of several large offshore islands. Wave height measure_ments can be substantially different over distances of a few miles (Newberger, 1982). The Channel Islands block waves approaching from the south and Point Conception blocks waves from the north. Waves that propagate into the chan_nel are severely refracted by irregular shallow bathymetry, producing large spatial variations in swell wave height and direction (Guza and O'Reilly, 2001). Tides California has a mixed semidiurnal tidal regime of two unequal high and low tides a day with total open-coast elevation changes of about 2.1 m in Crescent City and 1.6 m in La Jolla (Figure 5). The two daily high and low tides are unequal in amplitude with the lower-low tide of the day usu_ally following the higher-high. The monthly tidal variations are dominated by the spring-neap cycle with spring tides occurring during full and new moons and neap tides occur_ring during half moons. The highest monthly tides during summer and winter months are higher than the highest tides in fall or spring. Tide ranges increase from the south to north along the coast and higher tide ranges occur in San Diego and San Francisco Bays than in adjacent open coast areas (Flick, R.E., 1998). Tidal range influences beach processes along the California coast because it determines the extent of beach exposure and inundation throughout the tidal cycle. Espe_cially crucial to beach erosion episodes are the timing and height of the highest tides in conjunction with the maximum wave height and surge developed during storms. Com_parisons between impacts of the 1982-83 and 1997-98 El Ni_o storms on California's coast (Storlazzi and Griggs, 1998; Storlazzi and others, 2000) show that greater dam_age occurred in the 1982-83 event, in part, because the high tides were slightly higher and peak waves coincided with maximum high tides. The differences in tidal height between these two El Ni¤o winters is primarily due to the 4.4yr lunar perigee cycle with a small contribution from the 18.6 year lunar node cycle (Flick, 1998). This cycle enhanced peak high tides in 1982-83, 1986-87, 1990-91, 1995-96, and 1999-2000. Long-term trends in California tide records are consistent with the general rise in mean sea level over the last century recorded throughout the world. Figure 8 shows long-term trends in mean sea level for selected California tide gauges as reported by the National Ocean Service (http://140.90.121.76/sltrends/sltrends_states.shtml?region =ca), and analyzed by Flick and others (2003). Both San Francisco and San Diego exhibit about 20 cm of sea-level rise over 93 years whereas Los Angeles exhibits a lower total rise of only 6 cm over 76 years. Crescent City, an exception to the sea-level rise trends, shows a relative sea-level fall of 3.2 cm over 66 years that is most likely the result of local tectonic uplift. Winds Winds are important for generating waves that drive littoral sediment transport and for blowing sand off beaches resulting in potential sand loss from the littoral system. The wind climate of California is strongly influ_enced by the North Pacific High that creates a predominant northwesterly air flow over most of the state. The intensity and position of the North Pacific High often determines the direction and strength of winds affecting coastal California. The North Pacific High is stronger and located more north_erly during summer months and moves south and weakens during winter months, allowing storms to reach the state. Coastal wind flow is predominantly parallel to the coast either from the northwest or the southeast. When winds are from the northwest, flow is along the coast but Ekman trans_port induces an onshore component, which is strengthened by local sea breezes (Zhiqian and others, 1997). Associated with northwest winds is the creation of a jet effect in the vicinity of some of the more prominent headlands. Strong jets of air and large eddies are projected around headlands such as Point Reyes, Point Sur, and Point Arguello (Figure 5). Wind speeds in the immediate vicinity of these major headlands can be two or three times as great as the wind flow at nearby areas. In general, wind flow from the north is more aligned with the coast, the strongest flow is pushed offshore, and there is usually no associated daily sea breeze. During strong northern flows high wind speeds extends over a large area from Northern California to Point Conception. During northeast or east flow conditions the along-coast variability is significantly larger and winds are weaker (Zhiqian and others, 1997). These wind patterns are altered by the passage of Pacific storms most of which arrive in the winter. As a storm approaches, the cold fronts are marked by strong easterly or southeasterly winds that can reach speeds of 50 km/hr or more. After the storm passes winds turn toward the southwest. With the passage of the rare warm front, storm winds can reach 30 - 40 km/hr. In Southern Califor_nia, after the passage of a cold front, Santa Ana winds will often blow down from the north to northeast. These winds are sometimes extremely intense and can blow between 90 to 145 km/hr and extend 160 kilometers seaward. Although it is rare, when Santa Ana winds blow during the summer they produce hot and dry conditions that increase the fire danger. Currents The California current system forms the eastern edge of the North Pacific gyre and is primarily driven by wind stress patterns over the North Pacific Ocean (California Coastal Commission, 1997). Changes of the ocean circu_lation pattern are caused by the interactions between the sub- tropical high pressure cell over the North Pacific and the atmospheric thermal low over California and Nevada. The interaction between these pressure regions results in a dominant southward-directed wind along the coast of Cali_fornia in spring and summer driving the California Current towards the equator (Hickey, 1979). Maximum southward wind along the California coast occurs between Cape Men_docino and San Francisco in the spring (Nelson, 1977). The associated Ekman transport moves water perpendicularly away from the coast allowing cold, nutrient-rich waters to upwell from the deep (Newberger, 1982). During the late fall and early winter southward winds weaken, reducing upwelling and allowing a near-shore, northward-flowing current north of Point Conception called the Davidson Current, to prevail (Hickey, 1979). In the Southern California Bight (coastal Southern California and offshore waters) a counter-clockwise eddy occurs called the Southern California Countercurrent. It is a northwestward- flowing current south of Point Conception and inshore from the Channel Islands. The current runs shore parallel until it reaches Point Conception where one branch flows southwest joining the California Current and the other branch con_tinues northward as a narrow countercurrent (Newberger, 1982). The Southern California Counter current occurs in all seasons but is best developed in winter (Maloney and Chan, 1974). Littoral cells and transport directions The prevailing southerly transport direction for Cali_fornia littoral sediment is driven by North Pacific swell and northwest wind waves (see Figure 7). There are local reversals in this prevailing direction due to orientation of the coast and/or southerly wave events. Littoral cells are seg_ments of the coast with distinct sediment sources, defined longshore transport pathways, and sinks where the sediment is removed from the littoral system. Conceptually, the cell boundaries delineate an area where the sediment budget can be balanced for quantitative analysis. Southern California littoral cells were first defined by Inman and Chamberlain (1960) and statewide littoral cells were identified by Habel and Armstrong (1978). In California the cells are typi_cally bound by either prominent rocky headlands that block littoral transport around them, or, submarine canyons that cross the continental shelf to a shallow enough depth as to intercept alongshore moving sediment. Submarine canyons are clearly the largest sink for beach sand loss in California with an estimated removal from some of the larger canyons at: Scripps and La Jolla - 270,000 m3/yr, Mugu - 765,000 m3/yr and Monterey - 230,000 m3/yr (Griggs and others, 2005b). Once sediment enters the submarine canyon system it is permanently lost from the littoral system. Another large sediment sinks are coastal dune fields where wind-blown sand is removed from the active littoral system. However, unlike submarine canyons, subsequent erosion of the dunes can re-supply adjacent beaches with sand. Figure 9 shows the boundaries of littoral cells and coastal watersheds along the California coast. The boundaries between many of these cells, however, and the amounts and rates of sediment transport are poorly understood. Long-term harbor dredge records are one of the best long-term sources of longshore transport rates (Table 8) where harbor dredging is under_taken. Sand Sources The primary sources of coastal sediment for California are the fluvial drainage systems that reach the coast. These systems range from short, steep, ephemeral streams that deliver a wide mix of sediment grain sizes, to more mature rivers which often have well-developed estuaries. Califor_nia's coastal streams have exceptionally high sediment loads due to the steep landscapes, geologically young and tectoni_cally active terrain, and, in central and Southern California, relatively sparse vegetation cover (Willis, 2002a). Sediment yield per size of drainage basin for California rivers is typi_cally very high when compared to other regions of the U.S. Estimated sand and gravel discharge for the major California streams that enter the open coast is shown in Figure 10. Average annual bedload discharges range from a few thousand m3/yr for the smaller creeks to nearly 3 million m3/yr for the Eel River in northern California (Wil_lis, 2002a and b) (Figure 5). These estimates should be considered maximum estimates of beach-quality material supplied from coastal streams because of numerous uncer_tainties and assumptions, and the fact that they include sand finer than 0.125 mm which is unlikely to remain in an energetic beach environment. In addition to the rivers shown in Figure 10, the large Sacramento and San Joaquin Rivers empty into San Francisco Bay (Figure 5), the largest estuary on the west coast. A large ebb-tidal delta has formed at the entrance to the bay. Numerous smaller ungauged streams also reach the coast and can supply significant amounts of sediment because of their steep, easily erodible watersheds (Willis, 2002b). On average 70 to 95% of the beach sand in California is derived from coastal streams (Runyon and Griggs, 2002; Willis, 2002a). In general, sand and gravel discharges from coastal watersheds decrease from north to south (primarily rainfall controlled), although the Trans_verse Range has a relatively high sediment discharge (lithol_ogy and vegetation controlled). Infrequent severe floods are thought to be responsible for delivering the majority of sedi_ment to the coast and a single large storm can deliver more sand to the beaches than years of low to moderate rainfall. In addition, sediment discharge during extreme events can lead to open-coast ephemeral delta formation (Richmond, 1988) and delivery of abundant coarse-grained sediment. El Ni¤o years are typically times of significant sediment introduction to the coast because of increased likelihood of extreme rainfall events (Inman and Jenkins, 1999). Coastal cliffs are the next major source and sand sup_ply varies with cliff lithology and strength. In some areas such as the Oceanside littoral cell, the coastal bluffs have been found to be a major source of beach sand (Young and Ashford, 2006). Softer cliffs composed of coastal sand deposits provide the most beach quality sediment when eroded. Subordinate sources of coastal sediment include marine planation of submerged rock, material of biologic origin such as shells, and possibly onshore transport of relict shelf sediment. GEOMORPHOLOGY OF THE CALIFORNIA COAST The California coast encompasses a wide range of coastal landforms a product of complex geology and dynamic coastal processes. Coastal landforms include steep cliffs, uplifted terraces, beaches, dunes, barrier spits, estuar_ies and lagoons (Figure 11). Cliffs Nearly three-fourths of the California coastline are backed by cliffs which fall into two broad general catego_ries: high steep cliffs and marine terraces. High cliffs occur where mountains directly border the coast such as along the Big Sur coast and most of northern California. The high cliffs may be hundreds of meters or more in height, they occupy about 13% of the California coastline, and are typi_cally composed of more resistant rock types such as granite and rocks of the Franciscan Complex (Griggs and Patsch, 2004). Marine terraces and coastal bluffs, which were dis_cussed earlier, form about 60% of the remaining coast and are common from Mendocino to San Diego. Where tectonic uplift has persisted, multiple terraces are often preserved. Beaches Beaches are ubiquitous features of the California coast and are important for a number of reasons: a) they act as a natural buffer that protects coastal land during storms, b) they are a valuable recreational and economic resource, and c) they provide habitat such as nesting sites for the endan_gered snowy plover and haul-outs for protected marine mammals. California beaches are not as long and continu_ous as those along passive margins (e.g. the U.S. South Atlantic and Gulf coasts) in part because the young and steep nature of the coast has not allowed enough geologic time for extensive sandy coastal plains to develop. Beach types found in California include pocket beaches, long expanses of linear to gently curved beaches, barrier spit beaches at stream mouths, and cuspate headlands. Pocket beaches are bound by headlands, and occur in both small stream valley and cliffed-coast settings. Pocket beaches are probably the most common beach type in California, although their total length is smaller than the total length of California's linear beaches. Long expanses of beach typi_cally front the major dune complexes, larger stream valleys, and coastal plain or concaved areas (e.g. Monterey Bay and Santa Monica Bay). Cliffed coastlines can be fronted by both permanent and seasonal beaches. Permanent beaches occur where there is abundant sediment supply, both along_shore and offshore. Seasonal beaches, which typically are present in the summer months and are lost during winter months, are common along exposed coasts with a limited offshore supply of sand. Because of the relatively high wave energy and a steeper and narrower continental shelf along the California coast, pronounced deltas do not form. Instead, barrier spits and ebb and tidal bars develop where the streams reach the sea. Beaches that form at the mouths of stream valleys and embayments are typically a mixture of both fluvial- and littoral-derived sediment (mostly sand). The barriers are typically barren to sparsely vegetated indicating an unstable substrate prone to occasional marine overwash and breach_ing. Seasonal changes in wave climate and rainfall result in a concomitant change in barrier style. In winter, periods of high waves and heavy rainfall cause overwash and chan_nelization of the barrier spits, reducing their overall size. During the summer months, there are smaller waves and low precipitation and the barrier spits may completely block stream mouths due to reduced stream flow and beach accre_tion. Seasonal beach change in California is caused by annual variations in wave climate that produce narrow beaches during winter months and wide beaches during the calmer summer months (Dingler and Reiss, 2002). Dra_matic beach erosion, both in rate and amount, occurs during large storms. Subsequent recovery is less rapid, often requiring several months for the beach to achieve its pre-storm configuration. Beaches without an abundant offshore sand supply take much longer to recover (Brown and others, 1998). Figure 12 illustrates the seasonal change in beach profile shape during the intense 1997-98 El Ni_o winter season at Cowells Beach in Santa Cruz. Coastal Dune Complexes Cooper (1967) mapped the coastal dunes of Califor_nia and recognized that coastal dune complexes are best developed where there is: a) a nearby source of fluvial-sup_plied sediment, b) a structural high at the coast, such as a headland, to trap littoral drift or a low-relief stretch so dunes can migrate inland, and, c) strong consistent onshore winds. Orme (1992) further noted that dune fields are best preserved in coastal areas that have undergone net tectonic subsidence or limited uplift in the Quaternary. Aeolian deposits are often interbedded with fluvial and nearshore facies and the larger complexes represent multiple episodes of dune building. In many areas the modern dunes represent surficial deposits overlying older, and larger, dune systems. There is some indication (Orme, 1992) that significant coastal dune building occurs at lower sea-level positions when large quantities of sand are exposed on the emergent continental shelf. Modern dune building removes sediment from the lit_toral supply; in some places this can be a substantial portion of the littoral sediment budget. For example, it has been estimated that about 150,000 m3 of sand are blown inland each year along the 55 km stretch of coastline from Pismo Beach to Point Arguello (Figure 5) (Griggs and others, 2005c). Where the present coastline is undergoing retreat, such as in southern Monterey Bay, the dunes are reworked and supply sediment to the beach. The major dune com_plexes of California are shown in Figure 13 along with their effective wind directions. Estuaries and Lagoons U.S. West Coast estuaries and coastal lagoons typi_cally form in drowned-stream valleys cut below the level of the uplifted coastal plain (Peterson and Phipps, 1992) or in subsiding coastal blocks. Four general types of estuarine/lagoon embayments occur in California: a) large embay_ments with high freshwater inflow, b) large embayments with relatively low freshwater inflow, c) large freshwater bodies with limited intertidal environments, and, d) ephem_eral streams with limited estuarine environments. The largest California estuary is the San Francisco Bay system that forms the outlet for the contiguous Sacramento- San Joaquin Delta watershed. This is a large embayment (~4,100 km2) with high freshwater inflow that drains more than 40% of the land area of the State of California (Chin and others, 2004). Bay environments include marshes, intertidal mudflats, and subtidal channels. The remainder of estuaries in California are much smaller in size but typically contain the same depositional environments. Embayments currently with low fluvial input, such as Bolinas Lagoon, Drakes Estero, Tomales Bay, Bodega Bay, Elkhorn Slough, and Morro Bay (Figure 5), appear to be structurally controlled depressions not presently connected to a major fluvial source. In these examples, the embayment size is large compared to present stream discharge. Embay_ment downcutting was probably enhanced during lower sea-level stands and the subsequent period of higher Holocene sea level resulted in bay infilling. At present, many of the morphologic bay features appear to be tidally controlled. The major rivers of California are typically character_ized by relatively high flow but narrow confined estuaries. These drainages are characterized by well-defined stream channels entering a restricted coastal depositional plain, and the location of the stream mouth is often controlled by a geologic feature such as a resistant headland. Inter_tidal environments are relatively limited in size because of extensive floodplain deposition (abundant sand). The rivers are the main suppliers of sand to the California coast (Figure 10). Fine-grained sediment typically bypasses the coastal zone and is deposited in deeper water. Ephemeral streams are similar to the larger rivers but on a smaller scale. General Characteristics of the California Coast Sections Northern Section: Oregon Border to Tomales Bay The coast of Northern California (Figure 14A) can be characterized as a rugged landscape with high rainfall and low population. Steep coastal cliffs dissected by numer_ous streams result in high sediment loads delivered to the coast. Franciscan Complex rocks are common and the more resistant units often result in an irregular coast with steep cliffs, small offshore islands and sea stacks. Barrier spits and beaches are common features at stream valleys and embayments with the largest barrier in the region extending across Humboldt Bay. Large dune complexes occur south of Smith River, between the Little and Eel Rivers, and south of Tenmile River (Figure 13). Other large dune fields are present north of headlands at Point Arena and Bodega Head, and at the entrance to Tomales Bay (Figure 5). Marine terraces and wave-cut bluffs are common between the areas dominated by the steep mountain cliffs. The terraces south of Cape Mendocino are Holocene features that are undergo_ing rapid uplift. According to Savoy and others (2005), as much as 1 m of uplift occurred during a single earthquake in 1992 along the Cascadia subduction zone. The heads of Mattole and Delgada submarine canyons reach into shallow water where they can intercept littoral transport. Central Section: Tomales Bay to Point Conception Central California (Figure 14B) is the most diverse coastal region of the state having characteristics of both the north and south regions plus a few unique features of its own. This section represents the transition zone between the relatively wet and high wave energy north and the drier and lower wave energy southern section. Unique embayments at Tomales, San Francisco, Monterey, and Morro-Estero Bays (Figure 5) form natural harbors along the rugged coastline. Marine terraces and coastal bluffs are well devel_oped south of Point Reyes, in the Monterey Bay region, parts of the southern Big Sur coast, and stretches along the San Luis Obispo County coast. High relief coastal slopes occur at the Marin Headlands and Devils Slide north and south of San Francisco respectively, and along most of the Big Sur coast. Between Morro Bay and Point Conception, coastal mountains of the San Luis Range, Point Sal Ridge, and the Santa Ynez Mountains of the western Transverse Ranges alternate with intervening basins forming the greater Santa Maria basin. There are large dune complexes at Point Reyes, southern Monterey Bay, Morro Bay, and near the mouths of the Santa Maria and Santa Ynez Rivers (Figure 5). The Santa Maria and Santa Ynez Rivers are presently dammed resulting in a significant reduction of sediment reaching the coast from past conditions. The heads of Mon_terey, Carmel, and Partington submarine canyons lie just offshore where they are thought to be major sinks for beach sand moving alongshore. Southern Section: Point Conception to the Mexican Border The coast of Southern California, extending from Point Conception to the Mexican border (Figure 14C), is markedly different from the rest of the state. Point Concep_tion marks a dramatic change in coastal orientation due to tectonic movement along the Transverse Ranges that has resulted in an east-west trending coast. Further south, the coast gradually returns to the northwest-southeast trend. Coastal cliffs and marine terraces are widespread and are typically fronted by narrow beaches. Unusual boulder deltas occur in the Santa Barbara area, notably at El Capitan and Rincon (Figure 5), and are thought to be remnant flood deltas at the mouths of steep mountain creeks. The larg_est river in this section in this section is the Santa Clara River with an estimated average annual sand and gravel discharge of 912,000 m3 (Figure 10). Other notable rivers are the Ventura, Los Angeles, and Santa Ana. There are a number of submarine canyons with heads near the lit_toral zone, including: Mugu, Hueneme, Redondo, Dume, Newport, Scripps, and La Jolla canyons (Figure 5). The narrow coastal plains of the Santa Barbara area are replaced by broader plains in Ventura-Oxnard, Santa Monica - Los Angeles Basin, and Mission Bay to Imperial Beach. The dune complexes are not as well developed as those in the rest of the state but moderately large dune fields occur near Oxnard, north of Palos Verdes, and at Silver Strand - Impe_rial beaches (Figure 5). This section is the most urbanized stretch of coast in California. HISTORY OF INFRASTRUCTURE DEVELOPMENT The first European to visit the coast of California is widely held to be Juan Rodriguez Cabrillo, a Portuguese explorer who is credited with the "discovery" of Califor_nia in 1542. The first permanent European settlement was established in what is now San Diego in 1769. Settlement of the coastal areas proceeded slowly in California, partly because of the dangerous nature of the Pacific coastal waters, and partly because access from inland was inhibited by the steep and rugged terrain. Northern California was settled primarily by Russian fur traders, and most coastal development in the State was restricted to large natural har_bors such as San Diego Bay, Monterey Bay, San Francisco Bay and Humboldt Bay. By 1850, the total population of California was only 93,000. Population grew over the years, but there was an explosion following World War II; the State's population increased from 10 to 20 million between 1950 and 1970 (Pincetl, 2004) and in 2005 is about 36 million. Today, California is the most populous state in the union, and it is estimated that 80% of California residents live within 50 km of the coast (Griggs, 1994). Much of the coast of Central and Northern California is very rugged, inaccessible and therefore undeveloped. This results in the focusing of developmental pressures over a smaller percentage of the coast resulting in variations in coastal hazards. Along much of the Northern California coast, the most important coastal hazards are large land_slides that can damage coastal roads, and the rapid retreat of coastal cliffs where community infrastructure exists at the top or base of the cliff. Central California has a mixture of hazards; in addition to large coastal landslides and coastal cliff erosion, there are linear stretches of sandy shoreline that have been developed with homes and infrastructure. Southern California, which has the greatest percent of sandy shorelines also has the greatest percent of coastal armor_ing, engineering structures and nourishment programs. The wide, sandy beaches that exist today in Southern California were created and are maintained through a variety of coastal engineering projects and nourishment programs (Flick, 1993). Practices such as damming coastal rivers and building various coastal engineering structures (groins, jetties and breakwaters) may be adversely affecting beach resources. During the post-World War II building boom, many homes and communities were established on or near the coast, with houses often built on the sand, especially in Southern and parts of Central California. Eventually these homes were threatened by shoreline erosion, and the response was fre_quently to construct some type of protection structure. The California Coastal Act was passed in 1976, and with it the California Coastal Commission was formed. The Coastal Act requires statewide regulation and planning for coastal development, but also allows local governments to imple_ment policies for coastal erosion hazard mitigation. The Coastal Commission has slowed the widespread emplace_ment of shoreline protection structures, but the Coastal Act states that such structures shall be permitted to protect exist_ing development if it is threatened by erosion. The post-World War II rapid increase in population and construction also coincided with a period of relative climatic quiescence on the West Coast. The period from the 1940s through the early 1970s had no major El Ni_o events and average or below average number of damaging coastal storms (Storlazzi and Griggs, 2000). Rapid build_ing took place near the coast during this time because it was considered desirable and not a high-hazard zone. This period also coincided with the development of several major coastal engineering projects in Southern California, which resulted in the addition of large volumes of sand to the beach systems. In the mid-1970s, the West Coast entered into a climatic period when the intensity and number of severe storms substantially increased. The destructive El Ni¤o winters of 1982-83 and 1997-98 are evidence of this stormier period. Widespread damage to both public and private property occurred during those winters. According to Griggs and Fulton-Bennett (1988), total losses during the winter of 1982-83 reached $200 million (in 2006 dollars), and numerous houses, businesses and existing coastal pro_tection structures were damaged. HISTORICAL SHORELINE CHANGE ANALYSIS This section presents the results of the California sandy shoreline (herein referred to as shoreline) change analysis and discusses, where applicable, the effects of engineering structures and beach nourishment programs on the rates of shoreline change. Each California section (Northern, Central and Southern) is subdivided into regions (Figure 1), which are based broadly on littoral cells and breaks in data cover_age. Tables 6A-C summarize both long-term and short-term average rates of shoreline change within each region. Addi_tionally, Tables 7A-C present the maximum and minimum erosion and accretion rates for each region in California. The description of shoreline change includes informa_tion and discussion on human-induced changes. Most of the substantial erosion/accretion trends and/or reversals in trend are related to human intervention within the natural coastal system; these are virtually inseparable topics of discussion. In California, shorelines are eroding primarily because of an increase in storm intensity, sea-level rise, climatic changes, and as a consequence of human activities that disrupt the natural sediment supply. In the discussions below, rates are referenced from Tables 6A-C and 7A-C, where shoreline change rates are presented as the region-average net rate for the long-term (1800s- 1998/2001) and short-term (1950s/70s - 1998/2001) analysis, as well as by the magnitude of the erosion-only and accretion-only rates. Errors and uncertainty values are not shown in the text for clarity; refer to Table 6A-C for these values. To compare how net trends and rates may have changed from the long-term to the short-term, a statistical t-test was performed to determine whether the long-term and short-term rates were significantly different from one another at the 90% confidence interval. The t-test results found that in all regions except the San Pedro region, the change from long-term to short-term was statistically significant. Within the remaining 14 regions, the net shore_line change rate became more erosional from the long- term to the short-term with the exception of the Russian River region. The average net rate of long-term shoreline change for California was 0.2 m/yr, an accretional trend. This is based on shoreline change rates averaged from 14,562 individual transects, of which 40% were eroding. Our analysis found that the only regions in California that experienced long-term negative net shoreline change were in Central Califor_nia (San Francisco South and Monterey Bay regions), both with region-averaged rates of -0.2 m/yr (Table 6B). The highest region-averaged net rate was measured in the San Diego region (0.9 m/yr). Overall, Central California had the lowest overall net long-term shoreline change, likely because of the lack of major coastal engineering projects, such as those that result in more accretional rates in South_ern California by adding sediment to the littoral system. In addition, the high volumes of sediment input from rivers likely contribute to the lower overall erosional trend in Northern California. When the erosion versus accretion rates were separated out, the average long-term erosion rate for the state was found to be -0.2 m/yr. The average net rate of short-term change for Califor_nia was -0.2 m/yr, based on 16,142 transects, along which 66% were eroding. Negative (erosional) net short-term shoreline change was measured in 10 of the 15 regions. For those transects along which erosion was recorded, the aver_age short-term erosion rate was -0.8 m/yr. The short- term average erosion rates were highest in Central California (Table 6B). It is important to keep in mind that the change rates discussed in this report represent change measured through the date that the lidar was collected and thus may not reflect the most recent trends in shoreline change. In addition, although erosion rates in some areas are relatively low, many of California's beaches are narrow and even a small amount of local erosion may present serious hazards to the coastal resources and community infrastructure in a given area. Northern California The Northern California analysis extends from the Ore_gon border to Tomales Bay, a distance of approximately 550 km (Figure 1). For the presentation of the shoreline change analysis Northern California was divided into four regions: Klamath, Eureka, Navarro and Russian River (Figure 14A). Much of Northern California is highly crenulated, rocky coastline with small sections of pocket beaches, except for near major river mouths such as the Klamath, Smith, Eel and Russian Rivers, and a few areas where steep coastal cliffs are fronted by narrow beaches. As a result of this geomorphology, there were many gaps in the data; the long- term change was measured along only 148 km of the shoreline, and short-term change over 168 km. Both long- term (0.5 m/yr) and short-term (0.3 m/yr) net shoreline change rates were accretional when averaged over all of the Northern California transects. Of the 2,966 transects along which long-term shoreline change was measured, 23% had erosional trends, with an average erosion rate of -0.3 m/yr (Table 6A). For the short-term analysis, the percent of beach eroding more than doubles, increasing to 47% and the average short-term erosion rate was -0.6 m/yr. Klamath Region The Klamath region covers approximately 112 km of coastline and extends from the Oregon border to Patrick's Point (Figure 1). This region lies within the Smith and Klamath littoral cells (Figure 9), where rivers of the same names supply abundant sediment to the beach systems. The coastline here is sparsely developed, except for the area around Crescent City, and includes long stretches of State and National Park lands. The only significant engineering structures are the breakwaters protecting the Crescent City Harbor. According to Clayton (1991) there is harbor sand by-passing every several years; however the frequency is not consistent. In addition, the harbor was dredged in the 1970s, and material was placed north of the harbor to attempt to slow chronic bluff erosion (Savoy and others, 2005). For the Klamath region, long-term change rates were measured along 71.5 km of shoreline. The net long-term rate, averaged over 1,430 transects, was 0.7 m/yr. Along those transects with a long-term erosional trend, the average erosion rate was -0.4 m/yr and was found along 25% of the coast. The average long-term accretion rate, which occurred along 75% of the coast in this region, was 1.0 m/yr (Table 6A). The long-term accretion rate in the Klamath region was the highest in Northern California. The maximum long-term erosion rate (-1.2 m/yr) occurred on the shoreline of a dynamic spit that extends across much of the Klamath River mouth (Table 9A). The average short-term net shoreline change rate in the Klamath region is accretional (0.4 m/yr). Forty-eight per_cent of the coast along which short-term shoreline change was measured was erosional, and the average erosion rate was -0.6 m/yr. The remaining 52% of the measured coast in this region had a short-term accretion rate of 1.3 m/yr. The highest short-term erosion rate, -2.6 m/yr, was along Big Lagoon Beach, north of Patrick's Point (Table 9A; Figure 15). This area was heavily impacted during the 1982-83 El Ni_o winter storms (Figure 16), which may have influenced the short-term erosion rate. The rate of net shoreline change in the Klamath region decreased from the long-term (0.7 m/yr) to the short-term (0.4 m/yr), and the percent of the coastline eroding increased from 25% in the long-term to 48% in the short- term (Table 6A). North of the Crescent City Harbor, shoreline change becomes increasingly erosional in both the long- and short-term periods (Figure 15) near the harbor as opposed to areas further north. The highest accretion rates north of Crescent City were located immediately south of the Smith River mouth where there are extensive dune systems. There were local increases in the rates of accretion adjacent to the north and south breakwaters of the Crescent City Harbor (Figure 15). This area is composed of broad tidal flats, and the high rates may have been a function of the tide level when the shoreline data were collected. South of the Klamath River mouth, the magnitude of shoreline change increased substantially. At the northern end of Redwood State Park where Ossagon Creek empties onto the beach at Gold Bluffs Beach, the highest accretion rates in the State were observed (4.8 m/yr long-term and 7.3 m/yr short-term). While most of the high accretion rates in other parts of the State were associated with engineering structures or beach nourishment, the accretion rates here were apparently natural. Eureka Region The Eureka region, which begins 6 km south of Trini_dad Head and extends 74 km south to Cape Mendocino (Figures 14A and 17) falls within the Eureka littoral cell. Most of the coastline consists of sandy beaches as compared with the other Northern California regions. Long, linear beaches, dunes systems and spits have formed through deposition of sand by the Eel, Mad and Little Rivers. While still sparsely developed by California standards, the Eureka region, which includes the towns of Arcata and Eureka, is the most developed and populous coastal area of the North_ern California regions. Humboldt Bay Harbor lies between Eureka and a seaward barrier spit and is the largest harbor in Northern California; jetties were constructed there in the 1800s to keep a channel in the spit open for boat traffic. North Spit and South Spit now converge at the Humboldt Bay jetties; unfortunately there is a gap in the lidar data for the spits, except immediately adjacent to the jetties. There_fore, we were unable to calculate long- or short-term rates for 18 km of sandy shoreline along the spits. The long-term net shoreline change rate for the Eureka region was 0.7 m/yr, an accretional trend similar to that measured for the Klamath region. Virtually all of the shore_line was accreting at a long-term average rate of 0.7 m/yr, observed along 96% of the measured shoreline. The average long-term erosion rate for the Eureka region is - 0.2 m/yr (Table 6A). Of the total 24.7 km of sandy shoreline that was measured, long-term erosion occurred along only 4% of the coast. The highest erosion rates were measured on either side of the Eel River mouth, where a maximum long-term rate of -0.4 m/yr, was observed (Table 9A). Short-term net average shoreline change rates for the Eureka region, measured along 32.6 km of coastline, were 0.4 m/yr, a less accretional trend from the long-term rates. The average short-term erosion rate was -0.9 m/yr (Table 6A) and was measured along 51% of the analyzed coast. The average short-term accretion rate, 1.8 m/yr, was the highest average accretion rate in the State, and was mea_sured along 49% of the analyzed coast. Short-term change rates varied along coast, and were predominantly erosional near the Mad River, the North Spit of Humboldt Bay, and south of the Eel River. The maxi_mum erosion rate (-2.7 m/yr) in this region was along North Spit Beach, immediately north of the Humboldt Bay jetty (Figure 18). The highest accretion rates (both long- and short-term) occurred in the northern portion of the region, on the south side of the mouth of the Little River (Figure 17; Table 9A), along Little River State Beach. The beach here is backed by a substantial dune system. Similarly, in the southern part of the region, within the Eel River State Wildlife area, there was a strong accretional trend in both long- and short-term rates in an area that is backed by substantial dunes. Navarro Region The Navarro region extends along 207 km of coast_line and contains both the Ten Mile and Navarro littoral cells (Figure 9). This section begins approximately 11 km south of Point Delgada and ends at Point Arena (Figures 14A and 19). The towns of Fort Bragg and Mendocino are located within the Navarro region; otherwise, this stretch of coastline is very rugged, inaccessible, and there is little development. The only major coastal engineering structure along this coast is the breakwater at the Noya Harbor, on the south side of Fort Bragg. With a few exceptions, the coast in the Navarro region is crenulated and rocky with steep cliffs; there are some scattered pocket beaches and occa_sional narrow beaches fronting the cliffs that generally are not passable at high tide. Exceptions include the extensive beach and dune system south of the Ten Mile River mouth (Mac Kerricher State Park) (Figure 20), and several beaches formed in the vicinity of larger creek mouths, such as West_port-Union Landing State Beach, and Manchester Beach State Park (Figure 14A). Of the 207 km of coastline in this region, only 31.5 km had measurable sandy shorelines for our long-term analysis, due primarily to the lack of continuous beaches. The net long- term shoreline change was accretion that averaged 0.1 m/yr. This rate was much lower than the average rates for the Eureka and Klamath regions discussed above. Along those transects where erosion was measured, the average long-term erosion rate was -0.1 m/yr, averaged along 28% of the coast (Table 6A). Long-term accretion, measured along 72% of the coast, averaged 0.2 m/yr. The maximum long-term erosion rate (-0.6 m/yr) was located within DeHaven Creek Beach north of Fort Bragg (Table 9A). Long-term accretion rates reached a maximum of 0.7 m/yr along Ten Mile Beach south of the Ten Mile Rive mouth. Net short-term shoreline change, as averaged along 32.8 km of coastline, was found to be undetectable at the significant figures appropriate for this analysis and therefore are reported as 0.0 m/yr in Table 6A. Of the measurable stretches of sandy shoreline, 50% eroded and 50% accreted. The average short-term erosion rate was -0.5 m/yr, and the accretion rate was 0.6 m/yr. The maximum short-term erosion rate of -1.4 m/yr occurred at DeHaven Creek Beach (Table 9A), and the maximum short-term accretion rate of 3.3 m/y occurred just north of the Ten Mile River mouth (Figure 19). The maximum long- and short-term erosion rates in the Navarro region are both found along isolated narrow beaches in the northern half of the region. These beaches are difficult to access and no development is threatened as a result of the higher erosion rates. Both long- and short- term maximum accretion rates were near the Ten Mile River mouth north of Fort Bragg. The high long- and short-term accretion rates were likely related to the large volumes of sand discharged by the Ten Mile River (Merritts and others, 2005). Russian River Region The Russian River region begins 12 km south of Point Arena along a remote, rocky stretch of coastline that has little development, and extends 155 km south to Tomales Point (Figures 14A and 21). Similar to the other regions in Northern California, there are few linear stretches of sandy shoreline, especially in the northern half of the region. The most extensive beaches are formed near large rivers such as the Gualala and the Russian Rivers (Figure 21). In addition, a wide sandy beach and dune system exist at Salmon Creek Beach just north of Bodega Head. The Russian River region contains both the Russian River and Bodega Bay littoral cells. Of the 155 km of coast in this region, we were only able to measure long-term shoreline change along 21.8 km, primarily because of the lack of linear beaches; we did not measure shoreline change rates on pocket beaches smaller than 0.5 km in length. The average net long-term shore_line change rate for the Russian River region was 0.2 m/yr. Average long-term erosion was -0.2 m/yr and occurred over only 28% of the coast (Table 6A). Similar to the other regions in Northern California, a much higher percentage of the coast is accreting in the long-term than eroding. The average long-term accretion rate for the Russian River region was 0.3 m/yr and was observed along 72% of the coast. The maximum long-term accretion rate, 1.8 m/yr, was at Dillon Beach, and the maximum long-term erosion rate of -0.7 m/yr was along Sonoma Coast State Beach (Figure 22; Table 9A). Net short-term shoreline change averaged 0.4 m/yr, measured along 24.1 km of the coast. Short-term erosion occurred along 35% of the coast, and the average rate was -0.4 m/yr (Table 6A). This is the lowest percentage of eroding coastline of the four Northern California regions. The average short-term accretion rate was 0.8 m/yr. The maximum short-term accretion rate (3.5 m/yr) was at Dillon Beach (Figure 22; Table 9A) and the highest short- term erosion rate (-1.6 m/yr) occurred along Sonoma Coast State Beach ~1.5 km south of the Russian River mouth.. Sandy shorelines in the northern 100 km of this region are rare and occur only where small pocket beaches form at the mouths of rivers and creeks. Overall the shoreline change was accretional; the highest erosion rates were near the Russian River. This is the only region in the state in which the average rate of net shoreline change increased from the long-term to the short-term. Central California The Central California section begins approximately 5 km south of Tomales Point in Point Reyes National Sea_shore (PRNS) and extends south to El Capitan State Beach, just north of Santa Barbara, a total distance of approxi_mately 740 km (Figure 14B). Central California is divided into six regions including San Francisco North, San Fran_cisco South, Monterey Bay, Big Sur, Morro Bay and Santa Barbara North. The average net long-term shoreline change rate for Central California was found to be undetectable at the significant figures appropriate for this analysis and is reported as 0.0 m/yr. In the short-term, however, the aver_age net rate is strongly erosional (-0.5 m/yr). There are many gaps in our analysis along this coast, as much of the shoreline is rocky with isolated pocket beaches; there are a few continuous linear beaches such as in the Monterey Bay region. Coastal engineering structures and nourishment projects are limited to small harbor construc_tion (e.g. Port San Luis, Santa Cruz) and some harbor by-passing. Numerous seawalls and revetment exist along the coast but these are primarily related to issues of coastal cliff erosion mitigation, not to protect structures from erosion of the sandy shoreline. San Francisco North Region The San Francisco North region is 93 km long and includes the Point Reyes, Drakes Bay and Bolinas Bay litto_ral cells (Figure 9). This is primarily a rocky coastline, with narrow beaches backed by high coastal cliffs, small isolated pocket beaches between rocky headlands, and an expansive dune field at Point Reyes. There are two very small, devel_oped sections of the coast in this region at Bolinas and Stin_son Beach. Due to a data gap in the 1800s-era t-sheets we have no long-term shoreline change rates for either of these areas. Other than these two areas, the coast here is undevel_oped and remote, and falls entirely within either PRNS or the Golden Gate National Recreation Area (GGNRA). Net long-term shoreline change rates in this region, averaged along 45.1 km of coast, were low and averaged 0.1 m/yr (Table 6B). Forty-six percent of the shoreline was eroding in the long-term at an average rate of -0.2 m/yr. Long-term accretion rates of 0.3 m/yr occurred along 54% of the coast. The highest and lowest long-term rates occurred north of Point Reyes Headland where the long-term trend was largely accretional (Figure 23). However, the maximum long-term erosion rate for the San Francisco North region of -0.5 m/yr was at Limantour Beach, which is north of Point Reyes (Table 9B). Net average short-term shoreline change rates were measured along 51.9 km of coastline, and the average change rate was -0.5 m/yr. Eighty-one percent of the coast was eroding (short-term) with the rate of erosion averaging -0.7 m/yr. Short-term accretion rates, averaging 0.5 m/yr, occurred along only 19% of the coast. Short-term shoreline change trends north of Point Reyes Headland (Figure 23) were highly variable with accretion dominant in the north, changing to predominantly erosion in the south. The maximum short-term erosion rate of -3.1 m/yr was measured at Point Reyes Beach (Figure 23). This trend is driven by the position of the most recent (lidar) shoreline and may indicate a rotation of the beach during the 1997-98 El Nino where the dominant littoral transport changed directions from southward to northward in many local areas. In the San Francisco North region, the long- and short-term shoreline change rates were significantly differ_ent; there was an overall shift from a net shoreline change trend that was 0.1 m/yr in the long-term to a net shoreline change rate that was strongly erosional (-0.5 m/yr) in the short- term. In addition, the percent of coastline along which erosion was measured increased from 46% in the long-term to 81% of the short-term. As demonstrated in Figure 23, from Point Reyes headland to Point Bonita, the sandy beaches were relatively stable in the long-term. In the short-term, this section of beach as primarily erosional, with a few localized excep_tions (i.e south of Drakes Estero). The short-term erosion rates at Stinson Beach were the highest of those measured in the southern portion of the San Francisco North region. Winter storm waves frequently inundate Stinson Beach, moving large volumes of sand, and threatening homes built on the sand spit. Figure 24 shows a house buried by sand that had been eroded from a location further north. Riprap has been emplaced in many areas to protect the houses from beach erosion (Savoy and others, 2005). San Francisco South Region The San Francisco South region is 115 km long and extends from the mouth of San Francisco Bay to Davenport (Figures 14B and 25). The northern coast in this region is urban and includes San Francisco, Pacifica and Half Moon Bay; the southern half is largely undeveloped and agricultural. The San Francisco littoral cell is within this region (Figure 9). The geomorphology of the coastline is variable, with linear beaches backed by dunes, steep cliffs with narrow fronting beaches, rocky coast with small pocket beaches, and steep, high-relief coast with no sandy shore_line. There are no known beach nourishment projects in this region, although a dredge spoil deposit offshore may be contributing material to the beach (Barnard and Hanes, 2006). Additionally, wind-blown sand is regularly removed from the inland side of the dunes and from the adjacent highway and added to the south end of the beach (Wiegel, 2002). The most notable coastal engineering structures are the O'Shaughnessy and Great Highway seawalls along Ocean Beach in San Francisco and the Pillar Point Harbor at Half Moon Bay. For this analysis we calculated net average long-term shoreline change rates for 56.3 km of coastline, and the average rate was -0.2 m/yr. Long-term erosion occurred along 76% of the coast at an average long-term rate of -0.4 m/yr. For the 24% of the coast along which accretion occurred, the long-term average accretion rate was 0.1 m/yr. The maximum long-term accretion rate (0.4 m/yr) in this region was located 0.25 km south of Mussel Rock, and the maximum long-term erosion rate of -1.8 m/yr occurred on the north side of Point A¤o Nuevo (Figure 25; Table 9B). Net short-term shoreline change, with an average rate of - 0.5 m/yr, was measured along 57.5 km of coast in the San Francisco South region. Along the portions of coast where the short-term shoreline change was erosional, the rate was -0.7 m/yr, averaged over 81% of the coast. The average short-term accretion rate was 0.5 m/yr. Short-term change trends in the central portion of this region are more variable than the long-term trends; erosion was relatively high north and south of Pillar Point Harbor (Figu