Shorelines of the Florida west (FLwest) coastal region used in shoreline change analysis

An operational Mean High Water (MHW) shoreline was extracted from the lidar surveys within MATLAB v7.6 using a method similar to the one developed by Stockdon et al. (2002). Shorelines were extracted from cross-shore profiles which consist of bands of lidar data 2 m wide in the alongshore direction and spaced every 20 m along the coast. For each profile, the seaward sloping foreshore points were identified and a linear regression was fit through them. The regression was evaluated at the operational MHW elevation to yield the cross-shore position of the MHW shoreline. If the MHW elevation was obscured by water points, or if a data gap was present at MHW, the linear regression was simply extrapolated to the operational MHW elevation. A lidar positional uncertainty associated with this point was also computed. The horizontal offset between the datum-based lidar MHW shoreline and the proxy-based historical shorelines nearly always acts in one direction and the "bias" value was computed at each profile (Ruggiero and List, 2009). In addition an uncertainty associated with the bias was also computed, which can also be thought of as the uncertainty of the HWL shorelines due to water level fluctuations. Repeating this procedure at successive profiles generated a series of X,Y points that contain a lidar positional uncertainty, a bias, and a bias uncertainty value. Ruggiero, P. and List, J.H., 2009. Improving Accuracy and Statistical Reliability of Shoreline Position and Change Rate Estimates. Journal of Coastal Research: v.25, n.5, pp.1069-1081. Stockdon, H.F., Sallenger, A.H., List, J.H., and Holman, R.A., 2002. Estimation of Shoreline Position and Change using Airborne Topographic Lidar Data: Journal of Coastal Research, v.18, n.3, pp.502-513.

The lidar data were collected in projected coordinates (WGS 84 UTM zone 17N). The series of operational MHW points extracted from the cross-shore lidar profiles were converted into a calibrated route shapefile for use in ArcGIS using a Python script. The script generates a point shapefile, converts it to a polyline-M file, saves the uncertainty information in an accessory dBase (.dbf) file and finally generates a calibrated route for the newly-created polyline-M file. Calibration is based on the unique and sequential profile ID value provided with the point data and stored as the M-value. This value is also stored as an attribute in the uncertainty .dbf file and is used as the common attribute field linking the shoreline route file (FLwest_shoreline.shp) to the uncertainty table (FLwest_shoreline_uncertainty.dbf) storing the lidar positional uncertainty, the bias correction value, and the uncertainty of the bias correction for each point of the original lidar data. During the rate calculation process DSAS uses linear referencing to retrieve the uncertainty and bias values stored in the associated table. Visually identified HWL-type proxy shorelines are virtually never coincident with datum-based MHW-type shorelines. In fact, HWL shorelines are almost universally estimated to be higher (landward) on the beach profile than MHW shorelines. Not accounting for this offset will cause shoreline change rates to be biased toward slower shoreline retreat, progradation rather than retreat, or faster progradation than in reality (for the typical case where datum-based shorelines are more recent data than the proxy-based shoreline dates), depending on actual changes at a given site. Ruggiero, Peter, and List, J.H., 2009, Improving accuracy and statistical reliability of shoreline position and change rate estimates: Journal of Coastal Research, v. 25, no. 5, p. 1069–1081. Ruggiero, Peter, Kaminsky, G.M., and Gelfenbaum, Guy, 2003, Linking proxy-based and datum-based shorelines on high-energy coastlines—Implications for shoreline change analyses: Journal of Coastal Research, special issue 38, p. 57–82. For a detailed explanation of the method used to convert the lidar shoreline to a route, please refer to "Appendix 2- A case study of complex shoreline data" in the DSAS user guide: Himmelstoss, E.A. 2009. "DSAS 4.0 Installation Instructions and User Guide" in: Thieler, E.R., Himmelstoss, E.A., Zichichi, J.L., and Ergul, Ayhan. 2009. Digital Shoreline Analysis System (DSAS) version 4.0 - An ArcGIS extension for calculating shoreline change: U.S. Geological Survey Open-File Report 2008-1278. https://woodshole.er.usgs.gov/project-pages/DSAS/version4/images/pdf/DSASv4_3.pdf This process step and all subsequent process steps, unless otherwise noted, were performed by the same person - E.A. Himmelstoss.

Data from the previously-published National Assessment of Shoreline Change study (USGS Open-File Reports 2004-1043 and 2004-1089) were used as the starting point for the project’s continued efforts to compile as many quality shorelines as possible for the region. Historical shorelines from the original shoreline assessment publication were provided as individual shapefiles in the report. These data were merged in Esri's ArcToolbox (v.10), Data Management Tools > General > Merge.

Additional digitized shorelines were downloaded in vector format from the NOAA Historical Shoreline Survey Viewer, currently available as a Google Earth KMZ (https://nosimagery.noaa.gov/images/shoreline_surveys/noaa_shoreline_surveys.kmz). The extracted shorelines are digitized vectors of the mean high water (MHW) position derived from the scanned and georeferenced T-sheet raster imagery that NOAA also manages. Although NOAA labels historical shorelines as MHW, they are proxy-based and therefore considered HWL for the purposes of shoreline change rate calculation with the proxy-datum bias correction. The modern lidar shoreline used in the analysis is datum-based, not proxy-based, and designated as operational MHW which reflects this difference. Shoreline data from NOAA were used with preference at any location where the shoreline date had previously been sourced from FLDEP. Both sources extracted the shoreline from the same T-sheet, but the NOAA data have fewer process steps and therefore a reduced likelihood of additional error. Topographic survey sheets (T-sheets) and shorelines were quality checked and edited where necessary. Edgematching was corrected when discovered.

Shorelines were acquired from the Florida Department of Environmental Protection. Some of these were sourced from T-sheet data (not available from NOAA) while other shorelines were derived from georeferenced vertical aerial photographs. The shoreline data are organized by county and provided for download as CAD files (.dwg) (ftp://ftp.dep.state.fl.us/pub/water/beaches/HSSD/drawings/draw8390/). These files could be viewed within ArcMap (ArcGIS v.10 SP1 used in all parts of this process step), but first the coordinate system for the file had to be assigned (NAD83 State Plane Feet) in ArcCatalog. All shoreline data compiled from FLDEP were projected in Esri's ArcToolbox (v.10) > Data Management Tools > Projections and Transformations > Feature > Project. Parameters: input projection = NAD 1983 State Plane Florida West FIPS 0902 (Feet); output projection = NAD 1983 State Plane Florida West FIPS 0902 (Meters); transformation = none. There are three State Plane FIPS zones for Florida (Florida East FIPS 0901, which covers the entire Atlantic coast along with Collier and Monroe counties on the southwestern coast of the peninsula; Florida West FIPS 0902, which covers the Gulf Coast section of the peninsula between Lee county to the south and Levy county to the north; Florida North FIPS 0903, which covers the panhandle section of the Florida Gulf Coast from Dixie county in the east to the FL/AL border in Escambia county to the west) and the appropriate zone was assigned to the data based on location. The .dwg files consist of several layers. After the individual county .dwg files were added to the ArcMap document, the polyline layer was selected from each county file and exported to a shapefile (right-click > export) from the ArcMap table of contents panel. The exported polyline shapefile for each county was added to the ArcMap document, an edit session was started, and all non-shoreline features were deleted from the file using the “Layer” attribute field. Non-shoreline features were determined to be any entry in the “Layer” attribute field that did not begin with a date. This removed most but not all of the annotation. The “Entity” field was then used to query non-shoreline features. In all files, features that had an Entity attribute field coded as “Circle” were deleted, in some files both “Circle” and “Line” were deleted (in some files the “Line” code contained shoreline position data). In all files the “Entity” field coded as “Arc” contained shoreline position data. After all annotation had been removed from the shapefile, a check for duplicate features was run using XTools (v7.1.0). All duplicate features were deleted. The Dissolve tool was used (ArcToolbox > Data Management > Generalization > Dissolve) to dissolve each shapefile based on the “Layer” field. The box next to “Unsplit lines (optional)” at the bottom was selected. A final visual check of the shorelines was performed. In many of the files it was concluded that edge matching was likely performed between adjacent T-sheets due to noticeable offsets in the shoreline. There is no way to differentiate these line segments from legitimate shoreline vectors using the attribute fields, but any that were observed during the visual inspection were deleted from the file. Close attention was paid to the overlapping areas between adjacent county shoreline files. There were several instances where inconsistent overlap of shoreline from the same date and source occurred, which required manual editing of the two county shapefiles including splitting of some shorelines at the location where the overlap began and deleting the duplicate/overlapping segments from one file. No edgematching was performed during this process. The DSAS-required attribute fields were added to the shapefiles and populated. Date or year and the source of shoreline data (T-sheet or aerial photograph) information was verified by reviewing the details summarized in the bibliography files provided by FLDEP for each county (ftp://ftp.dep.state.fl.us/pub/water/beaches/HSSD/sourcebibs/).

If no digital shoreline vectors were available from NOAA or FLDEP, the original T-sheet scan and world file were downloaded from the NOAA Google Earth viewer. Vector shorelines were digitized from the georeferenced T-sheets using standard editing tools in ArcMap v10. Quality assessments were performed and shorelines were edited to remove any overlap between adjacent T-sheets from the same time period. No edge-matching between neighboring shorelines was attempted. The DSAS-required attribute fields were added to the shapefiles and populated.

Historical shorelines were merged in Esri's ArcToolbox (v.10), Data Management Tools > General > Merge. Then the merged file was projected in Esri's ArcToolbox (v.10) > Data Management Tools > Projections and Transformations > Feature > Project. Parameters: input projection = geographic (NAD 83); output projection = NAD 1983 State Plane Florida West FIPS 0902 (Meters); transformation = none.

The lidar shoreline calibrated route was projected in Esri's ArcToolbox (v.10) > Data Management Tools > Projections and Transformations > Feature > Project. Parameters: input projection = UTM zone 17N (WGS 84); output projection = NAD 1983 State Plane Florida West FIPS 0902 (Meters); transformation = NAD_1983_To_WGS_1984_1.

Historic shorelines were appended to the lidar route data in ArcToolbox (v.10) > Data Management Tools > General > Append to produce a single shoreline file for the region. The final shoreline dataset was coded with attribute fields (Date, Uncertainty (Uncy), Source, Source_b, Year_, Default_D, Location). These fields are required for the Digital Shoreline Analysis System (DSAS), which was used to calculate shoreline change rates.

Further edits were made to the shorelines file prior to rate calculation. The shorelines file was projected in Esri's ArcToolbox (v.10.2) > Data Management Tools > Projections and Transformations > Project. Parameters: input projection = NAD 1983 State Plane Florida West FIPS 0902 (Meters); output projection = UTM zone 17N (WGS 84); transformation = NAD_1983_To_WGS_1984_1. This process step and all subsequent process steps, unless otherwise noted, were performed by the same person - M.G. Kratzmann.

The appended shoreline file and the shoreline uncertainty table (.dbf) were imported into a personal geodatabase in ArcCatalog v.10.2 by right-clicking on the geodatabase > Import (feature class for shoreline file and table for uncertainty table) for use with the Digital Shoreline Analysis System (DSAS) v4.3 software to perform rate calculations.

The shoreline feature class was exported from the personal geodatabase back to a shapefile in ArcCatalog v.10.2 by right-clicking on the shoreline file > Export > To Shapefile (single) for publication purposes.

The data were projected in ArcToolbox v.10.2 > Data Management Tools > Projections and Transformations > Project. Parameters: input projection = UTM zone 17N (WGS 84); output projection = geographic coordinates (WGS 84); transformation = none.

Keywords section of metadata optimized for discovery in USGS Coastal and Marine Geology Data Catalog.

Added keywords section with USGS persistent identifier as theme keyword.