Marsh Shorelines of the Massachusetts Coast from 2013-14 Topographic Lidar Data

The mean tide level (MTL) for each marsh was found using VDatum (version 3.7). MTL varies from -0.08 to 0.08 m (above NAVD88) across the state. I used 0 m for all marshes. Mean High Water was determined for each marsh from Vdatum and Weber and others (2005). A single value was used for each marsh.

The processing steps outlined here were conducted for each marsh area separately. All marshes were chosen because there was a previous shoreline from the Massachusetts Shoreline Change Project update (Smith and others, 2013). Each marsh is geographically separated from the other marshes, so each was processed on its own. The goal of this first step is to get a mean tide level (MTL) contour to use as a baseline in the next step. The topographic lidar data is first downloaded from NOAA then loaded into ArcMap (version 10.4.1), gridded and contoured. This is done by the following steps:
A) The LAS (x,y,z) data files were downloaded from NOAA in UTM, zone 19 (NAD83, NAVD88).
B) The LAS files loaded into ArcMap: ArcToolbox >> Data Management >> LAS Dataset >> Create LAS Dataset.
C) A DEM was created: ArcToolbox >> Conversion Tools >> To Raster >> LAS Dataset to Raster.
D) The MTL was then contoured: ArcToolbox >> 3D Analyst Tools >> Raster Surface >> Contour.
E) This contour was smoothed: ArcTools >> Cartography Tools >> Generalization >> Smooth Line, using PAEK smoothing at 100 m. The amount of smoothing was occasionally adjusted to improve the results.
F) This line will be used as a baseline, so sections of the line in areas where we did not need to find a shoreline were removed (using the Arc editor). Where appropriate, disconnected sections of the line were joined with the Arc editor.
G) The final lines were converted to points: ArcToolbox >> X Tools Pro >> Feature Conversions >> Convert Features to points and ArcToolbox >> Data Management Tools >> Features >> Add XY.
F) The MTL contour was exported as a CSV file.
This line is the baseline that will be used in the next step.

All the work in this process step is done separately on each marsh area. In this step the baseline from the previous step is loaded into Matlab (version 2017a) with the elevation data. The baseline is used to define transects and the elevation along each transect is found from the elevation data. The max slope along each transect is found and defined to be the marsh edge. This is done by the following steps:
A) The same LAS files from the previous step were loaded into Matlab.
B) A DEM was made from them (using the function griddata).
C) The baseline made in the previous step was also loaded into Matlab.
D) Transects are made: Equally spaced points (at 5m intervals) were found along this baseline. At each point a line perpendicular to the baseline was calculated (this is called a transect). The line extends out from both sides of the baseline. The length of each transect varied but was usually about 20-40m long.
E) The elevation along each transect was extracted from the DEM (using the interp2 function).
F) The marsh edge was defined to be the maximum slope along each transect between mean high water and (mean tide level - 0.5 m).
Thus for each transect, we have an estimate of the shoreline position along that transect.

All the work in this process step is done separately on each marsh area. In Matlab (version 2017a), the marsh edge found in the previous step was smoothed using a Gaussian weighted moving average with a 25 m window. Spurious points from the marsh line were removed using an interactive GUI in which the LAS data were plotted as dots color-coded by height with the marsh shoreline plotted on top. To aide decision making, the marsh edge was also displayed in Google Earth. It was helpful to look at images from different stages of the tidal cycle. Once this quality control check was done, the data were exported to a new shapefile (using the function shapewrite in the Mapping Toolbox, version 4.5.1).

After the above steps were completed for each marsh area, the shapefiles for all regions were loaded into ArcMap (version 10.4.1) and plotted on NAIP imagery. Sections on sandy beaches were removed. There are many fringing marshes in the state. In these areas, both the mean high water (MHW) shoreline and the marsh edge were kept, because both have value. Since it is sometimes difficult in imagery to distinguish sand, mud and brown marsh grass, and since we did not want to delete valid marsh sections, when in doubt, we kept the marsh shoreline even if it might be on a small sandy area. A new attribute called “MarhType” was added. It was set to “platform” if the marsh looked like a platform marsh. It was set to “fringing” if the marsh looked like a fringing marsh. It was set to “unknown” if it was not clear if the area was really a marsh. Note that all choices were subjective and based only on visual interpretation of the imagery. All the individual shapefiles were joined into one shapefile for the whole state using Merge (ArcToolbox >> Data Management toolbox >> General >> Merge). The data were transformed from UTM zone 19 (NAD83) to geographic (WGS84) using ArcToolbox>>Data Management tools>>Projections and transformations>>Project. The software used the transromation WGS_1984_(ITRF00)_To_NAD_1983.

Added keywords section with USGS persistent identifier as theme keyword.