Slope Values Across Marsh-Forest Boundary in Chesapeake Bay Region, USA

This process step and all subsequent process steps were performed by the same person, Grace Molino, in ArcMap (ver. 10.6.2) using tools from ArcToolbox, unless otherwise stated. For complex operations, names of specific tools used are given in CAPITAL letters (any critical parameters used are given in parentheses, separated by a semicolon, immediately after the tool name). The input and output file names are provided in [square brackets] when necessary. Units for length and area calculations are meters (m) and square meters (m2) unless otherwise stated
Create subdomain tiles.The individual datasets are too large for analysis as a whole so all analyses are completed within subdomains (subregions 0-11) created in this processing step.
a) CREATE FISHNET (Template Extent=Chesapeake_Bay_Topobathy_DEM_1m_v2.tif; Number of Rows=6; Number of Columns=2; Geometry Type=polygon) To define tiles used in the analysis [BayWide_fishnet.shp].
b) PROJECT [BayWide_fishnet.shp] from NAD 1983 UTM Zone 18N to USA Contiguous Albers Equal Area Conic USGS version [BayWide_fishnet_Albers.shp].

Process Maryland and Virginia National Wetland Inventory datasets to obtain the desired wetland category for this study and then combine them for a complete file of estuarine emergent wetlands in the Chesapeake Bay region.
a) EXPORT features from NWI datasets after SELECT ("ATTRIBUTE" LIKE 'E2%EM%' OR "ATTRIBUTE" LIKE 'E2EM%' OR "ATTRIBUTE" LIKE 'E2AB3%' OR "ATTRIBUTE" LIKE 'E2SS%' OR "ATTRIBUTE" LIKE 'E2US4%') to select estuarine intertidal areas of 1) Emergent, 2) Scrub-shrub, 3) Rooted vascular aquatic bed, 4) Organic unconsolidated shore classes [MD_NWI_E2wetlands.shp and VA_NWI_E2wetlands.shp].
b) MERGE the state wetland files [MD_NWI_E2wetlands.shp and VA_NWI_E2wetlands.shp] to create [MDVA_NWI_E2wetlands.shp]. This will be used to erase any overlap with the forest features from Land Use and Land Cover datasets.
c) EXPORT features from NWI datasets after SELECT ("ATTRIBUTE" LIKE 'E2EM%') to select estuarine intertidal emergent areas [MDVA_NWI_E2EMwetlands.shp]. This is a subset of features in step (a) and will be used to establish the marsh-forest boundary.
d) PROJECT [MDVA_NWI_E2wetlands.shp] and [MDVA_NWI_E2EMwetlands.shp] from NAD 1983 Albers to NAD 1983 UTM Zone 18N [MDVA_NWI_E2wetlandsUTM.shp] and [MDVA_NWI_E2EMwetlandsUTM.shp].

Prepare Chesapeake Conservancy BayWide Land-Use dataset. The Land-Use dataset is a raster file with 19 different land use categories. We are only interested in the "forest" category and need the corresponding forest extent as polygons to analyze alongside the salt marsh polygons.
The following process was undertaken for each of the twelve subregions in the [BayWide_fishnet_Albers.shp]. This example is written for subregion 3 which means the process steps are for the Land-Use file created in step a) for region 3 [BayWide_1m_LU3.tif]. These are the same steps taken for all the subregions.
a) SPLIT (Input=BayWide_1m_LU.tif; Split Features=BayWide_fishnet.shp; Split Field=Zone) into twelve subregions. One example of the result which is used in the following process steps: [BayWide_1m_LU3.tif].
b) For each region, EXTRACT BY ATTRIBUTES (Input raster=BayWide_1m_LU3.tif; Where Clause=“Value”=8) to create [BayWide_1m_LU3_forest.tif]. Value 8 is forest.
c) Convert RASTER TO POLYGON (Input=BayWide_1m_LU3_forest.tif; Field=Value; Simplify polygons=checked) [BayWide_1m_LU3_forestpoly.shp].
d) PROJECT from USA Contiguous Albers Equal Area Conic USGS version to NAD 1983 UTM Zone 18N [BayWide_1m_LU3_forestpolyUTM.shp]. Then REPAIR GEOMETRY.

Prepare Chesapeake Conservancy BayWide Land-Cover dataset. The Land-Cover dataset is a raster file with 6 different land use categories. We are only interested in the "Forest Canopy and Shrubland" category and need the corresponding forest extent as polygons to analyze alongside the salt marsh polygons.
The following process was undertaken for each of the twelve subregions in the [BayWide_fishnet_Albers.shp]. This example is written for subregion 3 which means the process steps are for the Land-Cover file created in step a) for region 3 [Baywide_1m_LC3.tif]. These are the same steps taken for all the subregions.
a) SPLIT (Input=Baywide_LandCover_UVM_CC_WorldView.tif; Split Features=BayWide_fishnet_Albers.shp; Split Field=Zone) into twelve subregions. One example of the result which is used in the following process steps:[BayWide_1m_LC3.tif].
b) For each region, EXTRACT BY ATTRIBUTES (Input raster= Baywide_LandCover_UVM_CC_WorldView_3.tif; Where Clause=“Value”=2) to create [BayWide_1m_LC3_forestandshrubs.tif]. Value 2 is "Tree Canopy and Shrubland".
c) Convert RASTER TO POLYGON (Input=[BayWide_1m_LC3_forestandshrubs.tif]; Field=Value; Simplify polygons=checked) to [BayWide_1m_LC3_forestpoly.shp].
d) PROJECT from USA Contiguous Albers Equal Area Conic USGS version to NAD 1983 UTM Zone 18N [BayWide_1m_LC3_forestpolyUTM.shp]. Then REPAIR GEOMETRY.

Obtain forest polygons which border marsh.
a) MERGE [BayWide_1m_LC3_forestpolyUTM.shp] and [BayWide_1m_LU3_forestpolyUTM.shp] datasets [LULC3_forest.shp].
b) DISSOLVE (Input=LULC3_forest.shp; Create multipart features=unchecked) to combine overlapping features [LULC3_forest_diss.shp]. Then REPAIR GEOMETRY.
c) ELIMINATE POLYGON PARTS (Input= [LULC3_forest_diss.shp]; Condition=percent; Part area percent=99; Part option=contained only) to remove any interior holes in the forest polygons [LULC3_forest_d_nh.shp].
d) ERASE any overlap with [MDVA_NWI_E2wetlands.shp] which will remove any area considered generally to be wetlands [LULC3_forest_dnh_erase.shp]. Then DISSOLVE (Create multipart features=unchecked) again to itemize any small forest patches created in the previous step [LULC3_forest_raw.shp].
e) Remove any polygons that are too small (smaller than 900 square meters) for target resolution. ADD FIELD to [LULC3_forest_raw.shp] attribute table called “Area” and then CALCULATE GEOMETRY of each polygon from within attribute table. EXPORT forest polygons from [LULC3_forest_raw.shp] which are over 900m2 [LULC3_forest900m.shp].
f) SELECT BY LOCATION forest polygons from [LULC3_forest900m.shp] which are within 10m search radius of [MDVA_NWI_E2EMwetlands.shp] marsh polygons [LULC3_forest900m_w10m.shp]. These are the forest polygons which border salt marsh.

Prepare forest polygons. The following processing steps are needed to simplify the forest polygon edges and eliminate odd geometries which resulted from preceding processing steps such as the conversion of the forest files from raster to polygon.
a) ELIMINATE POLYGON PARTS (Condition=percent; Part area percent=99; Part option=contained only) [LULC3_forest900m_w10m.shp] to remove any interior holes in the forest polygons [LULC3_forest900m_nh.shp].
b) BUFFER (Input=[LULC3_forest900m_nh.shp], Linear unit=5m) to create [LULC3_forest900m_out5m.shp] and then again BUFFER (Input=[LULC3_forest900m_out5m.shp], Linear unit= -5m) to remove unwanted edge features created during previous merge [LULC3_forest900m_in5m.shp].
c) SMOOTH POLYGON (Smoothing algorithm=PAEK; Smoothing tolerance=30 meters) to get the smoothed forest polygons [LULC3_forest900m_sm30m.shp].
d) BUFFER (Input=[LULC3_forest900m_sm30m.shp], Linear unit=-10m, Dissolve type=NONE) to create [LULC3_forest900m_in10m.shp]. Then SMOOTH POLYGON (Smoothing algorithm=PAEK; Smoothing tolerance=20 meters) to remove sharp edges which complicate analysis [LULC3_forest900m_sm20m.shp].
e) Then DISSOLVE (Create multipart features=unchecked) again to separate any small forest patches created in the previous steps [LULC3_forest900m_sm20m_diss.shp].
f) ADD FIELD to [LULC3_forest900m_sm20m_diss.shp] attribute table called “Area” and then CALCULATE GEOMETRY of remaining polygon from within attribute table. EXPORT forest polygons from [LULC3_forest900m_sm20m_diss.shp] which are over 900m2 [LULC3_forestnew900m.shp].
g) SIMPLIFY POLYGON (Simplification algorithm=point remove; Tolerance=0.5m) to reduce file size [LULC3_forestnew900m_simp.shp].
h) ELIMINATE POLYGON PARTS (Condition=percent; Part area percent=99; Part option=contained only) to remove any interior holes in the forest polygons [LULC3_forestnew900m_s_nh.shp].

Create input files for Digital Shoreline Analysis System (DSAS). DSAS was designed to estimate shoreline change over time by casting transects to a stable baseline. We repurposed this tool to cast transects 30 meters apart and perpendicular to the marsh-forest boundary by creating artificial shoreline and baseline files which are based on the forest polygons within 10m of salt marshes.The shoreline and baseline files do not represent the actual extent of the marsh or forest, they are only intended as an approximation of the marsh-forest transition zone so we may estimate the slope across this area. They are 20m apart which provides a buffer for us to capture the boundary if the marsh and/or forest extent is inaccurate.
As DSAS was not created with this purpose in mind, there are several limitations that we established work arounds for in the following steps. For example, DSAS crashes if the number of transects in the output file exceeds approximately 66,000 transects. If the estimated number of features (polygon perimeter divided by the transect interval) is larger than this number, the subregion you are working in needs to be broken into smaller zones. This calculation needs to be done before step a). Additionally, DSAS errors often occur if the tool tries to cast transects when there are multiple segmented shorelines and baselines within the same files. To reduce the complexity of the baseline and shoreline files, we have the baselines and shorelines extend around the entire forest polygon which we use as an approximation for the location of the marsh-forest boundary, so DSAS treats them as islands, and later remove the transects not cast where forest borders marsh.
a) If [LULC3_forestnew900m_s_nh.shp] needs to be split, CREATE FISHNET (Input=[LULC3_forestnew900m_s_nh.shp]). The number of columns and rows is dependent on the shape and distribution of forest polygons in the input file. SPLIT [LULC3_forestnew900m_s_nh.shp] into zones, such as Zone A, [LULC3_ZoneA.shp].
b) BUFFER (Input=[LULC3_ZoneA.shp], Linear unit=20m, Dissolve type=NONE) to create [LULC3_ZoneA_out20m.shp].
c) ELIMINATE POLYGON PARTS (Condition=percent; Part area percent=99; Part option=contained only) to remove any interior holes in the forest polygons [LULC3_ZoneA_b_nh.shp].
d) BUFFER (Input=[LULC3_ZoneA_b_nh.shp], Linear unit=-20m, Dissolve type=NONE) to create [LULC3_ZoneA_in20m.shp].
e) FEATURE TO LINE (Input=[LULC3_ZoneA_in20m.shp]) to create baseline [LULC3_ZoneA_line.shp]. Then DISSOLVE (Create multipart feature) to create a single baseline [LULC3_ZoneA_baseline.shp].
f) FEATURE TO LINE (Input=[LULC3_ZoneA_b_nh.shp]) to create shorelines [[LULC3_ZoneA_bdnh_line.shp]. Then DISSOLVE (Create multipart feature) to create a single shoreline [LULC3_ZoneA_shoreline.shp].

Cast transects across the marsh-forest boundary using Digital Shoreline Analysis Systems (DSAS) Version 5.0. FOr questions not covered by the information here, see the DSAS User Guide available online. Slope data is then added to these transects from the CoNED topobathy. Note that slope value of -9999 represents a slope crossing an area of the topobathy with a NaN value. Slope values of 0 flag transects which cross over areas where the originators of the dataset used artificial fill values such as -1 so there is artificially no change in slope. These are removed in step i).
a) IMPORT [LULC3_ZoneA_baseline.shp] and [LULC3_ZoneA_shoreline.shp] to a personal geodatabase to create feature class files [LULC3_ZoneA_baseline.mdb] and [LULC3_ZoneA_shoreline.mdb].
b) Change attributes to meet DSAS requirements: ADD FIELD [LULC3_ZoneA_shoreline.mdb] for DATE (Text, Length=10) and UNCERTAIN (Short integer) and delete Id field. For [LULC3_ZoneA_baseline.mdb] set Id equal to Object ID and make last column.
c) Import files into DSAS. Set output file to same personal geodatabase as input files [Transects_LULC3_ZA.mdb].
d) Set the following parameters in DSAS Set Default Parameters. Baseline tab: Right-hand baseline orientation, Baseline placement onshore; Shoreline tab: Landward intersection. Set the following parameters in DSAS Cast Transects. Maximum search distance: 25m, Transect spacing: 30m, Smoothing distance: 0m, Clip transects to shoreline extent.
e) SELECT BY LOCATION transects from the resulting transect file [Transects_LULC3_ZA.mdb] which are within 10m of [MDVA_NWI_E2EMwetlandsUTM.shp] to create shapefile with transects only across the marsh-forest boundary [Transects_LULC3_ZA_mfb.shp]. Save file outside of personal geodatabase.
f) Examine transects cast along artificial edges created by fishnets, both the exterior and interior fishnets. Remove transects which were cast along these artificial edges.
g) MERGE transects from all zones in region [TRANSECTS_LULC3_mfb.shp].
h) ADD SURFACE INFORMATION (Input surface= Chesapeake_Bay_Topobathy_DEM_1m_v2.tif; Output property=Maximum slope, Average slope).
i) Remove all transects where Maximum and/or Average slope equals -9999 or 0 as this reflects artificial evaluation values in the topobathy file.
j) MERGE all regions to create one file of transects for Chesapeake Bay (TRANSECTS_CBslope.shp).

Create points at marsh-forest boundary with the average slope across the transect. The shoreline and baseline files were created 10m on either side of the approximated marsh-forest boundary, so the mid-point of the transects is considered to be the approximate location of the boundary. However, this step does not calculate the slope at the boundary, the value of the point is still the average of the slope across the entire transect. The intent is to provide a visualization of the boundary for users and a file for users to create a boundary line file if they choose.
a) GENERATE POINTS ALONG LINES (Input features= TRANSECTS_CBslope.shp; Point placement=percentage; Percentage=50) to create points along the marsh forest boundary with the average and maximum slope values associated with the underlying transect [POINTS_CBslope.shp].

Added keywords section with USGS persistent identifier as theme keyword (20200807). Added the cross-reference to a primary related publication (20230901).