Data Management Plan

This polygon data set includes raw values and normalized scores for thirteen landscape scale metrics that characterize marsh resilience to sea level rise within watersheds along the coast of the conterminous United States. These metrics fall into bins related to current marsh condition, marsh vulnerability, and adaptation potential. The data are summarized at the watershed scale (HUC-12 units).

Process Steps:

• 2020-09-01 00:00:00 - Marsh Unit Codes (MUC) were created from the C-CAP 30 meter land cover data. Classes 16-18 (the estuarine wetland classes) were extracted and recoded using a morphometric algorithm to distinguish between different components of the marsh units. The following classification scheme was established: gridcode1 = core wetland gridcode2 = vegetated edge of wetland gridcode3 = unvegetated edge of wetland Core-to-edge ratio computed from MUCs and aggregated by 12-digit HUC. [Core_Edge_ratio = gridcode1 / ( gridcode2 + gridcode3 )] Unvegetated edge to vegetated edge ratio was computed from MUCs and aggregated by 12-digit HUC. [UnvegVegEdge_ratio = gridcode3 / gridcode2] A simplified version of the C-CAP 30 meter land cover was generated to facilitate derivation of land cover metrics. The resulting data layer had the following classification scheme: gridcode2 = high intensity developed gridcode3 = medium intensity developed gridcode4 = low intensity developed gridcode5 = open space developed gridcode6 = agricultural classes (pasture/hay, cultivated) gridcode7 = natural cover types (grassland, shrub, forest) Percent impervious cover was computed from the simplified version of C-CAP 30 meter land cover data and aggregated by 12-digit HUCs. Analysis was performed within a 150 meter buffer around each marsh unit. [Perc_IC = ((gridcode2 * 0.8503) + (gridcode3 * 0.5768) + (gridcode4 * 0.2929) + (gridcode5 * 0.0941)) / Total_Area * 100] Percent natural cover was computed from the simplified version of C-CAP 30 meter land cover data and aggregated by 12-digit HUCs. Analysis was performed within a 150 meter buffer around each marsh unit. [Perc_Natural = gridcode7 / Total_Area * 100] Percent agricultural cover was computed from the simplified version of C-CAP 30 meter land cover data and aggregated by 12-digit HUCs. Analysis was performed within a 150 meter buffer around each marsh unit. [Perc_Ag = gridcode6 / Total_Area * 100] Soil erodibility was computed using Esri's USA Soils Erodibility Factor image service (https://landscape11.arcgis.com/arcgis/rest/services/USA_Soils_Erodibility_Factor/ImageServer, accessed March 2018). Analysis was performed within each marsh unit, not the entire HUC. The average erodibility factor for each marsh unit was weighted by the size of the marsh unit and aggregated by 12-digit HUC. Tidal range was computed as the height difference between Mean Higher High Water (MHHW) and Mean Lower Low Water (MLLW) measured in meters. The tidal datum data were extracted from the VDatum tool and interpolated across data gaps to provide complete coverage within the study area. Percent of marsh below mean higher high water (MHHW) was computed by intersecting all marsh units below MHHW, dividing by the total marsh area, and aggregating by 12-digit HUC. Percent of marsh below mean tide level (MTL) was computed by intersecting all marsh units below MTL, dividing by the total marsh area, and aggregating by 12-digit HUC. Percent hardened shoreline was computed using the Environmental Sensitivity Index (ESI) database. The ESI shoreline data were divided by and associated with 12-digit HUC codes. Within each HUC, all shoreline features that were armored (GENERALIZED_ESI_TYPE LIKE '%Armored%') were divided by the total shoreline length and multiplied by 100. Environmental Sensitivity Index (ESI) maps provide a concise summary of coastal resources that are at risk if an oil spill occurs nearby. Examples of at-risk resources include biological resources (such as birds and shellfish beds), sensitive shorelines (such as marshes and tidal flats), and human-use resources (such as public beaches and parks).
• 2020-09-01 00:00:00 - NOAA ESI National Shoreline (2017) Migration space was determined using NOAA's sea level rise inundation data. Within every 12-digit HUC and for each foot of inundation above MHHW, the area of potential future marsh was divided by the current area of marsh to generate a ratio of future to present "potential" marsh area. This operation was performed for 1-6 feet of sea level rise inundation scenarios. For each SLR scenario, the resulting migration ratio values were ranked and scored by a quantile distribution function. The six scenarios were were then averaged, a new quantile rank and score was generated and reported by AVG_migration_ratio. Wetland connectedness, at this national level analysis, was computed using the marsh unit data (MUCs) and an analysis of projected fragmentation/consolidation under the 4 foot future sea level rise (SLR) scenario. A region grouping process was used to group all connected marsh units under current and future scenarios. Within each 12-digit HUC, the number of unique future marsh units were subtracted from the number of unique current marsh units, and divided by the number of unique current marsh units. This became the unitless raw value for the Wetland_Connectedness metric. Shoreline sinuosity was assessed using the NOAA ESI shoreline vector data to characterize sinuosity within each 12-digit HUC. A sinuosity index was computed using a Sinuosity python script provided by Esri. High values represent linear shorelines and low values represent sinuous shorelines. Since the the sinuosity script generates values from 0-1, where 1 is a straight feature, we multiplied the sinuosity index by -1 so that values closer to 0 were scored higher in the quantile index. Quantile scores were generated using a python script developed by NOAA OCM. Each metric was ranked and scored 1-10 based on the quantile position in the ranking. The quantile scores were then inversed if required based on the intent of the metric (negative or positive contributor to resilience). The quantile scores were summed by category (current condition, vulnerability, and adaptive capacity) and all combined, and new quantile rankings and scores were generated. The numerical quantile scores were used to generate ordinal data describing the degree to which marsh units scored "low" or "high" for each resilience category (current condition, vulnerability, and adaptive capacity). Scores 1-5 were assigned "low" and scores 6-10 were assigned "high." An overall management category was generated using a concatenation of the three resilience categories, in the order just shown. Vulnerability scores were updated to fix an error caused by the erroneous inclusion of the hardened shoreline metric in the original scoring (3/23/2022).