/ForestedWetlands

Code for building raster files of elevation derivatives, python and R scripts for using these rasters for building and applying random forest models of wetland presence/absence in ArcGIS Pro

Primary LanguagePython

Forested Wetlands

Here you'll find resources for building and applying statistical models for identifying potential wetland locations using ArcGIS. These resources include tools to:

  • use a Digital Elevation Model (DEM) to calculate topographic attributes associated with wetland formation,
  • use a data set of point locations with known wetland presence or absence to build a random forest model to predict the probability of finding a wetland as a function of topographic and other attributes,
  • test the model by applying it to new data sets,
  • use the model to predict wetland presence in other locations.

These tools and the methods they use were developed by Dan Miller at TerrainWorks and Meghan Halabisky at the University of Washington Remote Sensing and Geospatial Analysis Lab as part of a wetland mapping project funded by the Cooperative Monitoring, Evaluation, and Research (CMER) Committee of the Washington Forest Practices Board and overseen by the Wetland Scientific Advisory Group (WetSAG).

To use these tools with ArcGIS Pro, click on the download button to save the contents of this repository to your hard drive.

The repository includes the following files and folders:

  • Wetland_Tools_for_ArcGIS.pdf, which contains instructions for installing and running the tools in ArcGIS.
  • Wetland_Intrinsic_Potential_Tool_Phase1.pdf, which contains the report for the first phase of this project, completed in 2017.
  • Wetland_Mapping_Tool_Phase2.pdf, which contains the report describing the methods developed for the tools provided here, with examples of application of the tools for building and applying a random-forest model for predicting wetland presence at four areas in western Washington.
  • DEMutilities, a folder containing files for the DEMutilities ArcGIS toolbox. These files include executable programs for calculating elevation derivatives and python scripts to use those programs with ArcGIS to create raster files.
  • WetlandTools, a folder containing files for the Wetland Tools ArcGIS tool box. These files include R scripts for building and applying a random forest model for predicting wetland presence.