Hydrologic Terrain Analysis: Project Examples
Automated Watershed Delineation and Land Use Summary for Stream Monitoring Stations
Description: Created statewide hydro-conditioned 10m Digital Elevation Model (DEM) using NHD+ flowline data and the AGREE method developed at the Center for Research in Water Resources at the University of Texas Austin. Developed memory-efficient Python routine using arcpy and pandas that delineates watersheds for multiple XY coordinate points. Each point represents a stream monitoring station. Note that nested watershed are fully delineated from their respective pour points (Figure 1). The script uses the delineated watershed shapefiles to extract and summarize additional spatial data for each point. In this case, the tool returns formatted tabular summaries of NLCD years 2001, 2006, and 2011; the output includes area in hectares and % watershed area for each NLCD class.
Figure 1. Example GIS output from automated watershed delineation script. White River basin, VT, USA.
Flow-Based Intermittent Stream Channel Prediction
Description: Created statewide hydro-conditioned 10m Digital Elevation Model (DEM) using NHD+ flowline data and the AGREE method developed at the Center for Research in Water Resources at the University of Texas Austin. Interpolated USGS runoff contours to create continuous runoff raster surface (Figure 2). Used TauDEM hydrologic tool suite to calculate d-infinity flow direction and accumulation rasters and estimate accumulated runoff. Applied flow threshold to identify potential intermittent channels (Figure 3). Integrated intermittent shapefile with NHD+ flowline data.
Figure 2. Interpolated runoff raster based on USGS Average annual runoff in the United States, 1951-80.
Figure 3. NHD+ flowline hydrography integrated with predicted intermittent channel data.
NHD+ v2 hydrologic catchment routing and multi-scale spatial data aggregation
Description: Wrote memory-efficient Python routine to summarize 2001, 2006, and 2011 National Land Cover Database (NLCD) land use and land cover (LULC) and percent impervious cover data for all 5900+ NHD+ catchments in Vermont. A catchment is a sub-HUC12 scale watershed with a unique ID (COMID); the median catchment area in VT is 2.1 km2 (Table 1). The Python script summarizes NLCD data as total hectares (ha) and % area for each land class at the NHD+ catchment scale; percent impervious cover is characterized by summary statistics (min, max, quartiles, mean, and sd) for each catchment. This geospatial routine was then repeated for LULC data within an estimated active river (riparian) zone for each catchment. Value added attribute (VAA) tabular data were then obtained from the NHD+ dataset and joined to the catchment layer; these data provide to/from node information that describes hydrologic routing between catchments. A routing algorithm was then written in R to account for hydrologic divergences and accumulate upstream values for each unique catchment COMID. This algorithm calculates total drainage area and LULC class area for each catchment’s upstream drainage (total and riparian). The result is LULC data (or any data that can be linked to catchment COMID) summarized at multiple spatial scales (Figure 4). A range of data - hydrologic, water quality, aquatic biota, etc. - can be associated with these multi-scale LULC values and used in additional modeling exercises. Tabular outputs were also organized in a relational database with defined key fields and relationships; this allows watersheds, catchments, and riparian areas to be queried by LULC characteristics; for example, find the % of a catchment drainage that is forested (Figure 5), or identify catchments where upstream riparian areas are < 50% forested.
Table 1. Summary statistics for NHD+ catchment areas in Vermont; units are km2
| Min | 25% | Median | Mean | 75% | Max |
|---|---|---|---|---|---|
| 9e-04 | 0.687 | 2.115 | 4.626 | 5.542 | 66.79 |
Figure 4. Spatial scales used in data summaries. For each unique NHD+ catchment COMID, data were summarized at multiple scales: NHD+ (local) catchment, total upstream drainage, local active river area, and upstream active river area.
Figure 5. Total accumulated % upstream forest (NLCD classes 41, 42, and 43) for each unique NHD+ catchment COMID. White River basin, VT, USA.