[21pt] Subgrid modeling for REM - testing for high-res FIM in HydroVIS

[CarsonPruitt-NOAA]

HydroVIS would like some 3m HAND data with which to do some testing and cost estimates for one HUC. I'd like to take this opportunity to create a tools/subgrid_rem.py module that would be a neat research opportunity for some possible tests in the future. This will not be integrated into fim_pipeline and we don't want to go as far as creating a high-res DEM download tool in data/ for this either. It will simply be a standalone script that we may look further into in the future.

Rough plan for steps:

1. Pick a HUC where there is 3m DEM data available from the USGS 3DEP. Ideally, there should also be some benchmark data in the selected HUC where we can do some testing.

2. Download the 3m data. If only 1m is available, then resample it to 3m.

3. Write the tools/subgrid_rem.py module:

   • Inputs:
     • fim_pipeline output directory
     • 3m DEM
     • Raw DEM used for HAND (10m)
   • Output: same fim_pipeline directory with modified, high-resolution REMs
   • Core logic:
     • Resample 10m HAND DEM to 3m
     • Subtract 3m USGS 3DEP from resampled 10m DEM
     • Add the difference grid into each branch REM in the fim_pipeline directory
     • This process can be summarized by the following psuedo-raster-calculator equation: subgrid_REM = REM_10m + (DEM_10m - DEM_3m)

4. EXTENSION: Evaluate inundation outputs from the new subgrid HAND against benchmark data.

[CarsonPruitt-NOAA]

As we discussed in our meeting, we're going to try to do this with 1m DEM data instead of the 3m mentioned above.