Adjust dem_10m.tif creation to not consider "masked" pixels in downsampling #6
Description
When producing unmasked strip DEM results, the post-process gdal_translate operation that downsamples the 2m _dem.tif strip result raster to the 10m _dem_10m.tif strip result raster does not attempt to apply the water and cloud components of the _bitmask.tif during the resampling process. Thus, potentially bad data values will mix with good data values in the bilinear resampling procedure.
Is there some way to isolate the two classes of DEM pixels ("valid" and "invalid"/masked) in the downsampling process?