Fix science issues in make_bcs (GEOSsurface_GridComp/Utils/Raster/)

#763 identifed the following issues with the generation of the EASE grid tile space:

• The Pfafstetter (catchment) ID assigned to a tile is that of the last contributing raster grid cell and not that of the dominant catchment within the EASE grid cell. An analysis by @zyj8881357 showed that approximately 36% of EASEv2_M09 tiles and 80% of EASEv2_M36 tiles do not represent the dominant catchment.
• The min/max lat/lon values for each tile in the catchment.def file seem to be associated with the EASE grid cell and not the desired lat/lon boundaries of the tile.
• It remains unclear if the adjustment of the land tile elevation properly accounts for lake and landice surfaces. This requires further examination.

Start in GEOSsurface_GridComp/Utils/Raster/ and trace make_bcs's EASE grid tile generation, then inspect how catchment.def records Pfafstetter IDs and min/max lat/lon. Compare the generated values with dominant catchments and the desired tile boundaries, and examine whether lake and landice surfaces are handled in the elevation adjustment. Done means all three reported science issues have verified behavior.