Utilities for working with the GEBCO + RTopo-2.x digital elevation models in MPAS-like models.
python3 dem_remap.py \
--mpas-file="path+name-to-mpas-mesh-file" \
--elev-file="path+name-to-DEM-pixel-file"
can be used to inject a (full-sphere) MPAS-like mesh with the bed_elevation, ocn_thickness and ice_thickness varaibles derived from a compressed GEBCO + RTopo-2.x 'pixel' file. A 'remap'-style formulation is used, ensuring consistent interpolation onto variable resolution voronoi-type grids.
python3 dem_trnsf.py \
--base-mesh="path+name-to-base-mesh-file" \
--part-mesh="path+name-to-part-mesh-file"
can be used to transfer the remapped elevation data from a (full-sphere) MPAS-like mesh onto a partial sub-mesh, generated via a culling operation or equiv.
python3 dem_pixel.py \
--elev-path="full-path-to-raw-DEM-assets" \
--save-path="full-path-to-output-storage"
can be used to create the compressed GEBCO + RTopo-2.x 'pixel' files required for the above workflow. The underlying data-sets are modified to use a 'pixel'-centred (rather than grid-centred) layout, compressed into int16_t arrays (to reduce memory use), and are updated to expose the thickness-type variables listed above. In the case of GEBCO, ice sheet/shelf elevation and thickness data is produced by 'blending' with the R-Topo-2.x data-sets at the ice fronts. To build the derived files, the following raw data-sets are needed:
- The
RTopo-2.0.4data due to Schaffer et al:doi.pangaea.de/10.1594/PANGAEA.905295. - The
GEBCO_v2024data due to Weatherall et al:doi.org/10.5285/1c44ce99-0a0d-5f4f-e063-7086abc0ea0f.