After working through fetchez, transformez, and globato, this final Day 2 exercise brings the workflow together by building a DEM for a custom coastal region. You will choose a small coastal area, apply the same globato build pattern used for the Fiji example, and review the resulting output for basic quality-control issues.
Export the Custom RecipeΒΆ
Choose a small coastal region and apply the same workflow pattern used for the Fiji DEM.
General command:
globato cudem build -R WEST/EAST/SOUTH/NORTH -E 3s -O custom_test -P epsg:4326+3855 -M ms_binary_cudem:algos=interp_gmt -X 0:5 global-bathy-topo --exportReplace:
WEST/EAST/SOUTH/NORTHwith the selected coastal bounding box.
Example centered on the NukuΚ»alofa, Tonga region:
globato cudem build -R -175.40/-175.00/-21.35/-21.00 -E 3s -O custom_test -P epsg:4326+3855 -M ms_binary_cudem:algos=interp_gmt -X 0:5 global-bathy-topo --exportThe --export option writes the generated YAML workflow to a recipe file instead of running the full DEM build.
Expected recipe output:
custom_test_recipe.yamlOpen the exported recipe in a text editor and compare it with the Fiji recipe from Day 1.
Run the Custom RecipeΒΆ
Run the exported recipe to generate the DEM:
globato cudem run custom_test_recipe.yamlExpected final DEM:
custom_test_final.tifInspect the OutputΒΆ
Inspect the custom DEM the same way you inspected the Fiji output in the globato module:
gdalinfo custom_test_final.tifCreate a color shaded-relief image:ΒΆ
globato perspecto hillshade --helpExample hillshade command:
globato perspecto hillshade INPUT_DEM.tif OUTPUT_HILLSHADE.tif --cmap etopo --exag 3 Basic Output ReviewΒΆ
Review the generated DEM and hillshade using the following questions:
Does the DEM cover the expected region?
Are land and water in the correct locations?
Are land elevations generally positive?
Are underwater elevations generally negative?
Are there missing-data areas or obvious gaps in source coverage?
Are there edge or coastline artifacts?
Are there interpolation artifacts or unexpected surface patterns?
Does the hillshade show reasonable terrain and bathymetric structure?
Do neighboring source datasets appear consistent?
Are there areas that should be checked against local or independent data?
This is not a complete validation procedure. The goal is basic output review and quality-control thinking about the relationship between source data, workflow settings, and DEM quality, using tools available within the workshop environment.
Modify One ThingΒΆ
After the first custom run, modify one part of the command or recipe.
Change the RegionΒΆ
Export a new recipe using a different bounding box:
globato cudem build -R WEST/EAST/SOUTH/NORTH -E 3s -O custom_region_v2 -P epsg:4326+3855 -M ms_binary_cudem:algos=interp_gmt -X 0:5 global-bathy-topo --exportThen run it:
globato cudem run custom_region_v2_recipe.yamlChange the ResolutionΒΆ
Export a 1-arc-second version:
globato cudem build -R 178.25/178.65/-18.30/-17.95 -E 1s -O fiji_test_1s -P epsg:4326+3855 -M ms_binary_cudem:algos=interp_gmt -X 0:5 global-bathy-topo --exportRun the exported recipe:
globato cudem run fiji_test_1s_recipe.yamlAlternatively, override the saved Fiji recipe:
globato cudem run fiji_test_recipe.yaml -E 1s -O fiji_test_1sCompare the outputs:
gdalinfo fiji_test_final.tif
gdalinfo fiji_test_1s_final.tifThe goal is not to produce a perfect DEM. The goal is to understand how changes to the region, resolution, sources, and workflow settings affect the resulting output.
Wrap-Up and Next StepsΒΆ
By the end of Day 2, participants should have:
Reactivated the
cudemenvironmentReviewed the Fiji test DEM
Seen how
fetchez,transformez,globato, and IVERT support the CUDEM workflowExplored source modules and YAML formatting
Reviewed the role of vertical harmonization
Generated or inspected a custom DEM recipe
Reviewed DEM outputs for basic quality-control issues
Understood how the Fiji recipe can be adapted to other coastal regions
Suggested Next StepsΒΆ
Explore additional source modules and module bundles
Inspect additional example recipes
Try a new coastal region
Compare different output resolutions
Add local or project-specific source data
Begin adapting the workflow to your own project area
Use IVERT or other independent data to assess the resulting DEM
Continue using Zulip for questions, support, and sharing workflow ideas