Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

2b - Inside the Workflow

Inside the Workflow

The Newport DEM should now be running.

Keep the terminal visible. You do not need to interpret every message—just recognize the major stages:

discover data
      ↓
prepare each source
      ↓
standardize reference systems
      ↓
prioritize and stack measurements
      ↓
interpolate across multiple scales
      ↓
write the final DEM

The tools working together

ToolMain role
FetchezOrchestrates the recipe and data pipeline
TransformezSupports horizontal and vertical reference transformations
GlobatoProvides elevation-specific filtering, gridding, and DEM construction

We launch the build with Globato, while Fetchez coordinates the data pipeline and Transformez supports reference-system transformations.

Our Newport recipe combines:

coupling-bathy-topo
        +
dav:survey_id=9693,weight=100
End-to-end coastal DEM workflow showing Fetchez orchestration, national and local source data, Transformez and Globato processing, stacked measurements, interpolation, the final DEM and spatial metadata, and independent validation with ICESat-2.

Coastal DEM workflow. The build moves from data discovery through source preparation, reference-system standardization, DEM construction, and final outputs.


1. Discover the data

The workflow first determines which configured sources overlap Newport.

Fetchez provides a common way to work with data from different catalogs, APIs, formats, and providers.

Explore Fetchez

Optional commands for a second terminal:

fetchez modules list
fetchez modules info dav

These let you inspect the available data modules and learn more about a specific module, such as the Digital Coast DAV source.

About the workshop cache

The command uses:

--shared-cache ~/workshop/newport_data

The workshop setup staged prepared source data there so the build can reuse them rather than download everything again.

~/workshop/newport_data/    → prepared source-data cache
~/workshop/newport_dem/     → processing files and outputs

Messages about reusing cached data are expected.


2. Prepare each source

Coastal elevation datasets are not all equivalent.

A reproducible workflow can process each source appropriately while keeping the overall workflow consistent.

Why does the coastline matter?

A coastal DEM combines measurements collected over both land and water, often from sources with different sensors and data characteristics.

Land/water constraints can help prevent topographic measurements or interpolation from extending inappropriately into bathymetric areas, and vice versa. The coastline is therefore part of the processing logic, not just a map feature.


3. Standardize the reference system

Our target Newport DEM uses:

Source datasets may arrive in different horizontal or vertical reference systems. Before they can be combined, they need to be expressed in the requested common system.

This is where Transformez supports the workflow.

Why is the vertical datum important?

The same physical location can have different numerical elevations when referenced to different vertical surfaces.

A vertical transformation changes the reference used to describe the elevation, not the physical terrain or seafloor.

Explore Transformez

In a second terminal:

transformez list

You do not need to run transformations manually. Fetchez coordinates the workflow and uses Transformez where needed.


4. Prioritize and stack measurements

Once prepared and standardized, measurements are brought into a common raster framework.

Where datasets overlap, recipe priorities help determine which information should dominate.

For Newport, we added:

dav:survey_id=9693,weight=100

so the selected local topobathymetric lidar receives strong priority where it overlaps lower-priority sources.


5. Interpolate across multiple scales

Measurement density can vary dramatically across a coastal region.

Dense lidar can support fine spatial detail, while sparse bathymetric observations require information to be integrated across broader scales.

Use fine spatial information where observations support it, and progressively broader information where measurements are sparse.


6. Write the final outputs

When processing is complete, Globato writes the final DEM and supporting products.

The two primary Newport products are:

~/workshop/newport_dem/newport_n44x64_w124x10_final.tif
~/workshop/newport_dem/newport_n44x64_w124x10_hs.tif
*_final.tif   → finished coastal DEM
*_hs.tif      → hillshade for visual inspection

In Module 3, we will also use:

~/workshop/newport_dem/newport_n44x64_w124x10_sm.gpkg

This spatial metadata GeoPackage lets us see which source data support different parts of the DEM.

Other supporting outputs

Globato can also create:

OutputPurpose
Source masksShow where individual input files contribute
Sources VRTCombines source-data coverage
Spatial metadataGIS view of source support
Spatial metadata stylePrepared QGIS symbology
Provenance rasterCompact cell-by-cell source provenance
Stack rasterAccumulated information used during gridding

You do not need to inspect all of these during the workshop.

If your build finishes early

Confirm the primary products:

ls -lh \
  ~/workshop/newport_dem/newport_n44x64_w124x10_final.tif \
  ~/workshop/newport_dem/newport_n44x64_w124x10_hs.tif

We will interpret them in Module 3.


One last look at the terminal

Which stage best describes what you see?


If your Newport build did not finish

Known-good Newport fallback files are available in:

~/workshop/reference_outputs/newport/

If an instructor directs you to use them, stop the running build with Ctrl+C if necessary, then run:

rm -rf ~/workshop/newport_dem
mkdir -p ~/workshop/newport_dem
cp -a ~/workshop/reference_outputs/newport/. ~/workshop/newport_dem/
cd ~/workshop/newport_dem
ls -lh

The fallback set contains:

newport_n44x64_w124x10_final.tif
newport_n44x64_w124x10_hs.tif
newport_n44x64_w124x10_sm.gpkg
newport_n44x64_w124x10_sm.qml

That is enough to continue with Module 3.