geotiff-to-wps 0.1.0

Convert GeoTIFF DEMs to WPS geogrid binary format for WRF

geotiff-to-wps

Convert GeoTIFF files to WPS geogrid binary format for use as custom static data in WRF. Supports topography, land use, soil type, green vegetation fraction, albedo, and other fields via built-in presets or custom configuration.

---

Source Code: https://github.com/geotiff-to-wps

---

Installation

pip install geotiff-to-wps

Or with UV:

uv add geotiff-to-wps

Usage

Command line

geotiff-to-wps input.tif output_dir/

The input GeoTIFF must be in EPSG:4326 (WGS84 geographic coordinates). The output directory will contain the WPS binary tiles and an index file ready for use with WPS geogrid.

Presets

Use --preset (or -p) to select defaults for common data types:

geotiff-to-wps -p dem dem_4326.tif /path/to/WPS_GEOG/topo_custom/
geotiff-to-wps -p landuse landuse_4326.tif /path/to/WPS_GEOG/lu_custom/
geotiff-to-wps -p greenfrac gvf_4326.tif /path/to/WPS_GEOG/gvf_custom/

Available presets:

Preset	Type	Wordsize	Signed	Scale factor	Tile border	Description
dem (default)	continuous	2	yes	—	3	Topography height (meters)
landuse	categorical	1	no	—	0	Land use category
soiltype	categorical	1	no	—	0	Soil type category
greenfrac	continuous	2	yes	0.001	3	Green vegetation fraction
albedo	continuous	2	yes	0.001	3	Albedo

Overriding preset defaults

Any preset setting can be overridden individually:

geotiff-to-wps -p dem --units feet --description "Elevation in feet" input.tif output/
geotiff-to-wps -p landuse --category-min 0 --category-max 40 input.tif output/

Available override options: --data-type, --signed/--unsigned, --wordsize, --tile-bdr, --missing-value, --scale-factor, --units, --description, --category-min, --category-max.

Python

from geotiff_to_wps import convert

# DEM (default preset)
convert("dem_4326.tif", "/path/to/WPS_GEOG/topo_custom/")

# Land use
convert("landuse_4326.tif", "/path/to/WPS_GEOG/lu_custom/", preset="landuse")

# Green fraction with custom description
convert("gvf.tif", "output/", preset="greenfrac", description="Monthly GVF")

Full pipeline: custom static data in WRF

This guide walks through the complete process of replacing WRF's default static data with custom GeoTIFF sources, using topography as an example.

Step 1: Prepare the GeoTIFF

If your data comes as multiple tiles, mosaic them first:

gdal_merge.py -o merged.tif tile1.tif tile2.tif ...

Reproject to EPSG:4326 with bilinear interpolation and zstd compression:

gdalwarp -t_srs EPSG:4326 -r bilinear -co COMPRESS=ZSTD merged.tif dem_4326.tif

Bilinear interpolation avoids stair-stepping artefacts in the reprojected elevation. For categorical data (land use, soil type), use nearest-neighbour instead:

gdalwarp -t_srs EPSG:4326 -r near -co COMPRESS=ZSTD landuse.tif landuse_4326.tif

The zstd compression reduces file size without affecting the conversion.

Step 2: Convert to WPS format

geotiff-to-wps dem_4326.tif /path/to/WPS_GEOG/topo_custom/

This creates a directory containing:

• Binary tile files with border overlap for interpolation (continuous data) or exact tiles (categorical data)
• An index file describing the grid geometry, projection, and tile layout

Step 3: Update GEOGRID.TBL

In your WPS installation, edit geogrid/GEOGRID.TBL (or GEOGRID.TBL.ARW). Find the entry for the field you are replacing (e.g., HGT_M for topography) and add a rel_path line pointing to your new dataset:

name = HGT_M
    ...
    interp_option = default:average_gcell+four_pt
    rel_path = default:topo_gmted2010_30s/
    rel_path = 3s:topo_custom/

The label before the colon (e.g., 3s) is a resolution tag. Geogrid selects the dataset whose resolution tag best matches each domain's grid spacing:

Label	Approximate resolution
30s	~1 km (30 arc-seconds)
10s	~300 m
3s	~90 m
1s	~30 m

The default entry is used when no better match is found. Higher-resolution data is automatically used for finer nest domains.

The interp_option controls how source data is interpolated to the WRF grid. See the WPS interpolation options documentation for all available methods.

Step 4: Run geogrid

Run geogrid as usual:

./geogrid.exe

Verify the result by inspecting the relevant field in the output geo_em.d0X.nc files:

import xarray as xr

ds = xr.open_dataset("geo_em.d03.nc")
ds["HGT_M"].isel(Time=0).plot()

Compare grid-cell values against known reference data to confirm the custom dataset is being used.

Development

Setup environment

We use UV to manage the development environment and production build.

uv sync

Run unit tests

You can run all the tests with:

uv run pytest

Format the code

Execute the following commands to apply linting and check typing:

uv run ruff check .
uv run black --check --diff .
uv run mypy --install-types --non-interactive geotiff_to_wps

To auto-format:

uv run black .
uv run ruff check --fix .

License

This project is licensed under the terms of the Apache Software License 2.0.