MAP-Statewide-ATO-VMT

This repo creates a simple web map to view the ATO and VMT data from all the models in the state.

The web application itself can be found in the docs folder. The ViteJS App can be found in the _site folder.

Project structure

• scripts/00_download_raw_data.py downloads and extracts the raw data archive into data/raw/.
• scripts/01_prepare_data_layers.py checks and documents the raw inputs needed to rebuild the ATO and VMT data.
• scripts/02_process_ato_layers.py creates the canonical processed ATO artifacts in data/processed/ato/.
• scripts/03_process_vmt_layers.py creates the canonical processed VMT artifacts in data/processed/vmt/.
• scripts/04_build_web_artifacts.py copies the processed artifacts into _site/public/data/ato/ and _site/public/data/vmt/ so the browser can load them.
• pipeline/ stores reusable data-processing code shared by the numbered scripts (config.py for repo paths/model areas/PMTiles settings, io_utils.py for parquet/PMTiles/geometry helpers, ato.py and vmt.py for the dataset-specific logic - these two don't import from each other).
• data/raw/ stores raw model outputs and shapefiles. Large/raw files are gitignored, so a new clone needs scripts/00_download_raw_data.py run before processing.
• data/processed/ato/ and data/processed/vmt/ store canonical generated web-ready data.
• _site/ is the Vite/Vue web app.
• _site/public/data/ato/ and _site/public/data/vmt/ store the browser-serving copies used by the Vite app.
• _site/VITE_GUIDE.md has the quick file-by-file guide for the Vite/Vue app.

Rebuilding the data

Install uv, then set up the environment:

uv sync

You'll also need the pmtiles CLI and ogr2ogr (from GDAL) on your PATH - these aren't managed by uv. (pmtiles can instead be pointed to via a PMTILES_EXE env variable.)

Run the numbered scripts from the repo root, in order:

uv run scripts/00_download_raw_data.py     # downloads + extracts raw data into data/raw/
uv run scripts/01_prepare_data_layers.py   # checks raw inputs are present
uv run scripts/02_process_ato_layers.py    # builds data/processed/ato/
uv run scripts/03_process_vmt_layers.py    # builds data/processed/vmt/
uv run scripts/04_build_web_artifacts.py   # publishes both into _site/public/data/

The current web map uses simplified TAZ polygons in PMTiles for speed, with one vector feature per ModelArea + CO_TAZID and year-prefixed metric fields. ATO fields are built from the Access to Opportunity CSVs. VMT fields are built from the TAZ-based produced/attracted VMT CSVs for Statewide (TAZ-Based-VMT.csv) and Wasatch Front (TAZ-Based Metrics.csv), filtered to Metric = VMT, then summarized by produced/attracted direction, period, and purpose with Daily rollups plus Person, Truck, and Other purpose-group totals. TAZ outlines are stored in separate, lightly simplified boundary PMTiles files so the fill layers can stay faster without losing readable outlines.

Running the web app

cd _site
npm install
npm run dev

npm run dev starts the local Vite preview server for editing. When changes are ready for GitHub Pages, build the app from _site:

npm run build

That command creates the production Vite bundle and overwrites ../docs/, which is the branch folder GitHub Pages serves. Commit the updated docs/ files along with the source changes.