Prerequisites

Download OSM extract

I used the following, but you can use any extract:

https://download.geofabrik.de/north-america/us/virginia.html

Convenience Script

For convenience, these steps have been converted into process_extract.sh. To run the script, edit the following lines to reflect the directory and filename of your .osm.pbf file:

OSM_PBF_DIRECTORY=/home/wcedmisten/Downloads
OSM_PBF_FILENAME=virginia-latest.osm.pbf

Then run ./process_extract.sh

Manual Set up

If you wish to run these steps individually, you can instead run:

PostGIS

Start up PostGIS

cd osm2pgsql
docker compose up -d

osm2pgsql

Install the osm2pgsql tool:

apt install osm2pgsql

Import the data. This will take a few minutes or many hours depending on the size and your hardware.

# replace /path/to/extract.osm.pbf with the actual file path, e.g. ~/Downloads/virginia-latest.osm.pbf
PGPASSWORD=password osm2pgsql --create --verbose -P 15432 -U postgres -H localhost -S osm2pgsql.style /path/to/extract.osm.pbf

Valhalla

Copy osm extract to valhalla_data/

cd ..
cp /path/to/extract.osm.pbf valhalla_data/

Run valhalla with:

docker run -dt -v $PWD/valhalla_data:/custom_files -p 8002:8002 --name valhalla gisops/valhalla:latest

Then run

docker logs -f valhalla

until you see the line INFO: Found config file. Starting valhalla service!

Find the hospitals and all isochrone data for them

Retrieves isochrones within distance of all hospitals.

python3 query_hospitals.py

Convert isochrone polygons to mbtiles

Install tippecanoe

Follow the instructions here: https://github.com/felt/tippecanoe

Convert the geoJson polygons to mbtiles. --simplification 6 is needed for displaying large isochrone polygons in mapLibreGL. Otherwise we hit the maximum vertices available in WebGL (65535) at low zoom levels. This also helps reduce the size of the tiles.

tippecanoe --simplification 6 -f polygon_0.json -f polygon_1.json -f polygon_2.json -f polygon_3.json -o virginia_iso.mbtiles

Install go-pmtiles

https://github.com/protomaps/go-pmtiles

Convert mbtiles to PMTiles format

git clone https://github.com/protomaps/go-pmtiles
cd go-pmtiles/
go install
go run main.go convert ../virginia_iso.mbtiles ../virginia_iso.pmtiles
cd ..

Generating a base map with planetiler

cp /path/to/extract.osm.pbf planetiler_data/

docker run -e JAVA_TOOL_OPTIONS="-Xmx4g" -v "$(pwd)/planetiler_data":/data ghcr.io/onthegomap/planetiler:latest --download --osm-path=/data/virginia-latest.osm.pbf --output /data/extract.pmtiles --maxzoom=8

This will automatically create a tiles file planetiler_data/extract.pmtiles in the PMTiles format. No conversion is necessary.

How to use these files

Three files will be created if these steps are followed succesfully:

• a base map PMtiles file (planetiler_data/extract.pmtiles)
• a driving distance PMtiles file (<extract_name>_iso.pmtiles)
• a GeoJSON file containing all hospitals and their names (hospitals_layer.json)

These can be used with a MapLibreGL project, similar to this demo:

https://github.com/wcedmisten/nextjs-protomap-demo/blob/main/pages/isochrone/index.tsx

Copy the two PMtiles files into the /public directory, and update the sources to:

"sources": {
    "openmaptiles": {
    "type": "vector",
    "url": "pmtiles:///nextjs-protomap-demo/extract.pmtiles"
    },
    "isochrone": {
    "type": "vector",
    "url": "pmtiles:///nextjs-protomap-demo/<extract_name>_iso.pmtiles",
    },
},

population.py

Examines population data within driving distance to a hospital.

You will need to download the GHS-POP dataset and modify the path in population.py

https://ghsl.jrc.ec.europa.eu/download.php?ds=pop

Select GHS-POP 2020 100m Mollweide

Update the GHS_POP_FILENAME variable to your downloaded file path.