eyeseast / baltimore-trees

A demo project for NICAR24 in Baltimore

Baltimore trees

This is a demo project for NICAR24, showing the location of trees in the city of Baltimore, layered on a self-hosted street map. Everything the map needs to run is in this repository. Once dependencies are installed, it should run without an internet connection.

Getting started

The Makefile in this project contains all the steps to build both a map and a Datasette instance to explore the underlying data.

To build and run everything, use the following commands:

# install both Python and JS dependencies
make install

# build the database
make trees

# download and build tiles
make tiles

# run both servers
make run

These commands depend on having Python, Poetry, the PMTiles CLI and Tippecanoe installed locally. To just see the finished map, run the following commands:

npm ci
npm run dev -- --open

Tiles and basemap

Protomaps offers free basemaps built on OpenStreetMap.

I used this site to find a bounding box of the area around Baltimore, and then the pmtiles CLI to extract tiles for that area.

pmtiles extract https://build.protomaps.com/20240211.pmtiles public/baltimore.pmtiles --bbox="-76.861861,39.096181,-76.360388,39.454149

Fonts

Fonts are copied from the Protomaps basemaps-assets repository. To use a custom typeface, generate PBFs with font-maker.

Tree data

The City of Baltimore publishes a public tree inventory on its data portal. Running make data/trees.ndjson will use a Python script to download the full dataset as newline-delimited GeoJSON.

Datasette offers a way to quickly explore data, including with maps. The Makefile in this project includes commands to build an SQLite database with SpatiaLite and run a Datasette instance.

# build the database, downloading data if needed
make trees

# run the server
make run

Improve this map

I've deliberately left some things off this map. If you're looking through this codebase, I encourage you to fork it and add:

• controls
• search
• a legend
• zoom-based scaling for tree points
• color-coding of trees, based on species or some other property
• other data-driven overlays (local income, redlining history, etc)

Trees on public streets, or lack thereof, are often an indicator of inequality. Use this map as a starting point to look for bigger stories.

Further reading

• Protomaps
• PMTiles
• PMTiles CLI
• MapLibre