Trees for Shadowmap 🌲 🌴 🌳

The following repo:

1. Downloads open data of trees defined in sources.js
2. Converts the tree data into GeoJSON format
3. Maps property names to names used at OSM and removes faulty trees
4. Creates tiles containing the trees

Prerequisite

• Ensure you have Node and Anaconda installed
• Run npm install to install all Node dependencies
• Run conda env create -f conda-trees.yaml to create environment containing all Anaconda dependencies
• Run conda activate conda-trees to activate the environment

Problem with the Anaconda environment: This might help.

Afterwards, run npm run all-stages to execute the scripts.

1-download.js

Run npm run stage-1 to execute the first stage. It downloads the data (csv) as defined in sources.js. If the download is a ZIP file it gets extracted. Every sources corresponds to one csv file and ends up under the folder 1/unzip.

2-geojson.js

Run npm run stage-2 to execute the second stage. It uses the binary ogr2ogr (installed with Anaconda) to transform every csv file into a corresponding geoJSON file. Every line in the final files corresponds to one tree in geoJSON format. The files itself are not valid JSON files but a so-called ndjson files. You can find them under the folder 2/.

3-transform-and-filter.js

Run npm run stage-3 to execute the third stage. It reads every line of the geoJSON files and transforms the properties into OSM format. Afterwards every tree has for example a height property containing the height of the tree in meters as number. Furthermore, there is a sanity check executed for every tree. If a tree, for example has a negative height, it gets discarded. All trees passing the sanity check are written into new geoJSON files, every tree in its corresponding file, under the folder 3/.

4-create-tiles.js

Run npm run stage-4 to execute the fourth stage. It reads all geoJSON files from stage 3 and splits the trees in it corresponding tile files depending on its coordinates. The tiles can be found under the folder dist/.