Practical and Impractical Uses of Terrain Data

Open Terrain

About Open Terrain

• Supported by a Knight News Grant
• Catalog Global Open Terrain Data
• Document Workflows for Processing Terrain Data
• Create a global version of Stamen Terrain

Open Terrain Outcomes

• Github Wiki on Open Terrain data sources
• Terrain Classic map tiles
• AWS Tiler

Terrain Classic

About Terrain Classic

• Global version of Stamen's U.S. Terrain map tiles
• Ported to CartoCSS from Cascadenik
• NodeJS, Mapnik, Make, Postgres & PostGIS, Tessera
• AWS Tiler creates separate tile layers: background, lines, labels, hillshade
• Main tiles are created by compositing with Mapnik
• Terrain data sourced from Mapzen Elevation Data API

Terrain Classic Outcomes

• General Purpose Map Style
• Free, Public Tile Source
• CartoCSS IS NOT DEAD
• Shout out to Tom Patterson's Web Shaded Relief Guide

AWS Tiler

Web Browser Request -> Tile {z}/{x}/{y} URL

AWS: Does the Tile Exist?

IF TRUE -> Grab from S3 Store

ELSE -> Create it & Write to S3 then grab

Send back to Web Browser

Link to Demo

AWS Tiler Stack

• AWS: Lambda, S3, EC2, API Gateway
• GDAL: Raster Overviews & Warped VRTs
• Python's Matplotlib & Rasterio
• NodeJS & Mapnik
• Docker & Apex.io (Lambda helpers)
• Data source (DEM, Aerial Imagery, PostGIS, etc.)

AWS Tiler Benefits

• Cost Efficient
• You Own Your Tiles
• No Tile Server
• Works Well with Raster Data
• Can Use Multiple Layers on a Single AWS Setup

AWS Tiler Caveats

• Set up & AWS configuration can be tricky
• Initial Tile generation is slow

Impractical Uses of Terrain Data