This is a global dataset providing basemap elevation coverage, tiled for easy usage. Tiles are available in multiple output formats, optimized for different use cases.

Data is aggregated from several open data providers including 3 meter and 10 meter 3D Elevation Program (3DEP) in the United States, 30 meter Shuttle Radar Topography Mission (SRTM) globally, and coarser Global Multi-resolution Terrain Elevation Data (GMTED) and ETOPO1. Data updates are provided by Mapzen.

Data is available in the "elevation-tiles-prod" Amazon S3 bucket. There are several file formats for the data:

Terrarium format PNG tiles contain raw elevation data in meters, in Mercator projection (EPSG:3857). All values are positive with a 32,768 offset, split into the red, green, and blue channels, with 16 bits of integer and 8 bits of fraction. To decode:

(red * 256 + green + blue / 256) - 32768

Normal format PNG tiles are processed elevation data with the the red, green, and blue values corresponding to the direction the pixel “surface” is facing (its XYZ vector), in Mercator projection (EPSG:3857). The alpha channel contains quantized elevation data with values suitable for common hypsometric tint ranges. High alpha channel values indicate lower elevation values (below sea level), making them more opaque.

Specifically, normal format alpha values are counted in (floored) elevation increments. Below sea level they start at -11,000 meters (Mariana Trench) and range to -1,000 meters in 1,000 meter increments, with more detail on the coastal shelf at -100, -50, -20, -10 and -1 meters and finally 0 (intertidal zone). Values above sea level are reported in 20 meter increments to 3,000 meters, then 50 meter increments until 6,000 meters, and then 100 meter increments until 8,900 meters (Mount Everest).

GeoTIFF format tiles are raw elevation data suitable for analytical use and are optimized to reduce transfer costs in 512x512 tile sizes but with internal 256x256 image pyramiding, in Mercator projection (EPSG:3857). Allow for the larger tile size by referring to the tile coordinate of {z-1} parent tile.

Skadi format tiles are raw elevation data in unprojected latlng (EPSG:4326) 1°x1° tiles, used by the Mapzen Elevation Service. Essentially they are the SRTMGL1 format tiles but with global coverage and compressed using gzip. See the SRTM guide for exact format specifications.

For more details, see Mapzen's documentation page here.

You can access any of the tiles directly using the following endpoints:





Category Elevation data
License Attribution Required
Storage Service Amazon S3
Location s3://elevation-tiles-prod in US Standard
Update Frequency New data is added based on community feedback

You can read more about Terrain Tiles on AWS in this interview with Mapzen.

Mapzen’s elevation service is an open-source web API (and C++ library) that provides digital elevation model (DEM) data as the result of a query. The elevation service data has many applications when combined with other routing and navigation data, including computing the steepness of edges or generating an elevation profile along a route.

To learn how to interact with the API, you can view the documentation here.

The elevatr package was written to standarize access to elevation data from a variety of sources, including Terrain Tiles on AWS. Available as an R package, you can now analyze powerful elevation data in a new way.

Learn more here.

Michael Fogleman used the DEM tiles to stitch together a highly-detailed rendering of the Grand Canyon. Including satellite imagery and rendering via OpenGL leads to a nice video tour of the area.

To see the video and explore the code used to render the data, see Michael's article here.

If you would like to show us what you can do with Terrain Tiles on AWS or would like to receive updates on the project, please fill out the form below.

Educators, researchers and students can also apply for free credits to take advantage of the utility computing platform offered by AWS, along with Public Datasets such as Terrain Tiles on AWS. If you have a research project that could take advantage of Terrain Tiles on AWS, you can apply for AWS Cloud Credits for Research.