Skip to content

Overview

Geo Engine is a server for processing and visualizing geospatial data. It reads raster and vector data, combines and transforms them in processing graphs, and delivers the results as maps, files, data frames or plots. Every piece of data has a time dimension, so time series are first-class citizens. Results are computed on demand and streamed, which means Geo Engine can work on datasets much larger than the available memory.

All clients talk to the same server through its REST API.

Who Interface Typical tasks
Data scientists geoengine Python library, often in Jupyter Build processing graphs, load results into GeoPandas or xarray, train ML models
Domain users Web apps: gis for exploring and analysing data, eodyssey, and domain-specific dashboards Browse layers, combine data with operators, view time series on a map
Administrators manager web app Manage datasets, layers, data connectors, users, roles and permissions
GIS software (e.g., QGIS) OGC services: WMS, WFS, WCS, OGC API Tiles Display and download Geo Engine results in existing tools
Developers of other software REST API with generated clients for Python, Rust and TypeScript Integrate Geo Engine into their own applications

Geo Engine reads files through GDAL/OGR, so it supports most raster and vector formats. Data enters Geo Engine in three ways:

  • Internal datasets are files that the server knows about, e.g., GeoTIFFs or GeoPackages on a volume. See Datasets.
  • Uploads are files that users send to the server and turn into their own datasets.
  • Data connectors connect to other systems, such as STAC catalogs, GBIF or Pangaea, and make their data available without copying it. See Data Connectors.

Layers and layer collections organize all of this data for browsing.