OrionAI now reads the land itself

Many of the questions that matter most, such as where a forest is thinning, which neighbourhoods sit low beside a river or what drains into a reservoir, can't be answered from a list of features. You need to read the land itself: satellite images, elevation and the way water moves.

OrionAI in Map Studio can now do this for you. Ask in plain words and it picks the data and the steps, runs them and maps the answer.

Satellite imagery, cloud-free. OrionAI builds a clean image of your area from Sentinel-2 (10 m, since 2017) or Landsat (30 m, since 1984) for the dates you choose, then calculates any of about 280 spectral indices: NDVI for vegetation, NDWI for water, NDBI for built-up land and many more.

Change you can see. "Map vegetation change in the Mau Forest between 2019 and 2024" gives a change map and, for each area, the share that lost or gained green cover. You can also track an index month by month, or classify land cover and compare it between years.

Terrain and water. Slope, hillshade, contours and what can be seen from a point, from the Copernicus 30 m elevation model. Trace the stream network, delineate the watershed that drains into Masinga Dam, or screen which parts of Nairobi lie within 2 m of a drainage line.

Values for every area. Population, rainfall, elevation or land cover summarised by county, ward or any area on your map, ready to style and compare.

Results that are grids of values are drawn as image layers, and they are yours alone until you publish them. As before, a request uses one credit, however many steps it takes.

Read the full OrionAI 2.10.0 release notes.