Skip to main content

Sentinel-2 Explorer for Indices and Time Series

The Sentinel-2 Composite canvas builds one clear image of a place. The Explorer goes further. Pick a place and dates, then map a vegetation or water index, build a composite, or follow an index through time on a grid of H3 hexagons. Images, hexagons and series can be downloaded, and one API call returns the same data.

Try it out​

Open the Canvas →

tip

No keys or account are needed to call the API on this page.

1 · Explorer​

The Explorer widget has three tabs:

  • Index map: an image of an index such as NDVI, NDWI or NBR. It uses the least cloudy pixels in the date range and shows statistics for the area.
  • Time series: the area as H3 hexagons, with a chart of the area average. Click a hexagon to chart its own series, and click a day in the chart to map that day.
  • Composite: a cloud-free median, mean, min or max of every clear pixel. Show it as true colour, false colour, SWIR and other presets, as an index, or as any three bands you choose.

Index map can show the result as an Image or as H3 hexagons. Composite can too, when it shows an index or one band. The download button at the bottom left saves what the map or chart shows. An image downloads as a GeoTIFF (an index keeps its values, bands are reflectance from 0 to 100) or as a PNG exactly as drawn. Hexagons download as CSV or Parquet.

The area can be at most 0.5° wide and 0.5° tall.

2 · One API for every product​

sentinel_api is the single entry point. Send it a JSON config with a product and its params:

{"product": "index_map", "params": {"bounds": [-93.25, 41.95, -93.15, 42.02], "date_range": "2024-07-01/2024-07-31", "index": "NDVI"}}
ProductReturns
index_mapAn index image and area statistics
compositeA cloud-free composite of a preset, an index or chosen bands
index_timeseriesMean, min, max and standard deviation of an index per interval
scene_searchThe Sentinel-2 scenes over the area
hex_mapH3 hexagons of an index for one day or all days
hex_seriesThe series of one hexagon, or of the whole area
series_exporthex_series as a CSV or Parquet file
hex_cellsThe table behind the three hex products

Supported indices are NDVI, NDWI, MNDWI, NDMI, NBR, NDSI, NDRE, EVI, SAVI and BSI. Send {"product": "list"} to see every product with its parameters and defaults.

Find the scenes over a farm field in Iowa in July 2024:

curl -G "https://udf.ai/fc_fused/Sentinel2_Explorer/sentinel_api/run/file" \
  --data-urlencode "format=json" \
  --data-urlencode 'config={"product":"scene_search","params":{"bounds":[-93.25,41.95,-93.15,42.02],"date_range":"2024-07-01/2024-07-31","max_cloud":20}}'

Get the area-average NDVI every 10 days in summer 2024:

curl -G "https://udf.ai/fc_fused/Sentinel2_Explorer/sentinel_api/run/file" \
  --data-urlencode "format=json" \
  --data-urlencode 'config={"product":"hex_series","params":{"bounds":[-93.25,41.95,-93.15,42.02],"date_range":"2024-06-01/2024-08-31","index":"NDVI","interval":"P10D"}}'

Add "fmt" to an index_map or composite config to get only its file, as a row with a signed download_url valid for one hour. Use "geotiff" or "png" for an image, and "csv" or "parquet" for hexagons. series_export takes fmt inside params instead, and scene_search and the other products have no file:

curl -G "https://udf.ai/fc_fused/Sentinel2_Explorer/sentinel_api/run/file" \
  --data-urlencode "format=json" \
  --data-urlencode 'config={"product":"index_map","params":{"bounds":[-93.25,41.95,-93.15,42.02],"date_range":"2024-07-01/2024-07-31","index":"NDVI","max_cloud":60},"fmt":"geotiff"}'

3 · How the hexagons are built​

The first time you ask for a place and dates, sentinel_core reads the Sentinel-2 scenes in parallel, one job per 10 days with .map(). It keeps, for each clear day and each small H3 cell, the clear-pixel count and the mean band values. hex_map, hex_series and series_export then read this table, so changing the index, the statistic (mean, median, min, max, std, p10, p90) or the hexagon size is quick.

About the data​

PropertyValue
ImagerySentinel-2 L2A, AWS Open Data
OutputGeoTIFF or PNG for images, CSV or Parquet for hexagons and series
ContainsModified Copernicus Sentinel data

See also​