Earth Observation (EO)

Search, render, and analyze Sentinel satellite imagery for an area over time.

Tool Reference: 3 tools with full schemas →

Use Earth Observation (EO) to work with Sentinel satellite data through Copernicus Data Space. Find a clear scene before you render it, create a true-color or spectral-index image, or measure how vegetation and water changed over time.

EO works especially well with the Geo toolkit: use Geo to define an area of interest, then pass its GeoJSON boundary to EO. You can also work directly from a bounding box when you already know the coordinates.

Prefer video? This walkthrough shows how to create Copernicus credentials, connect EO in Toolforest, and start working with satellite imagery:

Example prompts

Show me a true-color satellite image of Golden Gate Park from the least cloudy week in June.

How did the vegetation in this park change from June through August? Show me monthly NDVI statistics.

Find the clearest Sentinel-2 scene over this lake last month and render it with water highlighted.

Setup

EO uses your free Copernicus Data Space account. Toolforest uses its Sentinel Hub credentials to request short-lived access tokens; you provide the client ID and client secret, not an access token.

Create Copernicus credentials

Create or sign in to a free Copernicus Data Space account.
Open the Sentinel Hub dashboard.
Open User Settings, find OAuth clients, and click Create.
Give the client a recognizable name, such as Toolforest EO, and use the Client Credentials grant type.
Copy the client ID and client secret when Copernicus displays them. The secret is shown only once, so save it before closing the dialog.

Configure EO in Toolforest

Open Toolforest and select the MCP server where you want to use EO.
Find Earth Observation (EO) and click Configure.
Paste the Copernicus client ID and client secret into their matching fields.
Save the configuration and add EO to the MCP server if it is not already connected.

⚠ Copernicus quotas and imagery limits

Sentinel imagery is not live and cloud cover can leave gaps in the data. Start with a small area and a narrow date range, use the catalog search to choose a clear window, and check your Copernicus account for current processing limits.

Tools

Tool	What it does
search_catalog	Find available satellite scenes over an area or bounding box, including acquisition time, cloud cover, and footprint. Use it to choose a suitable date before rendering or analyzing imagery.
get_statistics	Calculate an NDVI or NDWI time series for an area. Use NDVI for vegetation greenness and NDWI for water presence; results include mean, minimum, and maximum values by daily or monthly interval.
render_image	Render a shareable PNG in true color, false color, NDVI, or NDWI for a selected place and date range. The returned URL expires after about a day.

For every parameter and response schema, see the EO Tool Reference.

Notes and limits

Sentinel-2 imagery has 10 metre pixels, so EO is best for parks, farms, lakes, neighbourhoods, and other field-scale areas—not individual buildings.
Use a GeoJSON area of interest or a WGS84 bounding box. Provide one, not both.
search_catalog finds scenes before you spend processing capacity on a render or time series.
Image URLs are temporary. Download or save an image elsewhere if you need it after it expires.
EO supports Sentinel-2 by default. Other collections depend on the capabilities of your connected account.

Cookbook

EO cookbook examples are coming soon. Good first candidates are a least-cloudy scene search, a seasonal NDVI comparison, and a Geo-to-EO workflow that turns an OpenStreetMap boundary into a satellite image.