Skip to main content

Sentinel-2: So much data, so little time

With the launch of the second satellite at the end of this year, the system that Thales has built in partnership with Thales Alenia Space will be processing a record two-and-a-half terabytes[1] of data every day

The Sentinel-2 mission’s two optical satellites will continuously scan Earth, collecting imagery of all of its land surfaces at resolutions of 10 to 60 metres every five days. This imagery will be used chiefly to monitor vegetation, water cover, soils and coastal areas, but also to observe and aid response to natural disasters such as floods, volcano eruptions and landslides.
On each satellite pass, every 50 minutes, acquired images will be downloaded to three receiving stations in Matera (Italy), Svalbard (Norway) and Maspalomas (Spain). Each download will take 12 minutes, for a total of one terabyte of data every day.

Technical challenge

This telemetry, compressed and encrypted to optimise transmission, first has to be “decommutated” to obtain the raw images. These images then undergo two types of processing: first, they are radiometrically corrected to match and smooth brightness values (correcting for such effects as glare, dropped rows of pixels and so on); and second, they are geometrically corrected to position them precisely. What’s more, all of these processing operations will need to be performed between satellite passes on vast volumes of data, calling for a great deal of computing power.
The processed imagery will then be forwarded immediately to two data archiving centres in Madrid (Spain) and Farnborough (UK) for viewing and downloading by end-users. The volume of imagery generated and made available for download every day will exceed two and a half terabytes.
Handling such a profusion of data round the clock for the seven-year lifetime of the Sentinel-2 mission represents a major technical challenge, one that Thales’s teams—drawing on the experience acquired on previous programmes like Gaia (see here), Pleiades and CSO —are ready to meet.
After the launch of the first satellite, the three-month in-orbit checkout phase served to make final adjustments to the system before rolling it out. Ultimately, all European citizens will be able to view and download imagery from the Sentinel-2 mission.

Thales in Sentinel-2

Thales is working with Thales Alenia Space, the joint subsidiary of Thales and Finmeccanica, on the Sentinel-2 Payload Data Ground Segment (PDGS).

For this project, Thales is in charge of:

- the Data Processing Centre, covering satellite telemetry to image production
- engineering and validation of subcontracted image processors
- baseline industrial platforms in France and Italy
- roll-out of data processing and distribution centres in Spain (Madrid and the Canary Islands), France, Italy, Norway (Svalbard) and the United Kingdom

Thales is also sharing responsibility for integration of the Sentinel-2 PDGS with Thales Alenia Space, the prime contractor tasked by the European Space Agency (ESA).

More at:
Sentinel 2: Facts & Figures
Copernicus - A European programme to take the pulse of the planet in real time

Designed to address the challenges of monitoring the environment and assuring the security of people and property, the Copernicus programme (formerly GMES)—led by the European Union and developed in partnership with the European Space Agency (ESA) and its member states—aims to give Europe its own independent Earth-observation and monitoring capability, providing a complete, continuous and global picture of the planet’s health.

The programme’s space component is built around a constellation of Sentinel satellites: Sentinel-1 (day-night radar imaging of land surfaces and oceans), Sentinel-2 (medium-resolution optical land monitoring), Sentinel-3 (multispectral, infrared and altimetric monitoring of oceans and land surfaces), Sentinel-4 (monitoring of atmospheric pollution from geostationary orbit) and Sentinel-5 (observation of atmospheric chemistry from low Earth orbit).
More at


[1] 1 terabyte = 1,000,000,000,000 bytes (1012 bytes) = 1,000 gigabytes