A vast volume of data

The boom of Earth observation by satellite has provided users with ready access to continuous time series data. These are obtained on a repetitive basis all over the world using optical sensors such as the VEGETATION instrument on board of the SPOT satellites. As such continuous recording inevitably produces a huge volume of data there is a need for automatic processing methods to be able to analyse them and extract the useful information. Two concerns are particularly important when developing concrete applications:
• on the one hand, to produce cloud-free images from these data to be able to examine any location on the Earth’s surface;
• on the other hand, to detect indications of changes to the vegetation in order to monitor them in real time.

The project was based on six full years of daily worldwide recordings. This generated an impressive volume of over 14 terabytes (1 TB = 1024 Gigabytes) of data to draw upon. The focus of this project was to meet the day-to-day needs of potential users in the field, which is why close scientific cooperation with a number of organisations took centre stage. One of these was the FAO(1) which got assistance in the active monitoring of locusts that cause such devastation to crops through the detection of conditions that favour their proliferation. The project also co-operated with NASA to identify the behaviour of small water bodies that are excellent indicators of approaching drought or flood risk. These research efforts reached a pre-operational stage with final products that are user-friendly and adaptable to specific users needs as well as to the conditions of the observed environments nearly at hand.

(1) United Nations Food and Agriculture Organisation

Automatic and global processing methods are needed to manage the large volume of continuous time series data. These must contribute to the production, at regular intervals, of clear and cloud-free composite satellite images. In this context, the project focused on one hand on the temporal synthesis of a large volume of data obtained from the VEGETATION instruments and, on the other hand, on the ability to detect any indication of significant change.