LIST researchers have developed a new modular analysis tool for remote sensing data within an ESA funded project.
As we look up to the sky, more than 2000 satellites are working around the Earth. Among them, Earth observation satellites record and map our planet at scales of accuracy down to the metre. Huge amounts of data are generated every day, providing valuable information for monitoring, for example, the occurrence of natural disasters, or assessing the impact of climate change on our ecosystems.
Remote sensing data are a valuable mean for both global and accurate monitoring of a given location over time. However, they are massive in volume and need to be analysed and translated into understandable information for end-users. This is where the EOTSA Toolbox (Earth Observation Time Series Analysis) developed by Martin Schlerf and Ulrich Leopold together with Thomas Udelhoven from the University of Trier and Benedikt Gräler from 52° North, comes in.
Users of the EOTSA web interface can analyse and visualise the evolution of the vegetation status within a selected area on the globe over time. The change of vegetation status over time can be linked to many processes at the land surface, such as land degradation, forest fire, change in agricultural practices, etc.
Already praised by the European Space Agency (ESA) and spotted by the World Food Programme, the EOTSA Toolbox has confirmed its usefulness in several application areas through different use cases. Researchers from the Academy of Science in Bulgaria have successfully applied this system for burnt area detection in Bulgaria. Another use case is phenological trends across the African continent. In other words, any user of the interface can visualise, at different levels, the evolution of vegetation as a function of seasonal and interannual climate variations, land use change, and disturbances.
"Through this ESA-funded project, we worked on the development, implementation and demonstration of a toolbox for analysing time series of satellite data," explained Ulrich. Their innovative tool focused specifically on analysing data from so called PROBA-V Earth observation satellite, which is the predecessor of the Sentinel-3 satellites. The main PROBA-V (Project on on-board autonomy – Vegetation) payload is the Vegetation instrument: a multispectral sensor with four spectral bands and a very large swath of 2285 km to guarantee daily coverage above 35° latitude.
One of the great advantages of this tool made by LIST is that it directly accesses the satellite archives. In other words, there is no need to download the data and store it locally. In addition, the interface offers multiple reading modes that can be adapted according to the profile and needs of the users. "A non-expert can produce a map showing areas where vegetation is declining in time, for example because of forest degradation. At the same time, a more experienced user can access a greater level of detail, e.g. looking for shifts in the annual green peak as indicator of land-use change (for instance, rotating crop type or changing irrigation practice)”, said Martin. Indeed, users can either benefit from EOTSA’s pre-designed workflows or code their own workflow using the existing functions and algorithms.
While the EOTSA toolbox is already fully operational with PROBA-V satellite data, it is expected to be extended to include Sentinel-2 satellite data in the near future. The developments of this collaborative interface do not stop there, however, and are aimed at increasing the resolution of the images to 30 metres, as well as pursuing new use cases in partnership with the European Space Agency and other globally operating partners, such as the World Food Programme.