Advanced space technologies and processing techniques supporting disaster management at CNES
Since its inception, 25 years ago, the International Charter: Space and Major Disasters has been reflecting the evolution of space technologies. Initially supported by three satellites and 10-meter resolution images with a revisit time of several days, the Charter aiming at supporting countries facing disasters through the provision of satellite data benefits today from 270 satellites offering very high-resolution images, a diversity of sensors, and intraday revisit capability. This wealth of data presents a challenge regarding processing capabilities but also an opportunity to improve disaster management, especially large-scale disasters and types of disasters such as the flood that occurred in Valencia, Spain, last year. Let’s review the technological advancements developed at CNES, the French Space Agency, that can be harnessed to support crisis management.
Technological Innovations and Applications
Among the technological innovations that are helpful to support disaster relief, advanced Earth Observation missions recently launched or soon to be launched will offer new types of measurements such as 3D capability or infrared observations. Among them, the CO3D constellation developed by CNES was launched in July 2025 and enables the acquisition of synchronous stereo images of Earth. The CO3D data (both images and Digital Elevation Models (DEM)) will be integrated into the Charter operations and will be useful for instance, to better assess damages in urban areas, to study geological hazards or model flood propagation. Let’s note that 3D value added products are already available in the Charter, with the flood-depth product developed by SERTIT, interpolating water surface elevation from sample point altitudes along the flooded area’s boundaries. Precise 3D Models that will be generated by CO3D will allow to improve the accuracy of flood depth estimations.
New products based on thermal infrared data that will be delivered by the ISRO-CNES TRISHNA (Thermal infrared Imaging Satellite for High-resolution Natural resource Assessment) mission are also being investigated for the purpose of fire management. This satellite mission will observe Earth's surface in different infrared and visible bands. Enabling monitoring land surface temperature or identifying vegetation suffering from hydric stress, TRISHNA data will be useful to map active fires, burnt areas, and vegetation indirectly affected by fires and anticipate fire propagation.
Needless to say, sophisticated measurements require sophisticated processing and high-level expertise. Data processing chains are carefully designed to take into account the mission characteristics as well as user needs. These processing chains allow the radiometric and geometric calibrations which are essential for the elaboration of value added products as reliable information is also needed for the purpose of crisis management.
With the advent of artificial intelligence techniques, the extraction of information from the calibrated data is being facilitated. In the framework of the celebration of the 25th anniversary of the Charter, an AI Challenge was organized in collaboration with ESA's Φ-lab, which engaged different teams around the world in a competition to provide fast and efficient damage assessment algorithms. This experience has proven that AI techniques can support rapid mapping tasks. As an example, the CNES Earth Observation Laboratory is currently developing a tool called PICANTEO, which has just been released as an open-source Alpha version on GitHub. This tool is a framework designed to make it easy to build or use pre-configured change detection pipelines. It includes, in particular, a 2D optical very high-resolution change detection pipeline capable of identifying destroyed buildings while providing an associated uncertainty estimate. When suitable image acquisitions are available, a 3D change detection pipeline can be used to complement and reinforce these detections, incorporating an uncertainty assessment based on the concept of stereoscopic ambiguity. The application of AI techniques cannot replace human expertise but will reduce the human photointerpretation workload by producing a first guess assessment of damages and thus allow faster delivery of information to the Charter users.
In the near future, these new technologies will be integrated into Charter operations. Indisputably, technologies will bring added value for monitoring and understanding disaster situations. However, it should not be forgotten that humans are always involved at various stages of the processing chains and decision-making process, whether they are end users or space technology experts. That is why capacity building and activities to support user uptake of technologies are crucial for effective disaster management.