Whatever region of the world is affected, it is currently difficult to assess flood hazard, in particular because of models that are too vague and a lack of information. Yet, it is one of the biggest natural risks in terms of frequency and impact, not to mention human and economic loss.
Therefore, it is necessary to make better use of the data from Earth observation satellites, available at the global level, to make forecasting models more reliable and the information that ensues from them more accurate.
In order to better evaluate flood risk in different regions in the world, partners of the CASCADE project will devise a new modelling chain, based, in particular, on satellite Earth observation. This chain aims to combine two models, which are considered successful today but are often used independently of one another: hydrological and hydraulic models. While the first type of model allows river discharge to be predicted based on meteorological variables and the physiographical characteristics of the catchment basin, the second type aims to define the way in which the water will be dispersed along the flood plains, considering elements such as the geometry of the riverbed. To put such models in place, a certain amount of information is necessary, such as the geometry of the waterway or the parameters determining the water flow and its storage in the soil. Unfortunately, this information is not available at the global scale and the setting-up of the hydrological and hydraulic models necessary for mapping out flood hazard is proving to be a sometimes difficult task. With CASCADE, the partners intend to use the long satellite image time series of the state of soil saturation and flooded areas, using the compilation initiated in several previous projects, in order to develop a flood prediction system that could be applied in diverse regions of the world.
In this way, the research team will configure the models by assimilating satellite data on several study sites. The River Severn, in the United Kingdom, which regularly experiences severe flooding and for which a comprehensive history of information is already available, will be used to develop and validate the scientific methodology. Once they have been fixed, the methods developed will be transferred and tested, for example at the Zambezi Basin (1,330,000 km2), which is also regularly affected by flooding but for which there is very little data.
The project should allow more reliable models to be created, allowing large-scale flood risks to be better estimated in different areas all over the planet, thanks to Earth observation information. The application of these models to large basins at risk around the world will thus allow not only the probability of the occurrence of flooding, but also the extent of the individual floods to be better determined in order to better comprehend the dangers.
With this new tool, it will be easier to determine the location of the populations most at risk and anticipate future developments in order to have a better knowledge of the risk, alert the populations at risk of flooding several days in advance, or even, in the most extreme cases, to better plan humanitarian aid.