Logo du site
Projet ANR FireCaster | UMR SPE 6134
The development of operational services dedicated to mitigate natural and anthropogenic risks is one of the objectives of this ANR call. In the frame of a previous project already led by SPE lab, IDEA (2010-2013 Forest-Fires, from combustion Emissions to Atmospheric transport), considered as a highlighted project by the ANR, several demonstrators dedicated to wildfire risk were developed (codes, approaches, services), aiming at proposing a new generation fire decision support system. Available thanks to recent technological advances in the field of meteorology, data assimilation, fire modeling and supercomputing these tools have only been tested and partially validated on a limited amount of case studies and over limited time periods. The goal of the FireCaster project is to extent the approaches at the national scale by prototyping a platform  that  allows  to  estimate  upcoming  fire  risk  (H+24  to  H+48)  and  in  case  of  crisis,  to  predict  fire  front  position  and  local  pollution  (H+1 to H+12).
The main challenge is here to deliver these new diagnostics immediately for any given territory and at any given forecast date. It requires to overcome a key issue: access to high resolution (50m) fuel models and data. In order to characterize these fuels and potential pollution products, it is planned to use new vegetation atlas and study smoke emissions for various fuel types and states of fuel.
These models will then be generalized to the whole French territory, not by developing specific codes, but for the first time by linking them to surface models, which simulate energy exchanges and water cycle in meteorological models. Surface models recently had a strong increase in resolution and accuracy that makes this link possible (SURFEX model -CNRM- operational).
In terms of risk, we propose a probabilistic approach, based on large sets of perturbed multi-model simulations (INRIA), to determine the distribution of potential fire sizes. This approach will provide a new diagnostic (fire burnt area) very different from the current indicator (risk of fire ignition with no indication on the potential size).
Fire fighting tools should help to estimate the benefits and risks of each intervention scenario as planned by crisis management centres. They should also evaluate the impact of fires on air pollution and smoke for fire-fighters and population alert.
Probability impact maps for each fire fighting scenario, showing areas where the passage of fire is highly expectable, will be obtained by ensemble simulations, taking into account interactively fire-fighting actions. Another, deterministic and more detailed coupled Fire/Meso-NH atmospheric crisis model will determine front position, smoke pollution and local micro-meteorology with data assimilation of aerial/spaceborne observation of fire contours; it will be implemented (CECI) to reduce the uncertainty of these deterministic predictions. Eventually, in order to link the resulting computations to innovative indicators, economic, human and environmental costs will be evaluated (LISA).
At the national French level, Météo-France, responsible for this public service mission, will supervise and test the project and evaluate the products within a steering committee also composed of the European Forest Fire Information System (JRC), the National Forestry Services (ONF), the French government space agency (CNES) and Corsican fire brigades (to test crisis tools). While the project success first requires a successful application at national scale, there exists a strong potential of development at the European level. All codes will be Open-Science with French SMEs interested in selling the knowledge required to apply the platform to other countries or areas in the frame of SAFE Cluster (former Pôle Risques).
Page mise à jour le 10/02/2017 par JOHAN JOUVE