ACRI-STstudies:

•   Satellite data assimilation:
  
  - ACRI-ST led the European Commission project MSDOL gathering SA/CNRS (Verrieres-le-buisson), FMI (Helsinki), IASB (Brussels), BAS (Cambridge) for the development of a chemical data assimilation system. Data assimilation is a technique that combines model forecasts and measurements to provide the best (in a statistical sense) estimate of the true state of the observed system, in our case the spatial and temporal distribution of ozone in the stratosphere. This project thus required the development of a model coupling the dynamics and the chemistry in the stratosphere (based a model originally developped at NCAR), and an "optimal interpolation" module that perform the statistical analysis.
  
  
- ACRI-ST, on behalf of CNES, the French space agency, led the OZVAL project, with Service d'Aeronomie du CNRS(Verriêres-le-Buisson), observatoire de Bordeaux and IPSL (Paris). It aimed at supporting the validation of the atmospheric chemistry mission of the European ENVISAT satellite using data assimilation. Three instruments (GOMOS, MIPAS, SCIAMACHY) indeed provide continuous measurements of various trace gases in the stratosphere. Data assimilation allows to interpolate them in a way consistent with the physics of the system to ease the comparison with ground-based measurements that are not collocated. To do so, ACRI-ST is member of the ESA Atmospheric Chemistry Validation Team. Within the framework of this project, ACRI-ST also provide the community with forecasts of the potential vorticity field using the isentropic transport model MIMOSA from SA/CNRS.
http://www.enviport.com


•   UV irradiance reaching the ground:


- ACRI-ST led the European Commission project UFOS (Ultraviolet Forecasting Operational Service) aiming at providing analyses and forecast of the UV index over the Mediterranean area. Assimilated observations of the ozone content of the atmosphere are used along with meteorological analyses and forecasts to compute the UV indice. In order to speed up the forecast, the radiative transfer is approximated by a neural network trained on Modtran outputs. The forecasts are continuously distributed on the Web for two years.
http://www.enviport.com
http://www.acri-st.fr/ufos