New generation of radiative transfer models for climate studies based on neural networks

We demonstrate that neural networks can be successfully used for accurately deriving the longwave radiative budget from the top of the atmosphere (TOA) to the surface. The reliable sampling of the Earth atmospheric situations in the TIGR dataset developed at LMD allows for an efficient learning of the neural networks. The dramatic saving of computing time based on the neural networks technique allows for using more sophisticated (hence more accurate) radiative schemes for computing the longwave radiative budget either in GCM simulations or from long time series of satellite observations such as those provided by the 16 years of measurements of the TOVS sounder aboard the NOAA operational satellites.