A progress report on the application of pattern recognition techniques in NWP and other quantitative forecasting methods
Pankiewicz, G.
This report describes progress that has been made on the application of pattern recognition techniques in using satellite imagery for NWP and other quantitative forecasting methods. An overview of initial achievements that were made in developing tools and undertaking pilot studies up to 1997 is provided. During 1997, a strategy report was written and discussed at the Expert Group on the use of satellite Imagery in NWP, which directed the work undertaken during 1998 and 1999. Recent developments include the use of a neural network cloud classifier in the operational version of the GANDOLF flood forecasting system; production of a probability of precipitation classifier for use in removing spurious radar echoes in Nimrod; and an investigation into the use of a neural network for transforming Meteosat water vapour imagery to potential vorticity, for adjusting initial NWP conditions. Further work is discussed, with seven applications which may have potential significant operational impact at the Met. Office in the next few years.
Accès à la notice sur le site du portail documentaire de Météo-France