Site-Specific Forecasting using Ensembles of Coupled 1-D/3-D Model Simulations
John Zack, President and Sr. Research Scientist of Meso, Inc., New York, U.S.
Sala de Graus de la Facultat de Fisica, Diagonal 647, Barcelona
Divendres dia 17 d'abril a les 12:00
A method is currently being developed which has the potential to improve short-term local forecasting by routinely producing probabilistic forecasts instead of the usual deterministic forecasts from a mesoscale model. The 3-D/1-D coupled approach combines a 3-D mesoscale model and a 1-D PBL model in order to "scale-down" mesoscale model results to a particular site which may have surface characteristics quite different from the mesoscale model grid box enclosing it. An ensemble of 1-D runs are made to express the results probabilistically, in an attempt to account for the uncertainties of the mesoscale simulation and/or those local surface characteristics which may be poorly known. This method can be understood as a way to extract additional information from a single mesoscale model simulation by an ensemble of subgrid-scale forecasts. An application of the method to the forecast of canopy temperature for an agricultural crop canopy will be described.