A monthly probabilistic forecasting system is experimentally operated at the ISAC institute of the National Council of Research of Italy. The forecasting system is based on GLOBO, an atmospheric general circulation model developed at the same institute. The model is presently run on a monthly basis to produce an ensemble of 32 forecasts initialized with GFS-NCEP perturbed analyses. Reforecasts, initialized with ECMWF ERA-Interim reanalyses of the 1989-2009 period, are also produced to determine modelled climatology of the month to forecast. The modelled monthly climatology is then used to calibrate the ensemble forecast of daily precipitation, geopotential height and temperature on standard pressure levels. In this work, we present the forecasting system and a preliminary evaluation of the model systematic and forecast errors in terms of nonprobabilistic scores of the 500-hPa geopotential height. Results show that the proposed forecasting system outperforms the climatology in the first two weeks of integrations. The adopted calibration based on weighted bias correction is found to reduce the systematic and the forecast errors.
First outcomes from the CNR-ISAC monthly forecasting system
D Mastrangelo;P Malguzzi;O Drofa;A Buzzi
2012
Abstract
A monthly probabilistic forecasting system is experimentally operated at the ISAC institute of the National Council of Research of Italy. The forecasting system is based on GLOBO, an atmospheric general circulation model developed at the same institute. The model is presently run on a monthly basis to produce an ensemble of 32 forecasts initialized with GFS-NCEP perturbed analyses. Reforecasts, initialized with ECMWF ERA-Interim reanalyses of the 1989-2009 period, are also produced to determine modelled climatology of the month to forecast. The modelled monthly climatology is then used to calibrate the ensemble forecast of daily precipitation, geopotential height and temperature on standard pressure levels. In this work, we present the forecasting system and a preliminary evaluation of the model systematic and forecast errors in terms of nonprobabilistic scores of the 500-hPa geopotential height. Results show that the proposed forecasting system outperforms the climatology in the first two weeks of integrations. The adopted calibration based on weighted bias correction is found to reduce the systematic and the forecast errors.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
