The quantitative precipitation forecasting is one of the main challenges of the modern meteorology. Several systems were designed and applied to this aim, based on statistical, dynamical and hybrid techniques (Johnson et al., 1998; Gaudet et al., 1998). The basic input for hydrological modelling is the precipitation field; its temporal and spatial behaviour must be accurately estimated and forecast for the success of the hydrological simulations (Pererira Fo. et al., 1999a). Hydrological forecasting systems are being applied which receive as input data not only the observed (estimated) but also the forecast rainfall (Pererira Fo. et al., 1999b). This means a considerable improvement in the prediction capacity of actual distributed rainfall-runoff models. RAMS is one of the most popular state-of-the art atmospheric modelling systems whose applications range from the simulation of the air flow in urban canyons to the planetary climate modelling (Pielke et al., 1992).The ability of RAMS to produce accurate rainfall forecasts was investigated in other works (e.g. Gaudet et al., 1998). The data used for the analysis of the quantitative rainfall forecasts were collected by a rain gauge network over the study basins in Tuscany, Italy. The area has a complex orography, extending from the plain to the Appennine chain with peaks over 1500 m a.s.l.. The three basins are the Casentino, Valdarno and Sieve, about 1000 km2 each in size. Nine rain gauges are present over the Casentino and Valdarno basins, thirteen rain gauges over the Sieve basin . The climate is typical of the northern Mediterranean area, hot and dry in summer, mild and rainy in autumn and spring, moderately cool and dry in winter. The analysis covers the period September 1, 1999 to April 30, 2000, when 164 days data were available for the verification. The daily (hours 00-23 UTC) rainfall data were averaged over any sub-basin by means of a simple arithmetic average. The quantitative precipitation forecasts (QPF) produced by the model RAMS during its operational runs, up to the lead time of thirty-six hours, in the study period were considered for the comparison against the observed rainfall. The rainfall fields were produced over a regular grid domain of elementary size 8 km covering a part of central-northern Tuscany and nested into a 40 km grid resolution domain covering most of central-western Mediterranean . Due to the averaging process leading to the dynamic and thermodynamic equations of model RAMS, the rainfall at any grid node is to be meant as a grid mesh area average value. After this consideration, the daily (hours 00-23) distributed rainfall is averaged over each of the target sub-basins by means of an accurate superimposition of the sub-basins in raster format at 30 m resolution to the quantitative precipitation forecasts domain.
Optimal configuration of the very high resolution atmospheric model RAMS as a component of an Hydrological forecasting system
Grifoni D;
2000
Abstract
The quantitative precipitation forecasting is one of the main challenges of the modern meteorology. Several systems were designed and applied to this aim, based on statistical, dynamical and hybrid techniques (Johnson et al., 1998; Gaudet et al., 1998). The basic input for hydrological modelling is the precipitation field; its temporal and spatial behaviour must be accurately estimated and forecast for the success of the hydrological simulations (Pererira Fo. et al., 1999a). Hydrological forecasting systems are being applied which receive as input data not only the observed (estimated) but also the forecast rainfall (Pererira Fo. et al., 1999b). This means a considerable improvement in the prediction capacity of actual distributed rainfall-runoff models. RAMS is one of the most popular state-of-the art atmospheric modelling systems whose applications range from the simulation of the air flow in urban canyons to the planetary climate modelling (Pielke et al., 1992).The ability of RAMS to produce accurate rainfall forecasts was investigated in other works (e.g. Gaudet et al., 1998). The data used for the analysis of the quantitative rainfall forecasts were collected by a rain gauge network over the study basins in Tuscany, Italy. The area has a complex orography, extending from the plain to the Appennine chain with peaks over 1500 m a.s.l.. The three basins are the Casentino, Valdarno and Sieve, about 1000 km2 each in size. Nine rain gauges are present over the Casentino and Valdarno basins, thirteen rain gauges over the Sieve basin . The climate is typical of the northern Mediterranean area, hot and dry in summer, mild and rainy in autumn and spring, moderately cool and dry in winter. The analysis covers the period September 1, 1999 to April 30, 2000, when 164 days data were available for the verification. The daily (hours 00-23 UTC) rainfall data were averaged over any sub-basin by means of a simple arithmetic average. The quantitative precipitation forecasts (QPF) produced by the model RAMS during its operational runs, up to the lead time of thirty-six hours, in the study period were considered for the comparison against the observed rainfall. The rainfall fields were produced over a regular grid domain of elementary size 8 km covering a part of central-northern Tuscany and nested into a 40 km grid resolution domain covering most of central-western Mediterranean . Due to the averaging process leading to the dynamic and thermodynamic equations of model RAMS, the rainfall at any grid node is to be meant as a grid mesh area average value. After this consideration, the daily (hours 00-23) distributed rainfall is averaged over each of the target sub-basins by means of an accurate superimposition of the sub-basins in raster format at 30 m resolution to the quantitative precipitation forecasts domain.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
