s u m m a r y
Accurate precipitation forecasts are required for accurate flood forecasting. The structures of different
precipitation forecasting systems are constantly evolving, with improvements in forecasting techniques,
increases in spatial and temporal resolution, improvements in model physics and numerical techniques,
and better understanding of, and accounting for, predictive uncertainty. Hence, routine verification is
necessary to understand the quality of forecasts as inputs to hydrologic modeling. In this study, we verify
precipitation forecasts from the National Centers for Environmental Prediction (NCEP) 11-member Global
Ensemble Forecast System Reforecast version 2 (GEFSRv2), as well as the 21-member Short Range
Ensemble Forecast (SREF) system. Specifically, basin averaged precipitation forecasts are verified for different
basin sizes (spatial scales) in the operating domain of the Middle Atlantic River Forecast Center
(MARFC), using multi-sensor precipitation estimates (MPEs) as the observed data. The quality of the
ensemble forecasts is evaluated conditionally upon precipitation amounts, forecast lead times, accumulation
periods, and seasonality using different verification metrics. Overall, both GEFSRv2 and SREF tend
to overforecast light to moderate precipitation and underforecast heavy precipitation. In addition, precipitation
forecasts from both systems become increasingly reliable with increasing basin size and decreasing
precipitation threshold, and the 24-hourly forecasts show slightly better skill than the 6-hourly
forecasts. Both systems show a strong seasonal trend, characterized by better skill during the cool season
than the warm season. Ultimately, the verification results lead to guidance on the expected quality of the
precipitation forecasts, together with an assessment of their relative quality and unique information content,
which is useful and necessary for their application in hydrologic forecasting.
s u m m a r yAccurate precipitation forecasts are required for accurate flood forecasting. The structures of differentprecipitation forecasting systems are constantly evolving, with improvements in forecasting techniques,increases in spatial and temporal resolution, improvements in model physics and numerical techniques,and better understanding of, and accounting for, predictive uncertainty. Hence, routine verification isnecessary to understand the quality of forecasts as inputs to hydrologic modeling. In this study, we verifyprecipitation forecasts from the National Centers for Environmental Prediction (NCEP) 11-member GlobalEnsemble Forecast System Reforecast version 2 (GEFSRv2), as well as the 21-member Short RangeEnsemble Forecast (SREF) system. Specifically, basin averaged precipitation forecasts are verified for differentbasin sizes (spatial scales) in the operating domain of the Middle Atlantic River Forecast Center(MARFC), using multi-sensor precipitation estimates (MPEs) as the observed data. The quality of theensemble forecasts is evaluated conditionally upon precipitation amounts, forecast lead times, accumulationperiods, and seasonality using different verification metrics. Overall, both GEFSRv2 and SREF tendto overforecast light to moderate precipitation and underforecast heavy precipitation. In addition, precipitationforecasts from both systems become increasingly reliable with increasing basin size and decreasingprecipitation threshold, and the 24-hourly forecasts show slightly better skill than the 6-hourlyforecasts. Both systems show a strong seasonal trend, characterized by better skill during the cool seasonthan the warm season. Ultimately, the verification results lead to guidance on the expected quality of theprecipitation forecasts, together with an assessment of their relative quality and unique information content,which is useful and necessary for their application in hydrologic forecasting.
การแปล กรุณารอสักครู่..
