Toward Real-Time Daily PQPF by an Analog Sorting Approach: Application to Flash-Flood Catchments

摘要: Heavy-rainfall events are common in southern France and frequently result devastating flash floods. Thus, an appropriate anticipation of future rainfall is required: for early flood warning, at least 12‐24 h advance; alerting operational services, 2‐3 days ahead. Precipitation forecasts generally provided by numerical weather prediction models (NWP), their associated uncertainty estimatedthroughan ensemble approach. forecastsalso have tobe adaptedto hydrologicalscales. This study describes alternative approach to commonly used limited-area models. Probabilistic quantitative precipitation (PQPFs) through analog sorting technique, which directly links synoptic-scale NWPoutput catchment-scale rainfallprobabilitydistributions. One issue concerns the latest developments implementing a daily version this technique into conditions. It shown that obtained PQPFs depend on meteorological selecting analogous method has be reoptimized when changing source synoptic forecasts, because NWP output uncertainties. Second, evaluation demonstrates performs well warning heavy-rainfall provides useful information as potential input hydrological system. distributions can unreliable. A statistical correction observed bias proposed function no-rain frequency values, leading significant improvement PQPF sharpness.