High-Resolution Simulations of the 2010 Pakistan Flood Event: Sensitivity to Parameterizations and Initialization Time

Abstract

stimating the risk of flood-generating precipitation events in high-mountain regions with complex orography is a difficult but crucial task. Quantitative precipitation forecasts (QPFs) at fine resolution are an essential ingredient to address this issue. Along these lines, the ability of the Weather Research and Forecasting (WRF) Model, operated at 3.5-km grid spacing, to reproduce the extreme meteorological event that led to the 2010 Pakistan flood and produced heavy monsoonal rain in the Indus basin is explored. The model results are compared with Tropical Rainfall Measuring Mission (TRMM) rainfall estimates, the available ground measurements, and radar observations from the CloudSat mission. In particular, the sensitivity of the WRF simulations to the use of different convective closures (explicit and Kain?Fritsch) and microphysical parameterizations (WRF single-moment 6-class microphysics scheme and Thompson) is analyzed. The impact of using different initial conditions, associated with a different initialization day, is also examined. The use of the new-generation Distributed Simulation and Stimulation System NASA Earth Observing System Simulators Suite radar simulator allows a more accurate and extensive representation of the mesoscale processes and of the interaction with the complex orography. The results reported here indicate that the quality of the large-scale initial conditions is a prominent factor affecting the possibility of retrieving a realistic representation of this event when using a nonhydrostatic regional model.