One goal of this project is to determine whether a mesoscale model run at very high resolution can realistically simulate the evolution and magnitude of this flooding event. The Weather Research and Forecasting (WRF v3.5.1) model was used for these simulations using various initial conditions, physics, and domain configurations to test the sensitivities for the 18-h forecast starting at 0000 UTC 13 August 2014. The WRF was nested down to 1-km grid spacing in order to simulate the observed precipitation rates. Those WRF simulations that used the 0.5-degree Global Forecast System (GFS) analysis and its short-term forecast grids provided the most realistic WRF predictions for this event (as compared to using the NAM analysis/forecast). Numerous physics were tested to test the sensitivity, but this analysis will focus on a run using the WSM6 microphysics, RRTM longwave radiation, RRTMG shortwave radiation, ACM2 PBL, and Grell-3 cumulus schemes (for grid spacing > 5 km).
There was very large sensitivity to the WRF initial conditions, boundary layer and microphysics, domain size, and even grid nest interaction for this case, which will be demonstrated. Many members underpredicted the precipitation by a factor of two even down to 1-km grid spacing, but a realistic member at 1-km grid spacing was obtained in which over 285 mm of precipitation fell in narrow band at about the same locations as observed. Another goal of this presentation will highlight the impact of resolution on the evolution on the precipitation rates and circulations. Lastly, with a realistic high-resolution 1-km run, we will also highlight some of the low-level forcing and vertical circulations associated with the boundary that produced this band and mechanisms for its longevity.