Impact of soil moisture initialization on a simulated flash flood

We investigate the sensitivity of a flash flood simulation to soil-moisture initialization using a the Regional Atmospheric Modeling System (RAMS). The case simulated is the Fort Collins flash flood of 28 July 1997. Four nested grids are utilized to simulate the synoptic-, meso-, and convective-scale motions. Two soil-moisture initialization procedures are employed. The first utilizes the 28 July 12Z ETA soil-moisture analysis. The second relies on an antecedent precipitation index (API) derived from a three month record of precipitation gauge measurements preceding the event. Significant differences were found in the initial soil moisture field with the API method yielding 18% average soil saturation on the third nested grid compared to 54% in the ETA analysis. The effects of soil moisture initialization on the evolution of low-level temperature, dewpoint temperature, convective instability, storm initiation and storm motions are noted and discussed. In addition, two simulations with homogenous soil moisture, equal to the third nested grid average for the API and ETA simulations, were performed to assess the impact of soil moisture variability on the precipitation distribution.