Impact of microphysics parameterizations on simulation of 13 March 2003 bow echo

The Advanced Research Weather Research and Forecasting (WRF-ARW) model is used to simulate an isolated bow echo case over Oklahoma on 13 March 2003. Simulations utilizing 3-km grid spacing and a domain size of 2250 km by 2100 km are performed with different microphysics parameterizations, specifically the WRF Single Moment 5- and 6-class schemes (WSM5, WSM6), and the WRF Double Moment 5- and 6-class schemes (WDM5, WDM6). Each simulation produces a bowed line of convection, with initiation timing similar to the real-data case. However, differences in the development and intensity of the cold pool, rear inflow jet, and stratiform rain region, as well as system propagation speed, quantity of precipitation, and surface wind production are all caused by microphysical changes. These factors are more sensitive to the addition of graupel, between the 5- and 6-class schemes, than the change from single- to double-moment schemes, although both have an effect. Specifically, both the WSM5 and WDM5 schemes generate widespread concentrations of snow crystals in the upper-levels, resulting in unrealistically large stratiform regions both in advance of and behind the less intense, and less bowed, convective line. The 5-class schemes also produce much weaker cold pools, correspondingly slower propagation speeds, and weaker surface wind speeds than the 6-class schemes; the 6-class schemes are more realistic. Effects of adjusting parameters internal to the parameterizations, such as the intercept parameters of snow and graupel for the single-moment schemes, and the shapes of the rain and cloud water distributions of the double-moment schemes, are also examined. By observing these effects, it is expected further information about the connection between cold pool strength and new bowing development will be revealed.