The southeastern US receives a substantial volume of rainfall during the warm season, usually in the form of short-lived convective systems with limited spatial extent. These systems are driven largely by thermodynamic processes within the lower atmosphere, with moderate to weak dynamic forcings in the mid and upper levels. Due to the threat of flash flooding associated with the short but intense periods of rainfall, it is critical to quantify the probable frequency and distribution of rainfall for forecasting and historical analysis applications. Regional numerical weather prediction (NWP) models, such as the popular and widely used Weather Research and Forecasting (WRF) model, are often used to aid in quantitative precipitation forecasting (QPF) and analysis of rainfall events; however, NWP models are known to have difficulty in estimating precipitation extent and depth during thermodynamically driven events. This difficulty arises due to the inability to accurately portray small-scale convective and microphysical properties associated with warm-season rainfall generation. WRF utilizes a number of parameterization schemes for convection, microphysics, and planetary boundary layer processes, which along with choice of model core and initial conditions, can greatly influence simulated rainfall patterns. However, it is critical that the correct combination of parameterizations be applied to maximize the accuracy of the precipitation forecasts. To address this concern, this study will utilize a high-resolution (4-km) WRF domain over the southeastern US to simulate a selection of warm-season convective events, employing a parameter ensemble approach. The convective, microphysical, and planetary boundary layer schemes will be varied to test the sensitivity of the high-resolution WRF runs to parameterizations related to rainfall generation. The relative skill score and accuracy of the individual model runs will be quantified using the 4-km resolution NEXRAD multi-sensor precipitation dataset, utilizing a bivariate mixed log-normal distribution for comparison of the simulated and estimated gridded precipitation time series. Additionally, the difference in total precipitation volume and rainfall distribution will be assessed between the data sets using central tendency and cluster analysis, respectively, to ascertain which parameterizations are most capable of producing viable rainfall forecasts during thermodynamically driven warm-season events in the southeastern US. Results of the project will be used in developing operational regional WRF simulations for QPF applications, as well as the development of research WRF domains for analysis of historical extreme precipitation events.