A National Weather Service Central Region science goal is to improve Quantitative Precipitation Forecasts (QPF) for elevated convection. Inaccurate ETA QPF for a case of elevated convection over the Midwest during 20 August 2002 0000 UTC to 1200 UTC was observed. We will analyze sensitivity of QPF to convective parameterization schemes and horizontal grid spacing for this case.

Subjective analysis has suggests that initial ETA mesoscale features are accurate in respect to observed surface and upper air data. A warm front in southern Missouri at 0000 UTC was represented in the ETA model. RUC 0000 hour forecasts, used as a proxy for observed data to analyze the evolution of the ETA model, suggest that ETA correctly evolved mesoscale features. The warm front progression in the ETA model corresponded to the warm front in the RUC model through out the time in question. ETA wind profiles were compared to RUC 0000 hour profiles. The ETA simulation of the low level jet was realistic, so we ruled out errors in simulation of the low level jet as a cause of QPF error. Comparison of temperature with height indicated that the ETA model correctly represented temperature change, suggesting that the stability was represented.

At the conference, we will discuss results from sensitivity analysis.