Ensemble-Based Analysis of the 14 June 2010 Oklahoma City Flood

The extreme rainfall observed in Oklahoma City on the morning of 14 June 2010 was skillfully predicted by a real-time storm-scale ensemble forecast produced by the Center for Analysis and Prediction of Storms. This study ranks the members of that ensemble based on quantitative precipitation forecast skill for the event. Forecasts of higher-skill and lower-skill members are then compared to determine which factors were most crucial for an accurate precipitation forecast, in terms of both model settings and predicted meteorological features.

This study demonstrates the importance of initial radar data assimilation and proper treatment of microphysical processes in predicting cold pool strength and subsequent outflow boundary propagation in cases of this sort. It also highlights more transient and/or fine-scale features that contributed to greater combined intensity and duration in the forecast precipitation over the Oklahoma City area. The nature of these features suggests that attempting to forecast such events deterministically is unreliable.