How Might the May 2015 Flood in the U.S. Southern Great Plains Induced by Clustered MCSs Unfold in the Future?

The historic 22-26 May 2015 flood event in Texas and Oklahoma was caused by anomalous clustered mesoscale convective systems (MCSs) that produced record-breaking rainfall and $3 billion of damage in the region. A month-long regional convection-permitting simulation is conducted to reconstruct multiple clustered MCSs that lead to this flood event. We further use the pseudo global warming approach to examine how a similar event may unfold in a warmer climate and the driving physical factors for the changes. Tracking of MCSs in observations and simulations shows that the historical simulation reproduces the salient characteristics of the observed MCSs. In a warmer climate under a high-emission (SSP5-8.5) scenario, the Southern Great Plains is projected to experience a near surface warming of 4-6 K, accompanied by enhanced moisture transport by the strengthened Great Plains low-level jet. A warmer and moister lower troposphere leads to 36-59% larger convective available potential energy, supporting wider and more intense convective updrafts and rainfall production. Consistently, MCSs have wider convective areas and stronger rainfall intensities, producing 50% larger rain volumes during the mature stage. Extreme (99.5%) MCS rainfall frequency and amount will increase by threefold (Fig. 1). However, MCS stratiform rain area decreases as a result of elevated stratiform cloud bases that lead to stronger sublimation and evaporation of precipitation in response to warming, resulting in reduced weak-to-moderate surface precipitation. Results suggest that global warming greatly increases precipitation intensity of clustered MCS events under strong synoptic influence, with much higher potential to produce serious floods without additional climate adaptation.

Figure 1. (a) Frequency distribution of MCS grid-point hourly rain rates, and (b) normalized cumulative distribution of rainfall amount by hourly rain rates. The region of the data included is show in the inset.