Warm-Season Precipitation Organization in the Southeast United States: Current and Future Climate

The representation of precipitation systems is one of the major sources of uncertainty in model simulations of future climate, particularly the prediction of regional precipitation change. Mechanisms that control precipitation variability are in part linked to precipitation system organization. One approach to help reduce the uncertainty in regional climate prediction of precipitation is to link the more predictable aspects of climate, such as the diurnal variation of the thermodynamic structure of the atmosphere, with the statistical properties of precipitation organization. We focus on a simple size-based aspect of precipitation organization: whether the rain is organized as isolated thunderstorms that cover part of a city or as mesoscale rain systems that cover half a state.

Toward that end, this paper presents a seven-day case study (modeling and observations) of summer season diurnal convection in the southeastern United States for both current and future climate conditions, focusing on precipitation organization. For the future conditions, the pseudo-warming downscaling method is used, where initial conditions of temperature in the WRF model are adjusted by future (2090) climate temperature anomalies taken from ensemble runs of CMIP-5 GCM simulations assuming RCP 4.5 and RCP 8.5 carbon emission scenarios. An algorithm to identify isolated precipitation features (IPF) and mesoscale precipitation features (MPF) is applied to observed precipitation from the National Mosaic and Multi-sensor Quantitative Precipitation Estimation (NMQ) radar reflectivity and precipitation data sets,

In the current climate, WRF domain-mean precipitation totals compared well (within 6%) to observations. WRF captured quite well the observed pattern, phase and amplitude of afternoon diurnal maximum over land and nocturnal maximum over ocean, both associated primarily with IPF features. However, the WRF model produced almost 50% less MPF precipitation than observed, due mainly to the model’s failure to produced any MPF precipitation over the Gulf Stream off the Carolina coast. This problem, under investigation, may be related to the options chosen in the explicit convective scheme.

For the future climate runs, compared to modeled current climate, WRF RCP 4.5 produced about 6% more precipitation while WRF RCP 8.5 contained 30% more precipitation. For RCP 4.5, almost all of the modest increase was from IPF over land, while for RCP 8.5 the increase was large in both IPF and MPF, suggesting a threshold at or above RCP 4.5 for a large MPF precipitation increase. Modeled CAPE increased significantly (40%) over both land and ocean, with CIN only increasing over land (by 30%) and no change over ocean. This suggests that greater instability over ocean drove more IPF rain, while increase in CIN over land kept IPF land precipitation from increasing much in the future runs.

This case study is a feasibility study for a new project to perform multi-year WRF current climate and pseudowarming runs, to build robust statistics on future changes in precipitation organization for the SE US.