Evaluation of Mesoscale Convective Systems in Climate Simulations: Methodological Development and Results from MPAS-CAM over the United States

Abstract

AbstractA process-oriented approach is developed to evaluate warm-season mesoscale convective system (MCS) precipitation and their favorable large-scale meteorological patterns (FLSMPs) over the United States. This approach features a novel observation-driven MCS-tracking algorithm using infrared brightness temperature and precipitation features at 12-, 25-, and 50-km resolution and metrics to evaluate the model large-scale environment favorable for MCS initiation. The tracking algorithm successfully reproduces the observed MCS statistics from a reference 4-km radar MCS database. To demonstrate the utility of the new methodologies in evaluating MCS in climate simulations with mesoscale resolution, the process-oriented approach is applied to two climate simulations produced by the Variable-Resolution Model for Prediction Across Scales coupled to the Community Atmosphere Model physics, with refined horizontal grid spacing at 50 and 25 km over North America. With the tracking algorithm applied to simulations and observations at equivalent resolutions, the simulated number of MCS and associated precipitation amount, frequency, and intensity are found to be consistently underestimated in the central United States, particularly from May to August. The simulated MCS precipitation shows little diurnal variation and lasts too long, while the MCS precipitation area is too large and its intensity is too weak. The model is able to simulate four types of observed FLSMP associated with frontal systems and low-level jets (LLJ) in spring, but the frequencies are underestimated because of low-level dry bias and weaker LLJ. Precipitation simulated under different FLSMPs peak during the daytime, in contrast to the observed nocturnal peak. Implications of these findings for future model development and diagnostics are discussed.