Examining the Sensitivity of Simulated Midlatitude and Tropical Mesoscale Convective Systems to Horizontal Grid Spacing

Mesoscale convective systems (MCSs) are the predominant source of rainfall in the tropics and parts of the midlatitudes, produce extreme precipitation and severe weather worldwide, and contribute significantly to the global circulation. Continued advances in computing technologies have made global numerical weather prediction and regional climate simulations feasible at convection-permitting scales (Δx ≤ 4 km) and enabled regional weather prediction at large-eddy simulation scales. However, global climate models remain much coarser and thus still rely heavily on the parameterization of convective processes, with active efforts underway to refine their grid spacings to become convection-permitting. Although convection-permitting models have demonstrated drastic improvements in the representation of MCSs compared to those reliant on convective parameterizations, these grid spacings lie within the convective “gray zone”, wherein turbulent motions are incompletely resolved.

This work builds upon previous studies that have sought to determine the optimal model grid spacing for MCS simulations by further evaluating the simulated properties of convective updrafts and downdrafts in both midlatitude and tropical MCSs as a function of horizontal grid spacing. In this study, we leverage MCS observations collected at the Department of Energy’s Atmospheric Radiation Measurement (ARM) sites in the U.S. Southern Great Plains (SGP; Lamont, Oklahoma) and during the GoAmazon2014/5 field campaign (MAO; Manaus, Brazil) to assess the performance of the Weather Research and Forecasting (WRF) model in simulating 10 midlatitude and 10 tropical MCSs that passed over each site. For each MCS case, we conduct an ensemble of convection-permitting simulations at grid spacings ranging from 4 km to 250 m. Additionally, we perform simulations at 125-m grid spacing for two cases occurring within each regime. In this presentation, we will describe the sensitivity of various convective draft properties (e.g., intensity, vertical structure, and probability of occurrence) to horizontal grid spacing and how these sensitivities compare between midlatitude and tropical MCSs. Further, we will discuss how these simulated draft properties relate to the representation of physical processes at different grid spacings.