Evaluating the Relationship between Lightning and Large-Scale Environmental Parameters with Observational Data, Statistical Models, and Lightning Parameterizations in GCMs

Many links have been observed between the large-scale environment and precipitation, yet the relationship between large-scale environmental parameters and lightning remains relatively understudied with the exception of Convective Available Potential Energy (CAPE). CAPE has been long-thought to be a major contributing factor to lightning because of its proxy for updraft strength. However, CAPE values are similar for convection over both land and ocean (regions that show very different lightning distributions). Normalized CAPE (nCAPE) can potentially help discriminate between land and ocean thermodynamic environments and thus lightning production as it takes the “shape-of-CAPE” into account, where “fat CAPE” is typically observed over land and “skinny CAPE” is typically observed over ocean. Other variables, such as column saturation fraction (r), low-level wind shear (LS), deep wind shear (DS), and large-scale vertical motion (omega), have been shown to be correlated with precipitation and could also help distinguish the land/ocean lightning contrast as well as aid in the prediction of lightning in general circulation models (GCMs).

Using the Tropical Rainfall Measuring Mission (TRMM) Lightning Imaging Sensor (LIS) and MERRA-2 reanalysis data, this study investigates six large-scale environmental parameters and their correlation with lightning over the tropics and subtropics: CAPE, nCAPE, omega at 700 hPa, r, LS and DS. The lightning and reanalysis data is gridded at 0.5° spatial resolution and 3-hourly temporal resolution. We investigate the relationship between the six parameters and lightning using multiple regression and point-process statistical models. Preliminary results show that nCAPE and r have the highest correlation with lightning. Omega, LS, DS and CAPE are of secondary importance when compared to nCAPE and r but help in distinguishing land versus ocean lightning occurrence and intensity. We have also developed a lightning parameterization with our statistical models and plan to evaluate its performance in CAM5 to assess the ability of these environmental parameters to predict lightning in GCMs.