Predictive Proxies of Present and Future Lightning in a Superparameterized Model
Five lightning proxies are tested in a climate model capable of directly simulating clouds. The importance of cloud-scale physics with regards to lightning and its future projections is shown; we compare two methods of representing cloud microphysics. The more complex approach leads to a better match with present-day observations with all proxies, as well as greater fidelity in reproducing the land-ocean contrast. It also predicts global increases in future lightning while the simpler approach predicts decreases.
Previous studies have disagreed on future projections of lightning flash rates. They disagree when comparing predictors using large-scale variables, such as convective available potential energy and precipitation, and those using small-scale variables, such as cloud ice contents. This study shows that there can in fact be agreement when a more complex representation of cloud microphysics is used in climate models. Specifically, there is agreement in decreases over tropical land and increases over oceans and the midlatitudes.
Previous projections of lightning flash rates have been carried out either in coarse-resolution climate models or cloud-resolving models over limited areas. In addition, various predictors have been proposed, with some forecasting increases in lightning with global warming and others decreases. We use a superparameterized model, a compromise that allows both global coverage and higher resolutions, to evaluate several lightning proxies, including two based on the product of large-scale variables (convective available potential energy [CAPE] and precipitation [P]) and three based on cloud-scale ice contents. We also test a 1-moment and a 2-moment microphysics scheme. CAPE x P shows global increases with a 4 K sea-surface temperature warming, as do the ice-based proxies simulated with the 2-moment parameterization, while those simulated with the 1-moment scheme show decreases.