Predictive Proxies of Present and Future Lightning in a Superparameterized Model

Abstract

A superparameterized climate model is used to assess the global performance of several previously proposed proxies for lightning. In particular, predictors incorporating hydrometeor (ice, graupel) profiles and convective vertical velocities are compared to observations, then used to estimate changes in flash rates with global warming. The choice of microphysics parameterization is also investigated, with all predictors showing higher correlations with Lightning Imaging Sensor/Optical Transient Detector observations when using a 2‐moment scheme compared to a 1‐moment representation. All proxies generally agree in their response to warming over tropical land, with notable decreases in Africa, the Middle East, and northern South America, but disagree over oceans and the midlatitudes. The product of convective available potential energy and precipitation predicts increases over these latter areas, as do the 2‐moment ice‐based proxies, while those of the 1‐moment model tend to show decreases, highlighting the importance of cloud microphysics when using climate models to simulate lightning.

Lightning Over Central Canada: Skill Assessment for Various Land‐Atmosphere Model Configurations and Lightning Indices Over a Boreal Study Area

“…However, the southwest‐northeast gradient seen for the CLDN data is also seen for the CAPE × P index for both G4ICE configurations, albeit less pronounced. This also matches the general pattern seen from coarse‐scale global lightning predictions (Charn & Parishani, 2021). However, for the global predictions, much coarser resolutions are used, making the results of those studies not directly comparable with the fine‐scale, local patterns found here.…”

“…These finer resolution models allow for deep convection to be resolved explicitly, resulting in an improved representation of most convection related processes (Brisson et al., 2016; Lucas‐Picher et al., 2021; Prein et al., 2015). At the fine resolution, convection parameterization schemes become obsolete and other processes contributing to deep convection, such as microphysical (MP) processes, and their formulations become more important (Adams‐Selin et al., 2013; Charn & Parishani, 2021). The finer spatial resolution of convection‐permitting models also allows to represent more accurately the effect of land surface heterogeneities in the modeled land‐atmosphere interactions (Vanden Broucke & Van Lipzig, 2017).…”

“…There are numerous studies that compare the performance of different lightning indices (e.g., Allen & Pickering, 2002; Brisson et al., 2021; Charn & Parishani, 2021; Finney et al., 2014; Romps, 2019) but none of these make comparisons across different resolutions. Most studies focus on global lightning patterns, comparing the output of global simulations with lightning observations.…”

Amplification of Annual and Diurnal Cycles of Alpine Lightning

“…Classically, cloud top height (Price and Rind, 1992), the product of CAPE and precipitation (Romps et al, 2018), the square root of CAPE multiplied by deep layer shear (Taszarek et al, 2019), CAPE exceeding a threshold conditioned on the occurrence of convective precipitation (Taszarek et al, 2021), or vertical iceflux (Finney et al, 2014) have been used. Moreover, the content of such proxies depends on the micro-physical parameterizations used in the numerical models (Charn and Parishani, 2021).…”

Projection of Global Future Lightning Occurrence using only Large-Scale Environmental Variables in CAM5

“…However, these frameworks still require separate land and ocean equations and often need to be scaled to the current global mean lightning to provide a realistic prediction. In addition, Charn and Parishani (2021) found that the ice-based lightning parameterizations may be sensitive to the microphysics scheme used, not necessarily to the variables used to predict lightning, which adds motivation to avoid highly uncertain storm-scale variables as inputs for lightning parameterizations in GCMs.…”