Synthetic Lightning Generation Employing Autoregressive-Moving-Average (ARMA) Models

Lightning is a meteorological phenomenon that occurs frequently in numerous locations every single day on Earth. Comprehending the way it materializes spatially and chronologically is imperative to develop a realistic environmental scene. Live lightning data can be fed into a scene, but that process is costly. Therefore, this work explores if lightning data can be generated synthetically using vector autoregressive-moving-average (VARMA) models. Geostationary Lightning Mapper (GLM) data is used as the basis for the study. Lightning climatology is examined and compared to previous research to gain insight into the targeted areas. Individual lightning flashes are analyzed to inspect how well the process works on a smaller scale. Then, entire regions are evaluated to simulate lightning creation in a larger setting. The results of each step suggest that VARMA models are sufficient at lightning generation up to a certain degree. However, the techniques used in this study have the potential to beimproved and allow the models to mirror expansive scenes containing many lightning strikes.