A Gaussian Process Regression Method to Nowcast Cloud-to-Ground Lightning From Remote Sensing and Numerical Weather Modeling Data

Abstract

The climate change of the last decades is causing an increase in the frequency of extreme event occurrences. Numerous studies confirm the link between extreme meteorological events and lightning activity. The possibility of having short-term predictions of the intensity of lightning phenomena would allow the near-real-time monitoring of the evolution of these events. Such knowledge would help in defining operational strategies to mitigate effects on the population and the infrastructure. In this context, this article proposes a multidisciplinary approach aiming at developing a regression algorithm to nowcast the density of cloud-to-ground lightning strokes 1 h in advance. The algorithm is developed on the basis of a joint combination of remote sensing data, numerical weather prediction model outcomes, and lightning observations. The possible dependence between the meteorological data and the density of lightning is estimated using Gaussian process regression models. Separate models are created for positive and negative strokes and for strokes occurrences over the land or sea. The experimental results indicate that the model quite accurately estimates low numbers of strokes, whereas larger numbers of strokes are estimated with higher errors. Nevertheless, their presence is correctly detected. This suggests the potential of the method as a processing tool to support the management of weather-related hazards.

Keywords

• Lightning
• nowcasting
• Gaussian process regression
• numerical weather prediction
• satellite infrared imagery
• satellite precipitation radar