Exploring first and second-order spatio-temporal structures of lightning strike impacts in the French Alps using subsampling

Jean-François Coeurjolly,J Blanchet,Alexis Pellerin

2024-01-11

Abstract:We model cloud-to-ground lightning strike impacts in the French Alps over the period 2011-2021 (approximately 1.4 million of events) using spatio-temporal point processes. We investigate first and higher-order structure for this point pattern and address the questions of homogeneity of the intensity function, first-order separability and dependence between events. The tuning of nonparametric methods and the different tests we consider in this study make the computational cost very expensive. We therefore suggest different subsampling strategies to achieve these tasks.

Statistics Theory

What problem does this paper attempt to address?

The paper attempts to address the following issues:

1. **Spatio-temporal distribution characteristics of lightning strike events**: The authors aim to model cloud-to-ground lightning strike events in the French Alps from 2011 to 2021 using spatio-temporal point processes. There are approximately 1.4 million such events. The main objective of the study is to analyze the distribution characteristics of these events in time and space.

2. **Homogeneity and first-order separability of the intensity function**: The authors explore the homogeneity of the event intensity function, i.e., whether the events are uniformly distributed in time and space. Additionally, they investigate the first-order separability of the intensity function, i.e., whether the intensity function can be decomposed into the product of spatial and temporal parts.

3. **Dependence relationships between events**: The authors also examine the dependence relationships between events, i.e., whether there is some correlation or interaction between different events. This can be measured using higher-order statistics such as Ripley's K function or J function.

4. **Optimization of computational costs**: Due to the large dataset, the computational costs of tuning non-parametric methods and various tests are very high. Therefore, the authors propose different subsampling strategies to reduce computational costs, enabling effective analysis of the aforementioned issues.

Through the exploration of these issues, the authors hope to gain a deeper understanding of the spatio-temporal distribution characteristics of lightning strike events in the French Alps and the underlying physical mechanisms from a statistical perspective.