Spatial Patterns of Lightning at Different Spatial Scales in the Western United States during August of 1990 – A Case Study Using the Geographic Information Systems Technology

Abstract

Geographic information systems (GIS) have been widely used to study spatial variability in different atmospheric processes. In this study, we used a GIS approach to explore the potential to examine variation patterns of lightning strikes at different scales so that micro-, synoptic-, and planetary-scale processes can be linked in explaining and modeling the distribution patterns of lightning strikes. The data collected by the ground-based lightning detection system for an entire month in the western United States were used as an example. Lightning strike density surfaces were generated using different kernel bandwidths, or search radii. It has been recognized that density surfaces are useful in visual interpretation of spatial patterns at different scales, but there are insufficient data on how well such surfaces can be used in quantitative analysis of point distribution patterns. In our study, the resulting surfaces were compared quantitatively with gridded lightning strikes using meshes, or fishnets, of different cell sizes. The fishnet cell sizes ranged from 1 km for micro-scale processes to 50 km for synoptic- and planetary-scale processes. The results suggest that there is a threshold in the search radius or kernel bandwidth, above which a significant amount of errors would be introduced in quantitative analysis. It could be argued that it is possible to achieve a balance between the need for visual interpretation of distribution patterns and the need for quantitative analysis at different scales. We used the lightning data of August 1990 and digital elevation models of 1 km resolution to perform a case study on the relationship between lightning occurrence and topography. Our results indicate that at different spatial scales, the relationships between lightning density and topography may reflect different processes that influenced the spatial distribution pattern of the lightning occurrence.