Sensitivity analysis of lightning stroke-to-flash grouping criteria

Abstract

Lightning Location Systems provide detailed data on the lightning occurring within the area of the system's coverage. This kind of remote sensing equipment uses geolocation techniques to retrieve lightning cloud-to-ground return strokes. Prior to getting information regarding the number of lightning flashes, cloud-to-ground strokes need to be grouped into lightning flashes. The stroke-to-flash conversion is not straight forward and therefore, a stroke-to-flash grouping criteria (flash algorithm) is needed to obtain subsequent figures like the Ground Flash Density (NG). It is now more than twenty years since the most used stroke-to-flash grouping criteria was established: 1 s, 10 km radius and 500 ms inter-stroke interval.In the course of these years, new observational systems like the Lightning Mapping Array have been deployed to precisely depict the lightning flash extension and duration. The present work aims to introduce a new approach to the flash algorithm, taking advantage of the Lightning Mapping Array capabilities. Two methods based on this system have been used in this study as an alternative way for grouping cloud-to-ground strokes into flashes. These methods are compared with the classical flash algorithm. Results suggest shortening the total flash duration and enlarging the clustering radius. Besides, this work also analyses the sensitivity of the flash algorithm to the changes on the grouping criteria. Changes in these parameters will have a bearing on the retrieved annual number of lightning flashes, and thus will also influence derived lightning figures like the NG.