Evaluation of Product Algorithm Strategies for the Future Lightning Imager on MTG

Abstract:

The future European geostationary Meteosat Third Generation (MTG) will carry a lightning observation instrument - the Lightning Imager (LI), which will observe lightning in thunderstorm for the whole visible Earth disk with high temporal and spatial resolution. Its data will be of great benefit for nowcasting of lightning thread and severe weather, as well as for applications in lightning physics, chemistry and climatology.

Primary data output of the instrument are the 'lightning events’ corresponding to triggered pixels of the detector matrix. These primary data are processed into a set of products: hierarchically organized point data - groups and flashes, and various products of data accumulated over time and space bins. The algorithm strategies for generation of these level-2 products are evaluated in this contribution.

The clustering of the lightning event data consists of two steps: i) pixel events are clustered in a group representing the optical pulse, and ii) the groups are clustered in a flash representing the lightning flash. The algorithm strategies for clustering are discussed and evaluated for the various algorithm parameters. The distance function and linkage criteria are formulated in time and space. For groups the criterion is immediate adjacency in time and space. For flashes the gap to the nearest neighbor in time and space have to be lower a threshold value. Examples and statistics of the application of the clustering algorithms are presented for LIS data and for proxy data.

The second group of products accumulate the lightning point data in time (30s) and space (2km) intervals. The planned products are the accumulated flash number, flash area and flash radiance. All these products are created on a per flash bases using the pixel set for each flash. They are calculated on a grid which correspond cloud imager (FCI) grid on the same platform. The accumulated products represent an snapshot picture of the current lightning activity, valuable in nowcasting, and can be further accumulated into various lightning statistics and climatologies. The contribution discusses the interpretation and application of these products on different time scales.