Processing and Quality-Controlled Visualization of GOES-16 GLM Data

The Geostationary Lightning Mapper instrument on GOES-16 (and future satellites in the GOES-R series) generates lightning measurements in the form of hierarchically clustered, time-tagged, and geolocated detection events. Such hierarchical measurements are unusual in meteorology, and existing tools do not facilitate analysis or visualization of them. Experience indicates that most users treat GLM data as point data, and do not make use of the spatial footprint information implied by the hierarchy.

We describe a Python package, to be open-sourced for the meteorological community, which allows both experts and non-specialists to appropriately use the GLM data. The package is built on xarray, which was also patched to automate interpretation of the unique way in which GOES-16 datasets store unsigned integer data. The package reconstructs and makes readily available metadata implied by the hierarchical data format, as is necessary for quality control. The same reconstruction approach also facilitiates visualizations which capture the space-covering nature of the measurement. The end result is imagery-like loops of aggregated the GLM data that give a count of how many lightning flashes illuminated a particular spot at cloud top, as well as time series statistics and other derived properties. Such imagery allows for ready identification of meteorological signals such the contrast between rapidly flashing, spatially compact thunderstorm cores and infrequent, very large flashes in extensive anvils and stratiform rain regions adjacent to the parent storms.