A Radar Investigation of Storm Morphology and Precipitation Processes during Discrepancies between GOES-16 GLM and LMA Observed Lightning Flash Rates and Jumps

Preliminary analyses have identified some notable inconsistencies between optically-based GOES-16 Geostationary Lightning Mapper (GLM) and VHF-, ground-based Lightning Mapping Array (LMA) flash rates and trends, including from the two-sigma lightning jump algorithm (LJA), during several supercells over Northern Alabama, Colorado and West Texas. Given differences in spatial resolution, expected detection efficiency and measurement approach between the GLM and the LMA, it is not surprising to find some variance in lightning flash rates and trends. However, in these severe storms, the GLM flash rates are 2 to 3 times less than the LMA during prolonged periods. Furthermore, the timing and overall number of lightning jumps are often not in agreement between the GLM and LMA. Different hypotheses have been put forward to explain the observed inconsistencies in flash rates and trends. The storm periods in question are characterized by high LMA flash rates (> 30 min^-1) suggesting that differences in spatial resolution, detection efficiency and/or measurement approach could be important. The temporal correlation between observed storm flash rate and radar reflectivity-based proxy of storm intensity, such as Vertically Integrated Liquid (VIL) or Maximum Expected Size of Hail (MESH), is significantly higher for LMA than for GLM. The largest discrepancies in flash rates and trends between GLM and LMA are during periods with larger maximum VIL (> 20 kg m^-2) and maximum MESH (> 10 mm). Consistent with the radar metrics, large hail has also been reported on the ground during these times. One hypothesis that this study will pursue further is that GLM flash rates may be suppressed, likely in part due to decreased optical detection efficiency, during periods of high optical thickness associated with significant quantities of dual-polarization radar observed hail and other riming precipitation ice particles in and above the heights of LMA observed lightning initiation and propagation. The presence of hail is a rough proxy for high cloud water droplet concentration, which could increase the extinction of optical radiance from the GLM field of view and cause a reduction in optical flash detection efficiency. If occurring, the increased optical extinction and reduction in GLM flash detection efficiency could vary over time and space, thus also adversely affecting GLM lightning trends and the GLM LJA. To investigate, dual-polarization radar will be used to characterize storm morphology and hydrometeor types in supercells over Northern Alabama, Colorado and West Texas. Radar inferred precipitation properties and processes will be used to confirm the lightning jump conceptual model and to investigate inconsistencies between GLM and LMA flash rates and lightning jumps in these storms. LMA flash heights and extents will be analyzed relative to storm morphology and hydrometeor types for potential relevance. Convection embedded in a mesoscale convective system over central Oklahoma during which there was better agreement between GLM and LMA flash rates will be studied for comparison.