Ensemble Assimilation of Simulated Total Lightning Flash Rates

A set of observing systems simulation experiments (OSSEs) demonstrates the potential benefit from ensemble Kalman filter (EnKF) assimilation of total lightning flash rate data. The planned Geostationary Operational Environmental Satellite R-series (GOES-R) platforms are expected to include the Geostationary Lightning Mapper (GLM) to detect total lightning flash rate with high detection efficiency at a resolution of 8-12 km. Here, a truth simulation is conducted using multi-moment bulk microphysics, electrification mechanisms, and a branched lightning parameterization to produce 2-minute averaged synthetic pseudo-GLM observations at 1 km and 8 km resolutions.

The OSSEs use either perfect (two-moment bulk with graupel and hail) or imperfect (single-moment, graupel only) microphysics. One ensemble included the same electrification physics as the truth simulation to generate lightning flash rates, and the other ensembles used linear relationships between flash rate and graupel echo volume as the observation operator. Comparisons with pure forecast ensembles show that the assimilation of flash rates at 8 km horizontal resolution can effectively modulate the convection simulated at 1 km horizontal resolution by sharpening the location of reflectivity echoes and the spatial location probability of convective updrafts. Tests with zero flash rates show that the EnKF is able to limit spurious deep convection, as well. Tests using pseudo observations at 1 km further sharpen the analyses of storm location.