Improving Convective Precipitation Forecasting through Assimilation of Regional Lightning Measurements in a Mesoscale Model

Abstract

A technique developed for assimilating regional lightning measurements into a meteorological model is presented in this paper. The goal is to assess the effectiveness of cloud-to-ground (CG) lightning information for improving the convective precipitation forecasting. The main concept of the technique is that utilizing real-time location, timing, and flash-rate data retrieved from a long-range lightning detection network, a regional/mesoscale meteorological model is informed about the deep moist convection spatiotemporal development and intensity. This information is then used to nudge the model-generated humidity profiles to empirical profiles as a function of the observed lightning intensity. The empirical humidity profiles are assumed to be representative of convective regimes since they have been produced on the basis of atmospheric soundings obtained during thunderstorm days. Case studies from three thunderstorm developments in a warm-season environment over the Mediterranean are used to investigate the relationship between lightning density and different empirical humidity profiles, and consequently demonstrate the impact of the technique on model precipitation forecasts. Results show that assimilation of lightning data can significantly improve the model?s prediction accuracy of convective precipitation in the assimilation period, while maintaining the improvement in short-range (up to 12 h) forecasts compared to the control case. The approach is general enough to be applied to any mesoscale model, but with an expectable varying degree of success. Its advantage when applied in an operational setting is that real-time lightning data responding promptly to the occurrence of convection would continuously get assimilated to update the moist state of the atmosphere in the model.