Adaptive Observation Error Inflation (AOEI) and Adaptive Background Error Inflation (ABEI) for Convection-Permitting Ensemble Assimilation of All-Sky GOES-16 Radiances

The effective usage of high-spatiotemporal all-sky satellite observations, such as from the new generation geostationary satellites GOES-R (GOES-16), is one of the key promises to improve future severe weather and tropical cyclone predictions. Given strong nonlinearities and imperfectness in the forecast model and all-sky radiance observation operators for severe convective weather, there are strong sampling and representativeness errors in both the default observational error statistics and the ensemble-estimated flow-dependent background error. We have recently developed two empirical adaptive inflation methodologies, one called Adaptive Observation Error Inflation (AOEI) (Minamide and Zhang 2017) and Adaptive Background Error Inflation (ABEI) (Minamide and Zhang 2018), both of which are designed to mitigate the representativeness errors in assimilating all-sky radiances with highly nonlinear observation operators. Here we explore the effectiveness of AOEI and ABEI for the impacts of assimilating all-sky infrared satellite radiances from GOES-16 for convection-permitting initialization and prediction of Hurricane Harvey (2017) and a June-2017 Tornadic severe weather event over the US Midwest. It is found that the assimilation of the all-sky infrared radiance with these adaptive inflation methods can accurately constrain the dynamic and thermodynamic state variables; forecasts initialized the EnKF analysis and perturbations can lead to significant improvement in both deterministic and probabilistic forecasts of these severe weather events.