Convection-Permitting Analysis and Prediction of Hurricane Harvey (2017) through Ensemble Assimilation of All-Sky GOES-R Radiance

Hurricane Harvey caused catastrophic damages in the Gulf region in late August 2017, which was recorded as the first landfall major hurricane observed throughout its entire lifecycle by the newly launched next-generation geostationary satellite GOES-16 (first of the GOES-R series). Harvey is characterized by its reemergence and following rapid intensification (RI) process in the Gulf of Mexico, which was very challenging to predict by most of the operational models. Given the advantage of geostationary satellites that can seamlessly provide high spatiotemporal-resolution observations in particular over the inner-core vortex region, we explore the impacts of assimilating all-sky infrared satellite radiances from GOES-16, for convection-permitting initialization, analysis and prediction of tropical cyclones. We assimilate hourly all-sky radiances from one of the water-vapor sensitive channels of GOES-16 with the ensemble Kalman filter (EnKF) data assimilation system developed at Penn State University (PSU) and built around the Advanced Weather Research and Forecasting model (WRF-ARW) and the Community Radiative Transfer Model (CRTM).

Assimilation of all-sky radiances from GOES-16 remarkably improved the representation of the vortex structure from convective to mesoscales. Continuous cycling data assimilation enables the analysis to capture the developing individual cells, the convections with the rapidly intensifying vortex and the secondary eyewall formation. Deterministic forecasts initialized from the EnKF analyses (referred to as APSU) become able to simulate the rapid intensification of Harvey that reached the category-4 intensity more than 24 hours prior to the RI, which was not predicted by the operational forecast guidance models initialized at the corresponding timings (Figure). Also performed are the ensemble probabilistic forecasts at several earlier stages of Harvey to investigate the sensitivity of inner-core and environmental structures on RI, key dynamics and its predictability. This study highlights the potentials of next-generation geostationary satellites to analyze both the dynamic and thermodynamic conditions of tropical cyclone inner-core structures, which would be essential to understand and forecast the process leading to the development of tropical cyclones.