Regional Applications of the GSI-Hybrid Data Assimilation for High-Resolution Tropical Storm Forecasts

In collaboration with research and operational centers, the Developmental Testbed Center (DTC) works toward the improvement of Data Assimilation (DA) and initialization of numerical models for tropical cyclone forecasting. Recent work has been performed in the framework of the 2014 operational Hurricane Weather Research and Forecast (HWRF) model, which applied some critical updates to various aspects of the system from the 2013 HWRF system, including, but not limited to, increased model levels from 43 to 61 levels with the model top raised from 50mb to 2mb, the upgraded vortex relocation scheme, and some important changes to the hybrid variational-ensemble configuration of the Gridpoint Statistical Interpolation (GSI) system. The data assimilation is now performed for the two new intermediate domains (ghost_d02 and ghost_d03, with 9km and 3km horizontal resolutions respectively), with more observational data (satellite radiance, satellite derived winds, GPS Radio Occultation (RO), etc) in addition to the conventional observation and Tail Doppler Radar (TDR) data.

Multiple cases of tropical storms are run to investigate various aspects of the initialization of the HWRF system, with a focus on the GSI-Hybrid DA system, as part of the effort to advance the GSI-Hybrid DA system and initialization in HWRF. Diagnostics have been performed to study the impact of numerous configurations on hurricane forecasts, including the benefit of applying high resolution regional HWRF ensemble for the vortex scale DA, the impact of DA and vortex initialization, and investigation of the spin-down of the intensity forecasts.