Development and research of the GSI-based, Cycled, Dual Resolution Hybrid Ensemble-Variational Data Assimilation System for HWRF

A hybrid EnKF–variational data assimilation system was previously developed for HWRF based on the operational GSI. Experiments with the assimilation of TDR data have been conducted with a detailed study of hurricane Sandy (2012) and cases during 2012-2013 hurricane seasons. This hybrid system was able to correct both the wind and mass fields in a dynamically and thermodynamically coherent fashion with properly ingesting the radial velocity data from TDR onboard NOAA P-3 aircrafts. The hybrid system using self-consistent HWRF EnKF ensemble was found to improve both the track and intensity forecasts relative to GSI-3DVar and the hybrid ingesting GFS ensemble. More details can be found from Lu et al. (2015).

In this past study, the system was focused on assimilation during TDR periods and it was a single resolution system which had a more frequent assimilation (hourly or leg based) than operational HWRF (6 hourly). The HWRF hybrid data assimilation system was therefore further developed to include storm following, continuous cycling capability using a new directed moving nest strategy, dual resolution data assimilation capability, and vortex relocation capability. The primary goal of this study is to further improve the system to address the following questions: a) what is the best configuration for high resolution TC data assimilation; b) what is the impact of using a high resolution control analysis through dual resolution hybrid DA capability; c) what is the impact of vortex initialization and relocation processes before data assimilation; d) what is the best way to utilize the observations in hurricane DA, especially vortex scale observations.

Experiments were conducted with 32 DA cycles for hurricane Edouard (2014) and all operational observations including conventional in-situ data in prepbufr, satellite wind, tcvital, satellite radiances and tail Doppler radar observations. It was found that: a) the newly extended system improved in intensity forecast compared to the operational HWRF forecast; b) the dual resolution hybrid improved upon the coarser, single resolution hybrid; c) vortex initialization in the control is necessary for TC forecasts when TDR data are not available, and relocation of ensemble background can further improve upon the forecasts; d) TDR data is helpful for the inner core structures of the storms if it is used appropriately. These findings together with results from other cases will be presented in the conference.