Improving High-Resolution Tropical Cyclone Prediction using a GSI-based, Cycled, Dual Resolution Hybrid Ensemble-Variational Data Assimilation System for HWRF: system description and experiment results

Tropical cyclones (TCs) are always on the list of the most costly natural disasters. After decades of effort, we have now reduced the storm track forecast errors by as much as half from that 20 years ago (Zhang and Weng, 2015). However, in comparison with the dramatic improvements in track forecast skill, the intensity forecast has advanced very little (Berg and Avila, 2011). This can be attributed to deficiencies in models, deficiencies in current operational data assimilation systems, and lack of effective utilization of observations.

In our recent study (Lu et al. 2015), a hybrid EnKF–variational data assimilation system was developed for HWRF based on the operational GSI. In this initial effort, the system was first tested with a single, fixed domain. Experiments have been conducted with a detailed study of hurricane Sandy (2012) and cases during 2012-2013 hurricane seasons with the assimilation of airborne radar observations. It was found that this hybrid system was able to correct both the wind and mass fields in a dynamically and thermodynamically coherent fashion while only ingesting the radial velocity data from the Tail Doppler Radar (TDR) onboard the NOAA P-3 aircraft. 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. The hybrid method also provided the largest positive impact of the TDR data.

In Lu et al. (2015), the system only focused on periods of TC sampled by TDRs and only explored the hybrid DA system in a partial cycling mode. The HWRF hybrid data assimilation system was therefore further extended to conduct DA and forecast cycles for the entire life of TC assimilating all operational observations in addition to TDR. The system was also further developed to conduct an end-to-end, continuous DA cycling using a new developed directed moving nest strategy. The primary goal of this study is to use the newly extended system to address the following scientific questions: a) what is the impact of dual resolution over single resolution hybrid DA? b) what is the impact of vortex initialization and relocation? c) what is the impact of 3DEnVar/4DEnVar/hourly DA for vortex scale airborne radar observation? and d) how and why can the newly developed hybrid system help alleviate the “spin-down” issue?

Experiments were conducted with end-to-end cycles for hurricane Edouard (2014) in which all operational observations including conventional in-situ data, satellite wind, tcvital, satellite radiances, and tail Doppler radar observations were assimilated. It was found that: a) the dual resolution hybrid improved upon the coarser, single resolution hybrid; b) vortex initialization and relocation in the control and relocation of the ensemble background improved the forecasts; c) using 4DEnVar DA in the TDR-involved cycles improved the intensity forecasts for early lead times compared to 3DEnVar DA; and d) the hybrid system improved intensity forecasts relative to operational HWRF during the intensification period due to the alleviation of the “spin-down” issue because the hybrid better analyzed the structures of an intensifying storm. These findings together with results from other cases will be presented at the conference.