Impact of Issues of Vortex Initialization and HWRF Model Errors on Hurricane Inner Core Data Assimilation and Intensity Prediction

Inner core observations collected from recent field campaigns such as ONR TCI and NOAA IFEX have provided a unique opportunity to describe the 3-dimensional hurricane inner core structures. In our early study for hurricane Patricia (2015), using the newly developed hybrid EnKF-Var data assimilation (DA) system for HWRF (Lu and Wang, 2016), the inner core observations (including TCI dropsondes, IFEX SFMR, TDR, flight level data together with the high resolution CIMSS Atmospheric Motion Vectors (AMV)] describes the storm structures well.  However, the intensity forecasts initialized from the analysis produced from DA showed spin-down issue.  Therefore, the primary goal of this study is to explore how and why spin-down occurs.  In particular, we explore how issues in the vortex initialization (VI) which is integrated with the hybrid DA impact the DA and how errors in the model can contribute to the spin-down issue.  

Experiments are conducted for hurricane Patricia (2015) while Patricia went through rapid intensification and when all the TCI an IFEX inner core observations are available. It is found that: a) assimilating all the available inner core observations can produce a complete 3D analysis of the storm structure which is consistent with the observations at the surface, middle levels and the outflow regions; b) the dramatic size contraction during the inner core observation DA can result in “seemingly” inconsistent Vmax & MSLP change at the surface due to the spuriously large background storm from the cycling VI; c) the analysis produced by DA describes the TC structure much better than VI. The analysis produced by DA is also more consistent with the gradient wind balance whereas VI breaks the gradient wind balance; d) although the VI analysis is less consistent with observations and breaks the gradient wind balance, the spin-down issue is  less apparent than DA; e) frequent outputs from HWRF model suggests that  HWRF is not able to handle the better wind and pressure relationship obtained from the DA analysis and likely because the HWRF model is tuned to be consistent with the unrealistic VI analysis; and f) although wind structures above the surface and at surface are all corrected toward observations by DA, HWRF cannot maintain the large Vmax from DA due to HWRF model bias. These findings together with methods to mitigate the above issues will be presented at the conference.