Evaluation of Tropical Cyclone Relocation on Operational NCEP GFS/GDAS

With the ever-increasing resolution of global numerical weather prediction models and more realistic representations, it is becoming increasingly important to focus on tropical cyclone (TC) vortex initialization in order to realize improved intensity and track guidance. Currently, the operational GDAS/GFS uses a vortex relocation and minimum sea-level pressure assimilation provided by operational TC agencies. Recent sensitivity experiments have demonstrated for newer versions of the GFS that the vortex relocation leads to an increase in track forecast error for some cases. A qualitative evaluation of Hurricane Joaquin was preformed to determine the mechanism behind the increase in forecast track error. The 2015 hybrid 3D EnVar-based GDAS system initialized the operational resolution GFS was used as a control to be compared to an experimental run excluding the vortex relocation. The comparison revealed small but crucial differences, allowing the experimental runs without vortex relocation to correctly forecast a more southwesterly track towards the Bahamas early in the forecast resulting in a more easterly track later in the forecast and no landfall on the United States coast. Here, we explore the differences that yielded the significantly different forecast tracks.