The Impact of Supplemental Dropwindsonde Observations in the NCEP GSI/GFS System for Hurricane Joaquin (2015)

Hurricane Joaquin, which affected the central Bahamas as a major hurricane on 1-3 October 2015, was a major forecast challenge that resulted in above-average track forecast errors from much of the dynamical model guidance and official National Hurricane Center (NHC) forecasts. In particular, much of the guidance, with the exception of the ECMWF model, failed to bring Joaquin far enough south into the Bahamas in the short range and later showed Joaquin making landfall along the U.S. east coast from the Carolinas into the mid-Atlantic states for several forecast cycles initialized on 29-30 September.

In particular, the NCEP Global Forecast System (GFS) showed a large shift in its track forecast of Joaquin with the 0000 UTC cycle on 1 October, moving Joaquin over Long Island and southeastern New England, as compared to previous cycles that showed Joaquin making landfall in the Carolinas. This model cycle also marked the assimilation of numerous dropwindsonde observations from a synoptic surveillance mission flown around Joaquin by the NOAA Gulfstream-IV jet. Subsequent model runs after the 00Z cycle on 1 October continued an eastward shift in the track, and ultimately Joaquin remained well offshore of the U.S. East Coast.

Previous research has shown impacts in several cases from these supplemental dropwindsondes on the track, intensity, and structure of tropical cyclones. Here, the results from a data denial study using the NCEP GSI/GFS system will be presented to examine the impact of the G-IV dropwindsondes on the track, intensity, and structure of Joaquin, with a particular focus on the 0000 UTC cycle on 1 October.