Abstract
Abstract In 1994 the United Kingdom Meteorological Office (UKMO) introduced a new method for intitializing tropical cyclones (TCs) in its global model by use of “bogus” data. This has resulted in a major reduction in TC forecast errors and has made the model far more competitive in the field of TC forecasting. The unusually active Atlantic TC season of 1995 produced a stern test of the qualities of the model forecasts. A full assessment of UKMO forecasts shows that the mean forecast errors for TCs in the Atlantic in 1995 were lower than mean errors for the UKMO model in any previous year. Errors were also lower than those of the European Centre for Medium-Range Weather Forecasts model at all forecast ranges shorter than 120 h. In addition, assessment carried out at the National Hurricane Center (NHC), Miami, found that UKMO forecasts were better than all forms of guidance available to NHC, although they were not available to NHC in real time.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
