pdf |
Seasonal Tropical Cyclone Forecasts 
(Article in the WMO Bulletin October 2007 Issue)
In an article in the October issue of the WMO Bulletin, a group of tropical cyclone research experts composed of Dr Christopher Landsea (National Hurricane Center, Miami, FL), Anthony Barnston (Columbia University), Dr Philip Klotzbach (Colorado State University) and led by Dr Suzana Camargo (Columbia University), describe the current status and future outlook for operational seasonal tropical cyclone forecasts. Media interest and that of the general public in seasonal TC forecasts has increased markedly in recent years, especially after the active 2004 and 2005 hurricane seasons in the Atlantic and the heightened incidence of disastrous landfalling tropical cyclones in the Pacific in 2004.Statistical forecasts of the number of tropical cyclones that will occur during the season have come a long way since first being issued in the early 1980s.Along with predictions of total seasonal activity, some groups now provide individual monthly forecasts and predictions of probability of landfall. With the increasing availability of global datasets such as the NCEP/NCAR and ECMWF reanalysis, it is expected that statistical forecasts of tropical cyclone activity will continue to improve. An updated and homogeneous quality global best-track dataset would also contribute to more skillful forecasts.Computer models that are run for many months are now also providing skillful seasonal tropical cyclone forecasts for various tropical cyclone basins.For example, a combination of three seasonal forecasting systems at the European Center for Medium-range Weather Forecasts (ECMWF), the UK Met Office, and Météo-France has had more success than the statistical models in forecasting the active 2005 Atlantic season and the not-active 2006 season. A future goal of forecasting landfall probabilities using these computer models will require research on the limits on predictability, as well as faster computers. For shorter time periods (i.e,. weekly, monthly) the improvement of the dynamical and statistical models in forecasting large-scale waves such as the Madden-Julian Oscillation, and their effects on tropical cyclone activity should lead to operational forecasts in the near future. |
