How well forecast were the 2004 and 2005 Atlantic and US hurricane seasons?

Abstract

The 2004 and 2005 hurricane seasons with seven intense hurricane landfalls on the U.S. and an estimated total damage bill approaching US$ 200bn rank as the most active and damaging consecutive hurricane years on record. This high activity was due to a combination of warm sea surface temperatures in the tropical North Atlantic, low vertical wind shear and steering currents that favoured storms being steered towards the U.S. rather than recurving at sea. Currently there are four main organisations that provide seasonal forecasts of Atlantic tropical cyclone activity: TropicalStormRisk (TSR), Prof W Gray, NOAA and the Meteorological Institute, Cuba. How well did the publically available seasonal outlooks anticipate the record-breaking Atlantic and U.S. hurricane activity in 2004 and 2005, and would business have benefited from acting upon the forecasts made? This paper compares the performance of forecasts (deterministic and probabilistic) issued by different centres at different lead times. It also shows why seasonal forecast precision may now be high enough to offer practical benefits.