A modeling study of the extreme 2005 Atlantic hurricane season relative to recent years

The 2001-2006 base period has been adopted to investigate the occurrence of extreme hurricane activity in 2005 using the RegCM3 regional climate model. The study focuses on the investigation of the relationship between Atlantic hurricane activity and evolution of African easterly waves (AEWs) during the Atlantic tropical cyclone season (June-November). Filtered 700mb meridional winds are used to identify AEW activity during the research period. Sea surface temperature (SST) sensitivity experiments using the RegCM3 model were also conducted to investigate role of SSTs in the occurrence of such an extreme hurricane season during 2005.

RegCM3 realistically simulates the basic characteristics of the propagation, frequency and development of AEWs over the Western Africa-eastern Atlantic zonal corridor between near 10 and 20 degrees latitude. Compared to observations (NNRP1 and GFS reanalyses), RegCM3 exhibits moderate positive bias regarding the intensity of the AEWs. The reasons for this behavior are explored in the study.

The results show that the dramatic increase in AEW activity and their northward shift over the African exit region in 2005 was associated with the development of a warm pool (2ºC+) of SST anomalies adjacent the African coastline. The results also show change in the location and strength of the African easterly jet (AEJ). The 2005 SSTs resulted in a more northerly and weaker AEJ in comparison to the climatological SST RegCM3 run. As shown in previous studies, this study also confirms that a more northerly and weaker AEJ is consistent with stronger AEWs and enhanced precipitation over the Sahel region. It is suggested that the north-south basin-scale structure of the SST anomaly dipole pattern resulted in increased longevity of the AEWs but did not have much impact on the frequency of these disturbances. We infer that longevity of the AEWs is a more relevant factor in determining Atlantic hurricane activity than frequency. This has important implications and applications for stochastic models used to predict future Atlantic hurricane activity.