The frequency of tropical storms in the North Atlantic region varies markedly on decadal timescales1,2,3,4, with profound socio-economic impacts5,6. Climate models largely reproduce the observed variability when forced by observed sea surface temperatures1,8,10. However, the relative importance of natural variability and external influences such as greenhouse gases, dust, sulphate and volcanic aerosols on sea surface temperatures, and hence tropical storms, is highly uncertain11,12,13,14,15,16. Here, we assess the effect of individual climate drivers on the frequency of North Atlantic tropical storms between 1860 and 2050, using simulations from a collection of climate models17. We show that anthropogenic aerosols lowered the frequency of tropical storms over the twentieth century. However, sharp declines in anthropogenic aerosol levels over the North Atlantic at the end of the twentieth century allowed the frequency of tropical storms to increase. In simulations with a model that comprehensively incorporates aerosol effects (HadGEM2-ES; ref. 18), decadal variability in tropical storm frequency is well reproduced through aerosol-induced north–south shifts in the Hadley circulation. However, this mechanism changes in future projections. Our results raise the possibility that external factors, particularly anthropogenic aerosols, could be the dominant cause of historical tropical storm variability, and highlight the potential importance of future changes in aerosol emissions.
Access optionsAccess options
Subscribe to Journal
Get full journal access for 1 year
only $15.58 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
We are very grateful for discussion and input from N. Bellouin. We also thank G. Jones, P. Halloran and J. Hughes for setting up and running the CMIP5 integrations of HadGEM2-ES. We acknowledge the modelling groups, the Program for Climate Model Diagnosis and Intercomparison (PCMDI) and the World Climate Research Programme’s Working Group on Coupled Modelling (WGCM), which is responsible for CMIP, and we thank the climate modelling groups (listed in Supplementary Fig. S5 of this paper) for producing and making available their model output. For CMIP the US Department of Energy’s Program for Climate Model Diagnosis and Intercomparison provides coordinating support and led development of software infrastructure in partnership with the Global Organization for Earth System Science Portals. The authors were supported by the Joint DECC/Defra Met Office Hadley Centre Climate Programme (GA01101) and the EU FP7 THOR project.