Les cyclones tropicaux et le changement climatique

Abstract

Results from observations and modelling studies, a number of which having been used to support the conclusions of the IPCC fourth assessment report, are presented. For the past and present-day (since 1970) periods, the increase of strong cyclonic activity over the North Atlantic Ocean appears to be in good correlation with increasing temperature of the ocean surface. For regions where observational data are of lesser quality, the increasing trend is less clear. In fact, assessing long-term changes is made difficult due to both the multi-decennial natural variability and the lesser coverage of observations before satellites were made available. Indirect observational data, such as those derived from quantitative estimations of damage caused by tropical cyclones, suffer from many artefacts and do not allow the resolving of the issue either. For the future, only numerical three-dimensional climate models can be used. They nevertheless run presently with too-large grid-sizes, so that their results are still not converging. Various simulations lead indeed to different results, and it is very often difficult to find the physical reasons for these differences. One concludes by indicating some ways through which numerical simulations could be improved, leading to a decrease of uncertainties affecting the prediction of cyclonic activity over the next decades.