Numerous studies have analyzed tropical cyclone best track data for trends in tropical cyclone numbers and various measures of tropical cyclone intntensity. All recent work suggests that there is no current detectable trend in global tropical cyclone numbers, with numbers typically 80-90 per year (Emanuel 2005; Webster et al. 2005). Regional trends are somewhat more difficult to analyze, given the lower signal to noise ratio due to the high interannual variability of tropical cyclone numbers in many formation basins (e.g. McBride 1995). Analyses that have been performed show different trends in different basins. Most work has been performed in the Atlantic, since this is the basin with the best data. Trends in tropical cyclone indicators in the Atlantic have shown substantial positive trends since 1980, and these trends appear to be real (e.g. Mann et al. 2007). Kossin et al. (2007) show that since 1980, the Atlantic has experienced very large upward trends in an intensity-related variable, the Power Dissipation Index (PDI), a measure of the total integrated power in the storm. According to Kossin et al. (2007), trends in the northwest Pacific have been modest, have been downwards in the northeast Pacific and approximately flat in the other regions.
In other basins, where the data are poorer, there is more dispute regarding trends, particularly of variables like intensity that are more difficult to estimate, leading to uncertainty in detected trends. Intensity trends in the Australian region differ by geographical location, with the northwest region showing some increases in intensity since 1980, but the eastern region exhibiting no trend (Harper 2006; Hassim and Walsh 2008).
In contrast, Webster et al. (2005) found large trends in global tropical cyclone intensities in the period from 1970 to 2004, noting a doubling in the global number of intense category 4 and 5 storms over this period. The value of this paper was that it identified that there were such unexplained trends in the best track cyclone data, a fact that was not previously generally appreciated. Nevertheless, subsequent worked showed that these detected trends were likely at least partly the result of artifacts of the data. Kossin et al. (2007) showed that there was no apparent trend in global tropical cyclone intensity over that period, a time during which there was a substantial increase in category 4 and 5 storms numbers analysed by Webster et al. (2005). Instead, large decadal variations in global numbers of intense hurricanes were found. As already stated, there are a number of reasons to suspect that the best track data analysed by Webster et al. (2005) has artificial trends within it due to changes in observing practices. However, it is not clear to what extent correcting the existing data for inhomogeneities would alter the trends detected by Webster et al. (2005) and others: in other words, whether the trends would be eliminated, reversed or only modified.
In summary, the analysis of tropical cyclone trends is complicated by a lack of consensus regarding the state of the current tropical cyclone data used to determine such trends. The detected trends in the Atlantic ocean basin since 1980 appear to be real, however.
Was this article helpful?