Tropical cyclones (hurricanes and typhoons) are the natural events at the top of the scale for natural disasters. With a concern for the consequences of an increase in global temperatures it is natural for people to be concerned about any effect of global warming on either their frequency and/or their intensity. However, whatever global warming means there is one thing it does not mean and that is that temperatures area increasing uniformly everywhere on Earth in all seasons of the year at all times of the day and night.

The first question to ask on the issue of global warming and tropical cyclones is whether there has been a temperature increase in the regions where they form and develop during the times of the year where they form and develop. The regions where they form and develop is over the ocean during the later summer at about 30° of latitude north and south.

The following displays from the website of NASA's Goddard Institute of Space Studies show estimates for the seasonal as well as the global average trends.

The above information illustrates the fact that the global average temperature has increased over the last half century because of an extraordinary increase in temperatures in central Siberia and Northwest Canada during the winter. If the information separated day and nighttime temperatures the role of central Siberia and northwestern Canada would be even more sharply delineated. The places in the northern hemisphere which have the coldest temperatures are becoming less cold. This is not an adverse development.

The displays reveal that the sea surface temperature has not increased notably in the regions and in the seasons where tropical cyclones form and develop. The color scale is not linear so it is hard to discern precisely how much the temperature has increased in the relevant areas and how that compares with the natural variation.

Such information is available for land temperatures in the relevant latitude zones. This is shown below.

The t-ratios indicate that the observed temperature trends are not significantly different from zero at the 95 percent level of confidence.

After having reviewed the temperature trends in the region where tropical cyclones form and develop it is now appropriate to look at the time series statistics on hurricanes.

Visually there does not appear to be any significant trend in the total number of hurricanes on the number of category 3 or higher. The visual impression is confirmed for a compilation of the averages by decades, as shown below.

The trend in the more severe hurricanes is biased by changes in observational accuracy over time. Hurricanes fluctuate in severity over their life cycle. They are classified according to their maximum severity. Over time the tracking and observation of hurricane characteristics has improved so that over time there has been a greater opportunity to catch a hurricane in a higher level of severity.

The wind speed statistics for hurricanes in the Atlantic definitely do not show a perceptible upward trend. In fact, if anything according to Landsea's data they are decreasing.

Tornadoes

Although both tropical cyclones and tornadoes are vortices there are distinct differences which indicate quite different mechanisms for their formation and development. Tornadoes are much, much smaller in spatial extent than tropical cyclones and their wind speeds are much higher than those of tropical cyclones. The wind speeds of the hurricanes are awesome but most of their damage and loss of life comes from the storm surges accompanying them.

The statistics on tornadoes demonstrates the effect of improved surveilance. The time series on severe tornadoes shows no trend, as illustrated below:

But the observed numbers for all tornadoes do show an upward trend:

The most likely explanation for this trend is that the increased effort for providing advanced warning is picking up weak tornadoes that otherwise would not be observed.