One of the dire warnings of the effects of anthropogenic climate change is the increased frequency and severity of extreme weather events.
The recent spate of powerful hurricanes which it hit the U.S. and the Caribbean suggests that perhaps the frequency of hurricanes is increasing as predicted..
The Economist recently published data for hurricanes landing in the U.S. from 1850 onwards. These data suggested that the number of hurricanes is slowly decreasing while the number of category five hurricanes is slightly increasing. In this blog I will look at the same data with a little more rigorous analysis.
The chart shows the data presented by the Economist for all hurricanes landed on the U.S. The red line shows the best fit by linear regression which fits in well with the analysis showed in the Economist. The number of hurricanes hitting the U.S. is decreasing at a rate of 0.05 hurricanes per year. However the correlation is not good, with a regression coefficient of only 0.112 (a value of 1.0 indicates a perfect correlation).
The uncertainty in the slope can be calculated. The green lines on the chart show the same data with the slope at one standard deviation above and one standard deviation below the best fit line. The true trend has a 68% probability of being between the two green lines. The slopes of the two lines are a decrease of 0.006 and 0.040 per year. The Economists interpretation that the frequency of hurricanes is decreasing cannot be concluded with confidence.
The Economist also claimed that the frequency of major hurricanes is slightly increasing. A similar analysis, using the sum of category 3 to 5 hurricanes (the data for category 5 alone are too sparse for analysis) shows that the frequency of such hurricanes has a 68% probability of being between 0.019 and -0.004 , i.e. between increasing at 0.019 per year and decreasing by 0.004 per year. Again the data do not support the conclusion that such hurricanes are increasing in frequency.
To see if hurricanes are becoming more severe, the ratio of categories 3 to 5 to all hurricanes can be plotted. This is shown in the chart below. Again the red line is the best fit and the green lines represent the 68 percent confidence limits in the trend. The data do support the idea that the severity of hurricanes landing in the U.S. is increasing, but slowly.
The limited amount of data and the poor correlations mean that definitive conclusions cannot be reached. The data for all hurricanes in the U.S. and the Caribbean region would give more definitive answers