We are well into hurricane season with a dangerous storm lurking off the coast of Florida and now poised to make a run up the east coast of the United States. As happens every year at this time, the appearance of hurricanes provokes speculation about the role of climate change in the formation of these destructive storms.
Climate change theorists assert that warming ocean temperatures are increasing the number and strength of hurricanes that form and make landfall in the United States. As David Leonhardt writes this week in the New York Times, “The frequency of severe hurricanes in the Atlantic Ocean has roughly doubled over the last two decades, and climate change appears to be the reason.” He cites some statistics to support this conclusion, though his review of the facts is far from thorough.
As he notes, the underlying science holds that hurricanes develop in warm ocean waters in late summer, so that over time rising ocean temperatures will generate rising numbers of hurricanes, and stronger ones as well. According to scientists, average ocean temperatures have increased by about one degree Fahrenheit over the past one hundred to a hundred and fifty years, a finding that provides a foundation for the “hurricane hypothesis.” Thus, we hear the refrain that global warming is causing more storms with higher wind speeds, and that these storms last longer, are more destructive, and make landfall more often than in the past.
It is a plausible hypothesis and, unlike many claims in this area, is capable of being tested against the facts. The evidence for it turns out to be quite thin—at least in relation to the certainty with which it is usually expressed.
Looking at the historical data, one does not find a startling increase in hurricane activity in recent decades, and only modest evidence to suggest that hurricanes in the Atlantic basin are increasing either in number or severity.
Mr. Leonhardt cites a study from the National Hurricane Center that shows, based upon moving averages, an increase of Category 4 and 5 storms in the Atlantic from about 1.5 per year in the 1960s to one per year in the 1970s and 1980s, and then to about two per year in the last decade. (Note: These are all the hurricanes that form, not just those that make landfall.) The doubling in frequency that he cites is in comparison to the rates of the 1970s and 1980s, though the increase is more modest if we take the rates of the 1960s as a point of reference. In any case, these increases are not very impressive because the numbers are so small: Category 4 and 5 hurricanes are uncommon phenomena. Nor is the historical pattern consistent with the claims of steadily warming ocean temperatures, since the number of powerful storms increased in the 1950s and 1960s, declined in the 1970s and 1980s, and then increased again after 1990. Thus, we have a weak pattern but also a ragged one that is out of joint with the underlying theory.
The National Hurricane Center, a division of the National Weather Service, has compiled reliable information on hurricanes going back to the middle of the nineteenth century—though the information the nhc collects has grown much more reliable in recent decades with the development of satellite imagery and ever-more sensitive instruments with which to measure the strength and windspeeds of hurricanes. There is no shortage of information to test the claims about increasing hurricane activity.
- Are hurricanes in the Atlantic Ocean increasing in frequency with the passage of time?
The Hurricane Research Division (hrd) of the National Ocean & Atmospheric Administration has tabulated the annual number of Atlantic hurricanes going back to 1851, though it is likely that many hurricanes in the early years were missed if they did not make landfall. For that reason, it makes sense to chart the pattern going back to around 1950 when the modern era of hurricane tracking and measurement got underway.
From 1950 through 2018, the hrd tells us that on average 6.3 named hurricanes formed per year, with a high of fifteen storms in 2005 (the year of Katrina) and a low of two in 1982 and 2013. A named hurricane is one strong enough to be classified between 1 and 5 on the Saffir-Simpson Hurricane scale. Again, these are storms that formed in the Atlantic, but not all of them made landfall. Broken down decade by decade, the averages look like this:
1950-59: 6.9 per year
1960-69: 6.1 per year
1970-79: 5.0 per year
1980-89: 5.2 per year
1990-99: 6.4 per year
2000-09: 7.4 per year
2010-18: 7.0 per year
Over the past two decades, 2000–18, there has been an increase of one to two named hurricanes per year over the rates experienced during the three decades between 1960 and 1989—or, expressed differently, an increase of perhaps one per year over the long-term average of 6.3 going back to 1950. At the same time, the current rates of seven named hurricanes per year are very close to the rates (6.9 per year) experienced during the 1950s.
As with the data cited by David Leonhardt above, there has been a real increase in the number of hurricanes forming over the past three decades, but only if we compare these rates to the unusually low number of hurricanes formed in the 1970s and 1980s. If we move the reference point back into the 1950s, the rate of increase flattens out and nearly disappears—with average annual rates of 6.9 per year from 1950 to 1959 and an identical annual rate between 1990 and 2018.
Conclusion: There has been a modest increase in the number of hurricanes formed per year since 2000, but these rates are not significantly higher than the long-term average and are very close to the rates experienced in the 1950s.
- Are more hurricanes making landfall in the United States with the passage of time?
The hrd maintains an accurate list of hurricanes making landfall in the United States going back to 1851 and running through 2018. On average over this hundred-seventy-year period, between one and two hurricanes made landfall per year in the United States. The busiest years for hurricanes since 1950 were 1985, 2004, and 2005, as six named storms made landfall in each of these years. The busiest decade going back to 1850 was the 1940s, when twenty-six hurricanes made landfall; more recently, the busiest decade was between 2000 and 2009, when nineteen hurricanes made landfall.
Below is the decade-by-decade enumeration going back to 1950:
1950-59: 18 hurricanes made landfall
1960-69: 15 hurricanes made landfall
1970-79: 12 hurricanes made landfall
1980-89: 16 hurricanes made landfall
1990-99: 14 hurricanes made landfall
2000-09: 19 hurricanes made landfall
2010-18: 11 hurricanes made landfall
Average by decade, 1950 to 2018: 15
Average by decade, 1990-2018: 15
Average by decade, 1950-1989: 15
Conclusion: There has been no long-term increase in the number of named hurricanes making landfall in the United States.
- Are Atlantic hurricanes growing more powerful with the passage of time?
Over the hundred-seventy-year period, just four Category 5 hurricanes (the most powerful of all storms) have made landfall in the United States: The Labor Day hurricane that hit the Florida Keys in 1935; Hurricane Camille, which hit the Gulf coast in 1969; Hurricane Andrew, which hit south Florida in 1992; and Hurricane Michael, which hit Florida and Georgia in 2018. These events appear unrelated to changes in ocean temperatures.
With regard to the most damaging and destructive hurricanes to hit the United States, the hrd summarizes the history as follows:
The largest damage caused by a tropical cyclone as estimated by monetary amounts has been Hurricane Katrina (2005) as it struck the Bahamas, Florida, Louisiana, Mississippi, and Alabama: US $40.6 billion in insured losses, and an estimated $108 billion in total losses. This compares to $50 billion for Sandy (2012) and $37.5 billion for Ike (2008).
However, if one normalizes hurricane damage by inflation, wealth changes, and coastal county population increases, then Katrina is only the third worst, after the 1926 Great Miami Hurricane and the lethal 1900 Galveston Hurricane. If the 1926 storm hit in 2005, it is estimated that it would cause over $140 billion in damages, and the 1900 storm about $92 billion.
The two most destructive storms to hit the USA occurred in 1900 and 1926, long before the era of rising ocean temperatures.
From 1950 to 2018, an average of 2.7 major hurricanes have formed per year in the Atlantic basin, with highs of eight major hurricanes in 1950 and seven in 2005. There were several years in the period, most recently in 2013 and 1994, when there were no major hurricanes in the Atlantic. (The hrd defines a major hurricane as a storm classified as 3, 4, or 5 on the Saffir-Simpson Scale.)
The decade-by-decade pattern looks like this:
1950-59: 3.9 major hurricanes per year (Category 3, 4, and 5)
1960-69: 2.8 major hurricanes per year
1970-79: 1.6 major hurricanes per year
1980-89: 1.7 major hurricanes per year
1990-99: 2.5 major hurricanes per year
2000-09: 3.6 major hurricanes per year
2010-18: 2.9 major hurricanes per year
Average by decade, 1950-2018: 2.7 per year
Average by decade, 1950-1969: 3.3 per year
Average by decade, 1970-1989: 1.7 per year
Average by decade, 1990-2018: 3.0 per year
A total of twelve Category 4 and 5 hurricanes have made landfall in the United States since 1950, following no particular historical pattern or trend.
There were three such storms in the 1950s (King in 1950, Hazel in 1954, and Gracie in 1959); three in the 1960s (Donna in 1960, Carla in 1961, and Camille in 1969); none in the 1970s; one in the 1980s (Hugo in 1989); one in the 1990s (Andrew); one in the 2000s (Charley in 2004); and three in the 2010s (Harvey and Irma in 2017, and Michael in 2018). There were four such hurricanes in the 1940s, the business decade of thetwentieth century for Category 4 and 5 storms. Again, there is no pattern or trend in these hurricanes that would match them up with trends in ocean temperatures.
Conclusion: There has been a slight increase in the frequency of powerful hurricanes since 1990, but mostly in relation to the numbers of such storms from 1970 to 1989, a quiet period for hurricane formation. The frequency of powerful hurricanes from 2000 to 2018 (3.3 per year) mirrors the rates experienced from 1950 to 1969 (also 3.3 per year). Moreover, there is no pattern or trend in the frequency of Category 4 and 5 hurricanes making landfall over the 1950-2018 period.
How, then, in view of these data, should we assess the claims that Atlantic hurricanes are increasing in numbers and strength in recent decades in response to rising atmospheric and ocean temperatures, and are also making landfall at increasing rates?
There has been a modest increase in the frequency of Atlantic hurricanes in recent decades along with a slight increase in their strength from year to year, but no increase in the number of hurricanes making landfall in the United States and no increase since 1950 in the number of the most powerful hurricanes (Category 4 and 5 storms) to hit the U.S. mainland. Moreover, any trend that we find in the frequency and strength of hurricanes in the past few decades is mostly washed out when we compare those rates to the ones experienced in the 1950s and 1960s. This suggests that the frequency and strength, though perhaps increasing of late, are but loosely related to recent measured increases in Atlantic Ocean temperatures.