Hurricane Ian - a "500-year" Storm?
The history of major Florida hurricanes suggests another Ian should happen about once every 80 years - and another Charley about once every 30 years
This essay was to be about recent news articles describing "the big one" - a superstorm said to be due in California soon - and how to predict the likelihood and intensity of rare, major events.
But Hurricane Ian has made this more than a theoretical discussion. The devastation is overwhelming and the loss of life and property are on a scale rarely seen - but just how rarely? Some have been quick to say that this was a "500-year" event, implying that we never could have predicted it, and really don't have to plan for the next one, because there won't be another one like Ian for 500 years.
The use of the term is appropriate - the threat of disasters of a given intensity are often described in terms of likely return time or expected frequency - but the data on hurricanes in Florida do not support that 500 number.
The simplest and most intuitive method for estimating likelihood of major disasters is to look at existing data on frequency and intensity and use those data to extrapolate into the future.
But what numbers should we use to characterize a hurricane? Disaster reports often focus on number of fatalities and total value of property damage, and those are clearly the most important in terms of the human tragedy. But those numbers are also affected to a large degree by population density and property values under the storm's track, numbers that vary widely from place to place and over time, especially in a rapidly growing state like Florida.
Ian was so tragic in terms of loss of life and property in part because it made landfall on a major city and also in an area with high property values, and with much of that value built on a set of very fragile and also thickly settled barrier islands (like Fort Myers Beach and Sanibel/Captiva).
If loss of life and dollars of damage are so affected by factors outside of the strength of the storm itself, what can be measured consistently and be used to judge trends in storm frequency and intensity alone?
The physical strength of a hurricane can be described by both category (the 1-5 scale we hear on the weather reports) and by maximum sustained winds at landfall (in knots - The marine designation for speed - one knot is 1.15 miles per hour). Ian was a category 4 storm (almost a 5) with sustained winds at landfall of 150 knots. Have there been other storms of similar strength?
The data set on this is managed by the National Hurricane Center, part of NOAA (National Oceanic and Atmospheric Agency), and analyzed through a project called HURDAT. An accessible list of hurricanes of category 3 and above that have hit Florida is available here. That list is used in the analyses that follow.
Since 1851, there have been a total of 38 hurricanes of category 3 and above (Florida landfall only). With a total record covering 171 years, the likelihood of such a storm is 38/171 or 22.2% per year. Divide 1 by this number (invert it) and the likely return time, or time between such storms, on average, is 4.5 years.
Completing this table for category 4 and 5 storms you get return times of once every 10.7 years for category 4 and above, and once every 57 years for category 5. Not even close to 500 years.
Data on maximum sustained winds at landfall are divided more finely, into categories by 5 knots between 100 and 160 knots. On this scale, Ian is the second strongest storm to hit Florida.
With more available data points, a graph can be drawn that shows the change in return time or frequency as a function of windspeed at landfall. Hurricanes with windspeeds of 110 knots and above should occur about once every 8 years. Storms of 130 knots and above should occur about once every 30 years. Storms with sustained winds at landfall of 150 and above (2 storms, including Ian) should occur somewhere in Florida about once every 80 years.
The inverse relationship between frequency and intensity seen in both of these data sets (frequency declines and time between events increases as intensity increases) is seen for almost all kinds of natural disasters (floods, droughts, etc.).
And what of climate change? Reports often say that the frequency and intensity of hurricanes are increasing. The Florida data would support this. Plotting windspeed at landfall of each storm for the year in which it occurred, there is an upward trend, with average windspeed increasing over time from just over 100 knots to just over 120 knots (~ a 20% increase).
No one storm proves the impact of a changing climate system, but it may be worth noting that Charley, a category 4 hurricane with maximum sustained winds at landfall of 130 knots, followed nearly the same path as Ian. That was in 2004, just 18 years ago.
This trend is mirrored in a global data set that estimates total energy driving all tropical storms in either the Atlantic or Pacific (Accumulated Cyclone Energy or ACE). This data set for the Atlantic clearly identifies the hurricane season (August to October), and also shows an upward trend in total annual ACE since 1900 of about 70%.
The tragedies caused by hurricanes and other natural disasters rightly fix our attention on the immediate need for aid to those areas affected. Often lost in the aftershock is what might be done to either mitigate or prepare for the next one.
Claiming a 500-year return time argues against doing much of either. The actual data capturing the history of past storms for Florida suggests another Ian should happen somewhere in Florida every 80 years or so, and another Charley every 30 years or so. The return time for all storms of category 3 or higher is once every 4-5 years.
Given the trends in storm intensity over time shown in the figures above, we might expect those estimates to be conservative, with frequencies increasing and return times decreasing in the future. Rising sea levels could increase flood damage for each level of storm intensity as well.
Using existing information to predict the actual likelihood of hurricanes of different intensities might help decision-making about how to prepare for the next Ian or Charley. One can imagine changes in evacuation procedures, and in city planning, zoning, and building codes, but politics and policies are beyond the scope of these essays.
Whatever steps toward recovery are taken following the devastation of Ian, we can hope that those decisions will be based on an accurate assessment of the future frequency and intensity of major storms.
Sources
Data on hurricane occurrence and strength are accumulated by NOAA's National Hurricane Center: https://www.nhc.noaa.gov/
An accessible listing of Florida hurricanes, and the one used in these analyses, can be found here: https://en.wikipedia.org/wiki/List_of_Florida_hurricanes
The HURDAT project is described here: https://en.wikipedia.org/wiki/HURDAT
The Accumulated Cyclone Energy (ACE) database is described here: https://en.wikipedia.org/wiki/Accumulated_cyclone_energy
And the data are available here: https://psl.noaa.gov/gcos_wgsp/Timeseries/Hurricane/hurr.atl.ace.data