Was Melissa made worse by climate change?

I’m seeing links to the blog from “news” articles with titles like “Scientists blame climate change for Hurricane Melissa: Catastrophic storm was made 4 times more likely by climate change.” To be very clear, I am not associated with that research – the links are to my economic impact estimates (not physical damage estimates as the articles improperly state) of Melissa. While I think it’s possible storms like Melissa are more likely due to climate change, I do not think it’s possible with our present state of knowledge or history to prove any given storm was or was not due to climate change – much less compute the odds.

I’d be very curious as to how that four times number was arrived at. The biggest problem with probability studies is their reliance on historical data – data that is highly uncertain, periodically reanalyzed, which results in the data sets used in complex studies becoming obsolete sometimes before they are even published.

One example is the November 1912 storm that made landfall in about the same area as Hurricane Melissa. This storm, which is vital for computing the return period statistics for western Jamaica, was classified as a 130 knot (Category 4) storm until 2007. In their 2007 reanalysis, NOAA’s AOML downgraded the storm to only 100 knots – a 30% decrease in wind speed, but a 55% decrease in dynamic wind pressure and damage! For what it’s worth, I think this downgrade was too much – 130 was probably too high, but the reported storm surge would argue for a larger number than 100,, on the order of 115 knots.

In any event, this goes to show how hard it is to assess historical storms, if reasonable efforts and different, yet valid approaches come up with 10 or 20% differences in intensity (and far more in impacts).

Let’s look at another aspect of how complex this is, and how technology has changed hurricane intensity measurements. In 2000, a NHC/AOML study recommended the conversion factor between the 700mb aircraft flight level to surface wind based on the new GPS dropsonde profiles as 0.90. However, by 2022, NHC operationally had changed that to 0.93 for a variety of reasons. What that means is the 172 kt flight level wind sampled in Melissa before landfall is reported as 160 knots (184mph), but 20 years ago would have been reported as 155 knots (178mph).

At this point it’s just making the rubble bounce if you’re talking about damage, but it does make a difference when talking about records, climate change, etc. because the estimated increases due to climate change are in the 5% range so bumping up the intensity of modern aircraft measurements by 3.3%, whatever the reason, matters, especially if you are trying to compare those results to storms from 10 or 20 years ago, much less 50 or 150 years before there was satellite data.

I could go on, especially about satellite based intensity measurements, but you get the picture: it’s complicated, and assumptions that go in to different data bases (and the changing nature of those data sets) means there is a lot of room for uncertainty.

So what about Melissa? When you blend it all together, using data since 1871, and the fact this is a La Nina year, Melissa was likely about a 200 year storm (0.5% chance in a given year). However, due the uncertainty in the historical data, it could have been as little as a 1 in 300 year (0.3%) or as much as a 1 in 125 (0.8% chance). I think I can safely say there is no way climate change made it four times more likely. If we get another Melissa hit western Jamaica in the next 20 or 30 years then maybe you can start to argue that, but not based on what we know now.

To be clear, I think it’s likely, all things equal, it is quite possible Melissa was made worse by our changing climate. The theory is sound: more energy in the oceans means that rapid intensification is more likely, and brief episodes of favorable conditions means that some storms that otherwise might not have spun up have the chance to not only spin up, but become stronger than they would have 50 or 100 years ago. But the details of how and why a particular storm does what it does are hideously complex, and we can only talk about probabilities.

And therein lies the problem: while the theory is sound, and the data shows the trends are towards stronger hurricanes, our historical data is too uncertain, and the changes in measurement technology too great, to put numbers on it. To be blunt, saying a “Melissa” type storm is four times more likely now than 100 years ago is simply statistical absurdity. In my view, saying these kinds of things hurts the arguments that we need to do something about climate change because they are so exaggerated.

None of this is to diminish the horrific impacts we are seeing from Melissa. It is likely to be a record setting storm, and it will take years if not a decade for the region to recover. But science it what it is – and is an entirely separate question from the human tragedy. There is no need to exaggerate the storm characteristics in order to see that.

---

If you like this kind of commentary and analysis, by far the best way is to subscribe to the emails, which is the only really reliable way to get them. The commentary, maps, and snark (when appropriate) should be expanding to winter storms and other topics later this year. You can also donate to keep it going at this the same link, donations are responsible for the new automated maps subscribers are seeing previews of.

You can also follow us on X/Twitter (@EnkiResearch), Telegram, BlueSky, or even Facebook if that’s your preferred social media mudpit.