Scott K. Johnson
Science is relatively straightforward when you can compare a model prediction to data and assume that the former accurately represents our theories and the latter accurately represents what’s going on in nature. But the data isn’t always up to the task of providing a good representation of the natural world for one reason or another. That’s part of what makes the relationship between climate change and tropical cyclones something that requires a bit of explanation.
From climate-model projections (and basic physics), some things are pretty well established: sea level rise worsens storm surge damage, warmer temperatures increase the atmosphere’s ability to hold water vapor (and dump it as rain), and warmer temperatures increase the energy available to make tropical cyclones stronger. Sea level rise is obviously easy enough to measure, but checking records of tropical cyclones for trends in intensity is much more complicated.
This is partly because tropical cyclones in different ocean basins can be affected by natural variations in regional atmospheric patterns and ocean temperatures. But the larger problem is that our methods of collecting data on tropical cyclones have changed—and improved—over time. If a trend appears in the data, is it real or is it just an artifact of the advent of comprehensive satellite coverage? (A trend possibly extended by improvements in satellite resolution or instrumentation.)
For these reasons, the expectation that the strongest tropical cyclones should become more common has been hard to test against the data. In 2013, a group of researchers tried to overcome this limitation by producing a consistent, apples-to-apples dataset for the satellite era—which covered 1982-2009 at the time. This included an automated algorithm for identifying and quantifying storms and resampling the data at a common resolution.
That study found a trend toward stronger storms globally, but it fell short of the usual standard for statistical significance. The strength of the trend matched predictions, but a record of 28 years was not quite long enough to expect that trend to stand clear of the noise of natural variability.
More time, more data
In a new study, the researchers have extended their approach to span 1979-2017 and about 4,000 tropical cyclones. And the global analysis does now show a statistically significant trend toward stronger tropical cyclones. For example, if you compare the first half of that 39-year period to the second half, the share of storms coming in at Category 3 or higher has increased at a rate of about 8 percent per decade.
Somewhat surprisingly, the increase peters out at Category 5. However, this may just be a function of resampling all the data at 8-kilometer resolution—stronger storms tend to have smaller eyes and so are harder to pick out at this resolution. The data does show a trend toward smaller eye size, so that makes some sense.
Things get messier when you drill down to individual ocean basins, though. The biggest effect here is seen in North Atlantic hurricanes. Typhoons in the western North Pacific, on the other hand, actually show no change.
Why the differences? Breaking the dataset into smaller pieces increases the prevalence of noise—just as looking at a shorter timespan would do. And global warming isn’t the only factor at play. Regional variability that may more or less cancel out at a global scale has an impact when you zoom in. Volcanic eruptions and even greenhouse gas emissions can also have effects that vary from place to place—as a study published earlier this month showed, for example.
For the western North Pacific, specifically, the researchers note that the poleward migration of storm tracks there may also be contributing to the lack of trend in intensity. That migration has actually edged into an area where conditions are less ideal for typhoons.
“Ultimately,” the researchers write, “there are many factors that contribute to the characteristics and observed changes in [tropical cyclone] intensity, and this work makes no attempt to formally disentangle all of these factors.” But, they say, “the consistency of the trends identified here with expectations based on physical understanding and greenhouse warming simulations increases confidence that [tropical cyclones] have become substantially stronger, and that there is a likely human fingerprint on this increase.”