Although, conventionally, hurricanes are measured by their peak intensity, how quickly they reach that intensity and how rapidly they approach land matters enormously. Over the last century, the world has made remarkable if unequal progress in reducing human mortality from storms, even as tallies of property damage have appeared to grow larger. Much of that progress is thanks to better forecasting and early warning systems — having a few days, rather than a few hours, to prepare makes even the most brutal storms much easier to endure and survive. A tropical storm isn’t an insignificant threat, and what became Otis surely would’ve damaged Acapulco even if it hadn’t ever intensified. But a Category 5 is a threat of a different order, requiring an entirely different scale of preparatory response. You simply can’t evacuate a city of one million in just a few hours — at least, it’s never been managed before.
But it wasn’t just the speed of intensification that has marked Otis as different; after all, hurricanes are now routinely slowing and growing much stronger as they approach land, powered by ever warmer ocean waters. It was the surprise of the storm’s transformation, with few of the conventional forecasting models predicting any significant intensification at all.
How could this happen? A week later, it remains a meteorological mystery. Climate scientists have been noting for years the pattern of rapid intensification; a landmark 2017 paper was called “Will global warming make hurricane forecasting more difficult?” and new research on the subject, focused on North Atlantic hurricanes, was published just two weeks ago. But though Otis fits the general pattern, nothing about the conditions that gave rise to it seemed to raise the proper alarms in the conventional models. Local ocean temperatures were high, to be sure, but not crazily so. Low wind shear contributed, too, allowing the storm to gain strength somewhat undisturbed — but there, too, the metrics were not that unusual. The storm was relatively compact, which may have meant micro conditions were more important than more large-scale factors, which meteorologists tend to emphasize.
“Otis was the worst nightmare for forecasters,” says Kerry Emanuel, the author of that 2017 paper, whose own model, which isn’t included in the major ensemble projections, did show some chance of intensification — though not as sharp as what actually unfolded. “We used to say that the nightmare was that you go to bed with a tropical storm 36 hours from landfall and you wake up the next morning and it’s a Cat. 4 and you’ve got 12 hours left to evacuate. But what actually happened was worse than that imagined scenario: You woke up with a Cat. 5 and it was only a few hours from landfall. It’s a terrible tragedy,” he says. “I can’t imagine anything worse.”
For Emanuel, the key lesson from Otis is that we need to assess the possible course of storms probabilistically — understanding that, whatever the median projection for the intensity or course of a given hurricane, there is some chance of some low-likelihood, high-risk surprise. He also mentions a few variables that are not yet properly integrated into most forecasting tools — for instance, the relative salinity of the ocean surface and the way that it can affect the temperature gradients helping influence the intensity of a given storm. And over time, our models will surely improve, as modelers incorporate lessons from this hurricane and from the unprecedented environment in which it so spectacularly emerged.