Twelve years ago, when Hurricane Katrina put much of New Orleans under water, it touched off a massive dispute among scientists about the relationship between hurricanes and global warming.
Now, it's happening again. Harvey, with its unprecedented rains, has spawned an urgent debate over whether this is what a changing climate looks like.
Yet while the argument following Katrina lacked any satisfying resolution, today we may be getting closer to answering questions about how a changing climate worsens hurricanes.
At least two things have changed in the intervening years. One is the availability of new analytical approaches and more powerful supercomputers.
"The methodological frameworks were very much in their infancy at the time of Katrina in 2005," said Noah Diffenbaugh, a Stanford climate researcher who performs climate change "attribution" studies, seeking to determine how the probability of various weather events have changed due to the warming of the change.
There's another change. Today, despite recent charges from EPA spokeswoman Liz Bowman that they are "engaging in attempts to politicize an ongoing tragedy," many researchers are more willing to simply say that Harvey's record rains were worse because of our hotter, wetter climate.
"Climate change worsened the impact of Hurricane Harvey," stated Penn State climate scientist Michael Mann bluntly, in a much-quoted Facebook post about the storm.
The climate-hurricane debate of yesteryear
Shortly before and after Katrina, several researchers published studies suggesting that more intense hurricanes were linked to rising ocean temperatures. That made sense: Hurricanes draw heat energy from the oceans and turn it into powerful winds and rain.
But there was a big problem: Data. You don't know how strong a hurricane is unless you can measure it, and hurricane measurement techniques are highly inconsistent around the globe and were less reliable in the past than they are today. Accordingly, other scientists questioned the ability to detect strong hurricane trends.
The tone at the time was quite cautious. Even the scientists arguing for a climate connection tended to underscore that they weren't talking about Katrina, but, rather, about hurricanes in general.
Yet the context was hard to miss, as Katrina proved a devastating political event for George W. Bush, whose administration had made a habit of sowing doubts about climate change.
The researchers also emphasized repeatedly that global warming doesn't cause individual storm events. They're products of a swarm of atmospheric and oceanic conditions — and what perhaps matters most is the broader pattern of winds that steer a given storm toward its landfall in a particular location.
And nobody — at least, at the time — seemed to be saying that these steering winds were being altered by climate change.
Superstorm Sandy shifts the narrative
The tone began to change when another major U.S. city was flooded by another suspicious storm.
In the case of Superstorm Sandy in 2012, the storm's particular path — a beeline towards New Jersey, rather than out to sea — was abnormal. Atlantic hurricanes often "recurve" and flow away from the United States as they travel farther northward.
Sandy did the opposite.
In one analysis, NASA's Timothy Hall and Columbia University's Adam Sobel found that Sandy's sharp turn toward New Jersey is only expected to happen once every 714 years based on the history of Atlantic storms.
The result "implies either that the New York-New Jersey area simply experienced a very rare event (with climate change playing no significant role), or that a climate-change influence increased the probability of its occurrence," they wrote in a 2013 study on Sandy's angle of approach.
Some researchers, like Jennifer Francis of Rutgers University, suggested a potential climate role. Francis argued that the warming of the Arctic, at a much faster pace than the mid-latitudes, is slowing the jet stream and prompting extreme weather events when there's weak atmospheric flow. She specifically cited Sandy's strange behavior.
"We were just coming off of a record-smashing low sea-ice year in 2012," Francis told Climate Central in 2014. "The block that played such a key role in Sandy's 'weirdness' seemed like an amazing coincidence."
But other researchers strongly disputed Francis' jet stream idea — which continues to divide scientists even today.
Meanwhile, watching the floodwaters swamp New York, some observers began to make a relatively simple but resonant observation. Sea levels, they noted, are higher today than they were when similar storms affected New York in the past — and nobody really disputes that the warming of the planet is driving sea level rise.
That led to the inference that at least some of what Sandy threw at New York was in a sense, extra. And maybe that water made a difference.
As MIT's Kerry Emanuel, who published one of the key studies in the wake of Katrina, put it to the Post last week: "New York, when Sandy hit, the sea level was already about a foot higher than it was 100 years earlier. So if Sandy had hit in 1912, it probably would not have flooded Lower Manhattan."
Indeed, subsequent research found that the chances of a Sandy-style flood event are higher today and will only grow further as seas continue to rise.
Changing the burden of proof
Meanwhile, some scientists were questioning what they viewed as a default assumption that one must prove that a given storm is influenced by climate change — seeking to flip the question, and instead suggest that all storms today are so influenced.
Kevin Trenberth of the National Center for Atmospheric Research, has argued that storm thermodynamics — the heat and moisture that serve as fuel — are strongly climate dependent. In particular, Trenberth argues that hurricanes and other storms will rain more today because the warmer atmosphere carries more moisture.
"The environment in which all weather events occur is not what it used to be," wrote Trenberth and his colleagues John Fasullo and Theodore Shepherd in 2015. "All storms, without exception, are different."
All these threads come together with Harvey. We have a storm that rapidly intensified over warm oceans and had an exceedingly odd track that led it to move very slowly, lingering in place amid a lack of strong atmospheric steering currents and dumping record rainfall.
Some researchers today are taking Trenberth's broad approach, and simply highlighting that a warmer atmosphere holds more water vapor and thus, suggesting that climate change worsened Harvey's rains.
"Harvey was not caused by climate change, yet its impacts — the storm surge, and especially the extreme rainfall — very likely worsened due to human-caused global warming," said Stefan Rahmstorf, a researcher with the Potsdam Institute for Climate Impact Research, in a statement.
Rahmstorf said in a follow-up email that this is just basic physics, citing the Clausius-Clapeyron equation, which shows that the atmosphere holds more water vapor when it is warmer, setting the stage for more rainfall.
"It is about 7% more water vapor per degree C of warming," he said. "The [equation] tends to be a good approximation for extreme rainfall events because they tend to occur with moisture-saturated air masses."
Enter the attributionists
But it would be considerably more powerful to say that because of the warming of the planet, a storm like Harvey was statistically more likely to occur — and give the odds. And scientists just might be on the cusp of doing that.
Such analysis requires a so-called "attribution study," which often use myriad runs of high powered computer models to determine the odds of an event occurring with, and without, human-caused changes to the atmosphere. Attribution science has been evolving alongside the hurricane debate: One of the first such studies, published in 2004, examined an extreme and deadly 2003 European heatwave.
Such studies may allow researchers to say that some aspect of Harvey, such as its path or its rainfall, was more likely to occur in the current climate than in an unaltered one.
"I would actually be really surprised if you don't start seeing those kind of studies coming out within six to 10 weeks. My guess is it will be weeks, not months to years," said David Titley, a meteorologist at Penn State University who chaired a National Academy of Sciences report on the attribution studies in question.
Yet complex atmospheric and ocean phenomena like hurricanes remain a difficulty in such studies, in part because of the computing power required. Heat waves are far easier to attribute to a warming climate in a statistical sense.
"Attribution is still a young science, and there have been big disagreements in the field," said Columbia's Sobel. "This is still the bleeding edge of what we can do."
Still, Stanford's Diffenbaugh agreed that before long — if they haven't begun already — scientists will be looking very closely at the Harvey event, based on its thermodynamics (storm intensity and rainfall) and its "dynamics," or the broader atmospheric flow factors that led the storm to stall.
Titley thinks climate analyses will try to suss out the likelihood of Harvey's track, just as occurred with Sandy. "The fact that nothing in the atmosphere was there to kick it out or shear it apart or do anything with it, really is this kind of worst-case scenario here," he said. "So the question going forward is: Are these situations going to be more likely or did we just get unlucky?"
But not all scientists agree attribution studies are the way to go.
"I think they provide at best a limited view of the climate connections," said Penn State's Michael Mann by email. "Simple physics (effect of [sea-level rise] on storm surge) and simple thermodynamics (i.e. Clausius-Clapeyron) are valid whether or not we can trust the models to get the specifics dynamical linkages between climate change and extreme weather events right (and I'm deeply skeptical the models are up to this task at present)."
But in the meantime, watching the evolution of the debate over hurricanes and climate change, several things are clear. First, major storm events drive science. They force researchers to ask questions, and the ensuing inquiry improves our answers.
The second, meanwhile, is that we have had devastating flooding of a major city from a hurricane event in each of the last three presidencies, even as the climate hurricane debate has grown steadily more insistent. At this point, we surely have the right to wonder, just as scientists themselves are already doing, whether the burden of proof has shifted.