In All Probability: Climate Change and the Risk of More Storms Like Sandy

By modeling climate change's local effects, we may finally be able to grasp what it means.


Earlier this year, the journal Nature Climate Change published a paper that measures hurricane behavior in our warming world. The study was as innovative as it was prescient. Combining models to simulate thousands of hurricanes in likely future climates, the authors discovered that New York City, their test case, was at risk. And increasingly so.

The city's most severe storm flooding events historically have been quite rare. According to the paper, hurricane storm surge -- roiling coastal waters driven inland which are responsible for most of the devastation and destruction during storms -- can, in New York City, reach the exceptional height of about six feet once every century, and 10 feet once every 500 years. But as the Earth warms and sea levels rise, the study found, over the next few generations these extreme swells could batter New York Harbor as often as once every 3 years and 25 years, respectively.

That was in February.

Then, on the night of October 29, Hurricane Sandy drove a 14-foot wall of seawater deep into Brooklyn and lower Manhattan, flooding seven subway tunnels and shorting most circuits below Midtown. It eclipsed the city's previous storm surge record, set in the early 19th century, by nearly three feet, leaving many New Yorkers paralyzed, wet, and in the dark.

"This is not just coincidence," Ning Lin, the paper's lead author and assistant professor of civil and environmental engineering at Princeton University, told me. "We knew that New York City was vulnerable."

At the spout of a funnel formed by Long Island and the New Jersey Shore, the so-called New York Bight renders lower Manhattan a basin for water with no place else to go. A stunted bulkhead, not quite five feet above mean sea level, does little to safeguard the low-lying coast. (Even with all of these threats, the city only ranks eighth on a list of U.S. areas most vulnerable to hurricanes.)

"Already, a bad storm today is way beyond our ability to deal with effectively," Michael Oppenheimer, one of Lin's co-authors and a professor of geosciences and international affairs, also at Princeton, told me. "So what does that say about a bad storm -- and the much worse risks we face -- years from now?"

Years from now, though, is in fact when these findings may be confirmed. Only a string of Sandys could verify such a study, which forecasts trends, not any specific hurricane which might define them. (That Sandy was a hybrid of hurricane and mid-latitude storm, formed under an unlikely confluence of climatic events, puts it somewhat outside the study's scope, too.) The temptation to view Sandy, or any major storm, as having been foretold by climate scientists, must be resisted.

Capturing the effect of climate change on such high-impact weather is, in any case, not about predicting storms, but assessing risk -- the likelihood that, as burning fossil fuels continue to change the Earth's climate, these storms could strike more frequently and more violently. The business of climate science is probability, and rightfully so, because once predictions come to pass, it's too late.

Lin appreciates why this may be difficult to grasp. "Most people think because we had Irene last year, and now Sandy this year, it's because of climate change," she said. "They think risk is increasing only when they see storms. And then after a few years, they forget."

According to recent research in climate psychology, cognitive deficiencies may be numbing us to the urgency of such intangible, global threats. Complex dynamics like those at play in climate change -- worsening gradually and posing consequences mostly in the distant future -- are not easily factored in to our present deliberations. As a result, we hold out for the singular maelstrom to sound the alarm, which, in this case, is like waiting for the straw man: Its occurrence wouldn't wholly reflect the phenomenon we're looking for, and so its absence shouldn't be seen as indicative that the models are wrong.

But climate psychologists also find that this simple-mindedness is not entirely our fault. It depends on how the questions are framed -- how a reader's mind is "primed" by certain events or exercises, before approaching the issue. And one way to frame climate change is to model the science. We may find that how climate scientists study the planet today, applying innovative techniques to quantify local risk, is more palpable than the diffuse issues we couldn't quite grasp.

Anthropogenic global warming is, admittedly, an abstract phenomenon. It does not create individual hurricanes, but instead, more diffusely, ratchets up their threat. As the atmosphere warms it retains more moisture which, carried to landfall, can then be hurled ashore in heavier wind and rain. And according to one accepted theory, the so-called "heat-engine" hypothesis first suggested by Kerry Emanuel, another of Lin's co-authors and a professor of atmospheric science at MIT, the warmest water in more than half a century could be amplifying storms' power source.

"That's a good starting point for understanding how climate change might affect future hurricanes," James Elsner, a professor of geography at Florida State University, told me. His 2008 study of ocean temperature and hurricane strength was consistent with this idea. "But just because stronger storms are getting stronger doesn't mean we're getting more storms."

The latest climate models appear to be reaching agreement on this issue. Christopher Landsea, the science and operations officer at the National Hurricane Center, in Miami, is a leading expert on hurricane forecasting. In an opinion piece for the National Oceanic Atmospheric Administration, he aggregates recent studies on hurricanes and global warming.

"Hurricanes have been depicted as the literal poster-child of the harmful impacts of global warming," he begins, before confessing that "manmade global warming has indeed caused hurricanes to be stronger today." But going forward, he's just not sure it will be significant. "Simply linking hurricanes to global warming is not sufficient," he writes. We need to know by how much, and in that respect, the science appears relatively unthreatening: By the end of the century, tropical storms and hurricanes may grow stronger as the Earth warms, but only slightly, and may occur just as frequently, or even less so. "[T]he overall impact of global warming on hurricanes is currently negligible and likely to remain quite tiny even a century from now," he concludes. Whatever a warming basin's brew cooks up, it isn't necessarily being felt.

While storm frequency and climatology generally speaking don't appear to be amplified by rising seas and warming waters, the effect on storm behavior at landfall is another question. To Landsea, it may be among the most important. "The biggest immediate worry I have," he concludes, "is with the huge population increases of vulnerable coastal communities...causing massive damage increases and, unfortunately, large losses of life."

It was for similar reasons that Emanuel, in 2006, first proposed conducting regional risk assessments by simulating hurricanes in various climate conditions. Historical records of hurricanes had yielded scarce data for many locales, and if the climate continued to change, they wouldn't offer much indication of future storms, either. By running simulations, he could render those future storms in physical models -- built from numerical values assigned to the different mechanisms at play -- to predict the likely effects of climate change on storms in a given region.

Now, this year's study of surge in New York City by Lin and her team is the culmination of these efforts. "The paper got a lot of response from people thinking it was very important and a new approach," Lin recalled. "It got a lot of interest. So I think this is the direction to go."

Greg Holland, director of the Earth System Laboratory at the National Center for Atmospheric Research, seems to agree. "Up until recently, the big focus has been on global climate change, like warming," he told me. "As far as I'm concerned, that's now past science. We're coming down to regional scale." The approach is advancing, he explained, as "we're able to take advantage of rapidly developing computing capacity." But, "there are also some clever techniques." Lin's study, he thought, combined both.

Using projections from the Intergovernmental Panel on Climate Change, future climatic conditions were set on a lattice, capturing the end of the 21st century: Earth is roughly three degrees Centigrade warmer, oceans are about three feet higher, and amid an atmosphere thickening with carbon dioxide, new wind and rain patterns are emerging. Then, one by one, 45,000 proto-hurricanes -- sudden disturbances to air pressure over the Atlantic -- were virtually plunked down.

"Most die, according to a kind of natural selection of the fittest," Emanuel told me, referring to the realistic interaction between storms and their environment in the simulations. "But some turn into bona fide hurricanes."

Monitoring those that churned toward New York City, the researchers witnessed surges more severe than any in the city's history. While offering no indication of when a particular storm might strike, the findings describe a clear trend: The most extreme storm surges will penetrate the city more frequently as the climate changes. Storms like Sandy, then, may be only part of a larger threat -- from the worst inundations, like those still plaguing New Yorkers weeks after Sandy -- even if storms in general don't much change.

Over the past century, the warming Atlantic Ocean has risen eight inches from thermal expansion and melting of mountain glaciers and ice sheets. And even as the rates pick up -- to between one and three feet every century -- coastline along the Atlantic is sinking, worrisome for many of the 3.7 million Americans already at risk from flooding. Suppose, as most climate scientists do, that each additional foot of sea-level rise, along an East Coast beach with average slope, swallows 100 feet horizontally of land. "Now think about a roughly similar effect with storm surge," Oppenheimer said. "The water drives farther inland than it would have otherwise." While the issue of storm intensity allows wiggle room for debate, that of sea-level rise does not: The 28-foot surge during Katrina in 2005, a mere category 3 hurricane, was the highest ever recorded in the U.S., costing over $100 billion in damages and about 2,000 lives.

The way Oppenheimer sees it, "the risk is very high already. We're not very good at dealing with it. And the risk will get higher continuously until we reduce emissions, and slow global warming. So we better get cracking."

The cost of waiting -- as high as $50 billion for Sandy, and potentially $1.8 trillion annually for high-impact weather in the coming decades -- has not been missed by the Mayor's office, which has since approached Lin and her team for help redesigning New York City's Flood Insurance Rate Map. Their method, though, can and will be applied elsewhere -- to Tampa, China, the Persian Gulf. Until then, according to Oppenheimer, "We need to stay abreast of the scientific developments. And start defending ourselves."

Unless, of course, we're willing to assume the risk.