Son of Sandy
Are we going to just sit back and wait for the next superstorm, or do something about it?
Steven Cherry: Hi, this is Steven Cherry for IEEE Spectrum’s “Techwise Conversations.”
Hurricane Sandy devastated New York and New Jersey. There’s no stopping a giant coastal storm when it decides to turn inland. But much of the damage was, in the opinion of my guest today, preventable.
The flooding of New York City’s commuter tunnels and much of the subway system, for example, was due not so much to the heavy rains as the storm surges that were, on the night of Monday, October 29, amplified by a nearly full moon and high tide.
New York City is a collection of islands separated by rivers. A series of three levees, placed strategically at chokepoints where the Atlantic Ocean and the Long Island Sound meet the city’s inner rivers and bays, would have protected Manhattan. Remarkably, that’s not just the opinion of today’s guest. Seven years ago, in a September 2005 op-ed piece in The New York Times, he presciently wrote:
In the future, relatively modest storms riding on an ever-increasing sea level will do as much damage as rare, once-in-a-century storms do now. New York City, New Jersey, and Long Island, particularly the low-lying south shore, are at great risk.
Also in 2005, he coauthored a report entitled “Bracing for Super Floyd” that predicted a storm—and resulting disaster—that matched the one Sandy wrought. And it advocated for that levee system similar to the ones that protect the Netherlands and its incredibly valuable, and vulnerable, low-lying coastal lands.
Malcolm J. Bowman is a professor of physical oceanography at Stony Brook University’s Marine Sciences Research Center and heads the Storm Surge Research Group there. He’s originally from New Zealand, but his Ph.D., in engineering physics, is from the University of Saskatchewan, in Canada. He joins us by phone.
Malcolm, welcome to the podcast.
Malcolm J. Bowman: Well, thank you.
Steven Cherry: Malcolm, let’s first understand the problem. The phrase “storm surge” is a little misleading; it makes us think of a tsunami, maybe a smaller tsunami: a single giant wave. But that’s not what it is at all. What is a storm surge?
Malcolm J. Bowman: I’m glad you brought that up, because there is a lot of misunderstanding what a storm surge is. Basically, a storm surge are ocean currents that are pushed by the strength of the winds during a nor’easter, a winter nor’easter, or a hurricane. As we all know, storms create large waves, and they can crash against the shore and do a tremendous amount of damage. But that’s part of the storm surge, those waves that wash up against the beach. But in addition, the winds actually blow the surface waters of the ocean around, and, depending on the direction of the wind, their strength, their duration, that water, if it’s pushed up against the coast, it’s held there by these winds, and that’s what we call “storm surge.” But it’s not a tsunami or a wall of water—although the sea can raise very rapidly. It’s really like an extra high tide. And so the sea level starts coming up over a period of hours, and then it will tend to go down again as the storm passes. So if it’s an additional high tide, if you like, it’s a surge in that respect, but it’s not a breaking wave.
Steven Cherry: Part of your work at the Stony Brook Storm Surge Research Group is computer modeling of possible storms for the New York region. How closely did Sandy match any of your models?
Malcolm J. Bowman: We have what we call a short-term forecasting system that looks two to three days into the future. It starts off with a numerical weather forecasting program we have for the East Coast of the United States, and then, in turn, gets its input information, its boundary conditions, initial conditions, from the National Weather Service. And, running a global weather model, every night at midnight, the atmospheric weather model turns on, and it takes two hours to make a three-day forecast.
So by 3:00 a.m., underneath that weather model, we have an ocean model, and the winds and the low-pressure systems at sea level from the atmosphere start moving water around. The ocean model also has the tides; it correctly reproduces the coastal tides on the East Coast, and we run that model for three days, and that’s finished about 5:00 a.m. each morning. We then compare our predictions at a number of coastal tidal stations that are maintained by the U.S. government, the National Oceanic and Atmospheric Administration, and the U.S. Geological Survey. So they have real-time data-collecting systems.
We compare our predictions, see the results, and post it on the Web. So we have a daily refreshed prediction system. How well did it do? It did pretty well, actually. We predicted the height of the surge at the Battery—the southern tip of Manhattan—within about nine inches. So that’s pretty good.
Steven Cherry: Tell us about the levee system you proposed, and maybe you can go ahead and tell us about the levee system the Netherlands already has.
Malcolm J. Bowman: Well, the Europeans have long been protecting their cities against storm surges. It’s not just the Netherlands. The countries that have active systems—United Kingdom; London, England; the Netherlands; Venice, Italy; and St. Petersburg, Russia. I can just briefly mention what they are. The most interesting one to me is actually St. Petersburg. In 1953, there was a massive storm in the North Sea, which surrounds several thousand people in southeast England and the Netherlands. And so following that, the respective governments realized they had to do something, something bold, something imaginative, something more permanent. So the Thames River barrier was constructed to protect the city of London, and that was open 29 years later in 1982, and it’s used often. It’s used several times a year, actually, to stop surges from the North Sea propagating up the Thames River and flooding the city of London.
The Dutch have a bigger problem in that half of their country is below sea level. Amsterdam is about 2 meters below sea level. Rotterdam, the largest seaport in the world, is up to 3 meters below sea level. So they have constructed over the last 50 years a system called the “Delta Project,” and that’s a combination of strengthening their coastlines against future storm surges. They do it in two ways: by enhancing the natural sand dunes, by pumping sand from offshore, pumping that onto land, and strengthening their sand dunes up to 15 meters high or more. They’ve protected half the country that way, and the other half of the coastline is protected by a system of levees, seawalls, floodgates—it’s a remarkable engineering feat.
But they cannot mess around. It’s a question of national security. It’s a question of national survival. It’s a densely populated country. It’s very flat. There’s nowhere to run. There’s nowhere to hide. So they have faced up to this challenge, and they protect their cities against what’s known as a 1-in-1000-year storm. That is a storm that is so severe that it is likely to happen only once every thousand years. Now, of course, the statistics of such a thousand-year storm don’t exist. But they even think about 10 000-year storms. That’s sort of unimaginable this side of the Atlantic, where New York City building code requires construction of postal facilities to be protected against the 1-in-100-year storm.
Steven Cherry: You said the St. Petersburg system is particularly interesting to New Yorkers?
Malcolm J. Bowman: St. Petersburg is very interesting in that it is also built on a river delta like New York Harbor: the River Neva. It’s called the “Venice of the North.” It’s very susceptible to flooding at the eastern end of the Baltic Sea. The Soviets started building an elevated ring road, if you like, like a beltway, around the city. The Soviet Union collapsed. It stopped. And now it’s being finished with European Union money and was opened last year. It’s an elevated six-lane highway and extends out into the harbor, where they have what you might call “levees,” but it’s more than just a levee. And then there are the huge gates that allow the ships in and out.
So that’s the model I think would work best for New York Harbor, and where would these be? We’d need two gateway systems—storm barriers. One would spread from the tip of Sandy Hook, New Jersey, the northernmost tip of New Jersey, across to a place called Far Rockaway on Long Island, and this would be a 5-mile gap. It would be a multipurpose system. It would be an elevated interstate highway, bypass for New York City. There would be a storm barrier, a light-rail system for Kennedy Airport, and that would have to be augmented with nourished sand dunes at both ends—both the New Jersey and Long Island ends. It’s about five miles across, but it’s very shallow, only 20 to 25 feet deep, and apart from the main navigation channels in and out of the harbor, it would be relatively easy to construct. It would be elevated. It would be open underneath, like a bridge, where normal tidal flow flushes out the harbor. The Hudson River can discharge to the ocean.
There would need to be a second one in the upper East River, which is the second connection of New York Harbor to the ocean through Long Island Sound. Long Island Sound experiences quite severe storm surges, so you’d need a second one there near the Throgs Neck Bridge. It’s relatively narrow. It’s only 1 mile wide.
Steven Cherry: And I guess there would be a third one at the southwestern end of Staten Island?
Malcolm J. Bowman: The earlier research, we had a barrier proposed from the Verrazano-Narrows, and then a third one near Staten Island, but it turns out that’s a very difficult engineering assignment. The Narrows are very deep. It’s over 100 feet deep. The currents are very strong. It’s got a number of problems, and it does not protect the outer boroughs of New York City: Brooklyn, Queens. It does not protect the coastlines. It does not protect Kennedy Airport. So the favored solution now is this outer-harbor gateway that I just mentioned, from Sandy Hook to Far Rockaway, and then that would not need a third one behind Staten Island, because that would already be completely enclosed.
If this system I’ve just described were in place for Sandy, there would’ve been no flooding within the perimeter of that protection.
Steven Cherry: I’m wondering if the existence of the barriers would make it worse for other areas behind the barriers, since there’s sort of nowhere else for the water to go.
Malcolm J. Bowman: Well, that’s a question people always ask me, and the answer is, very little, because the surge, uh, if it’s closed off, a surge is not going to become—it’s going to spread itself over hundreds of miles of the East Coast. It’s not just concentrated in the corner between New Jersey and New York. So it might come up an extra 6 inches.
Steven Cherry: People have been building levees, I guess, since the Egyptians, but they didn’t have sensors and GPS and satellite photography and computerized control systems. Are levees very different from 50 or 100 years ago, or is it still mostly mechanical and civil engineering?
Malcolm J. Bowman: I think if you look at the catastrophe in New Orleans with Katrina, those levees there were grossly inadequate: They were vertical walls without proper foundation. They basically fell over. If you look at the Dutch levees, they are more like elevated superhighways. They are wide. They are wider than they are high. They are basically immovable. They are built on a firm foundation. They are built high enough. Nothing will move those levees. So if you look at some of the engineering practices of the Dutch—and they’re really sophisticated; they’re not just sitting on mud—they have proper foundations underneath, massive mats that they lay out. It’s spectacular engineering.
So the science certainly has developed. If you look at the big swinging gates that protect the port of Rotterdam—I call them “saloon door gates”—it’s all computer controlled in the sense that the computers actually make a decision about when to close the gates. If the storm brewing in the North Sea is severe enough, should they close the gates, which would cause a disruption to commerce, or not? And the reason they have this whole thing automated is because a lot of the people that work at this facility have their own families living inside this ring of protection, so they try and make it more rational, more analytical, and the decision probably comes from the prime minister’s office, but it’s taken out of the hands of the local employees.
Steven Cherry: Getting back to the computer modeling, I gather you think there are worse storms in New York’s future, in part because of climate change. The New York governor, Andrew Cuomo, said right after the storm, with Irene and Sandy back-to-back in back-to-back years, that this was the new normal. I guess your view agrees with him, that this is becoming the new normal.
Malcolm J. Bowman: Well, it’s an interesting statement from a political leader. As scientists and as engineers, we might say, “Well, our statistical sample is far too small.” And that’s, of course, true for hurricanes. We study these historical [unintelligible] in my research group. But there’s only a handful, over the last 20, 30, 50 years, and not that many where the data records are sufficiently detailed to do much analytical work with them. But I can say this: Especially with global change, we can expect not necessarily more hurricanes, but they are expected to be more intense. The weather extremes, both hot and cold in summer and winter, will be more intense. You can say Irene and Sandy are consistent with climate change, like we’re building up a case and these are exhibits A and B. I cannot say definitively Sandy is a consequence of climate change, but it certainly is consistent.
Steven Cherry: It sounds to me that besides the frequency increasing to maybe a 100-year storm becoming a 20-year storm, the consequence of say, Sandy, 40 or 50 years from now, might be much more severe. That is to say the same storm would have caused much worse damage. In fact, in one of your diagrams I saw, the lower Manhattan Island that was flooded by Sandy would be cut off from the rest of Manhattan.
Malcolm J. Bowman: That’s right. If you look at the areas that were flooded during Sandy, three weeks ago now, those were areas that were reclaimed by European civilization. There was water up to Water Street. Guess what? These were the low-lying areas that have been reclaimed with a minimal amount of work, so they are very vulnerable. Climatologists are expecting sea level to rise between 3 and 6 feet, 1 to 2 meters, by the end of the century, depending on how quickly the ice sheet on Greenland melts. That’s huge. As you said, a relatively weak storm will do the same flooding we saw with Sandy because it’s starting out at a higher base level. That doesn’t mean to say the storms will get less intense, but if they get more intense, then it’s compounding.
Steven Cherry: My producer tells me your middle name is, Noah. Noah built his ark on his own, but you can’t build your surge barriers without some help. I wish you luck in getting everyone to take this seriously, and thank you for joining us today.
Malcolm J. Bowman: You’re welcome.
Steven Cherry: We’ve been speaking with Malcolm J. Bowman about the engineering measures that might be taken to protect cities like New York from massive storm surges.
For IEEE Spectrum’s “Techwise Conversations,” I’m Steven Cherry.
Announcer: “Techwise Conversations” is sponsored by National Instruments.
NOTE: Transcripts are created for the convenience of our readers and listeners and may not perfectly match their associated interviews and narratives. The authoritative record of IEEE Spectrum’s audio programming is the audio version.