Science

The Century of Disasters

Meltdowns. Floods. Tornadoes. Oil spills. Grid crashes. Why more and more things seem to be going wrong, and what we can do about it.

A half-mile wide twister tore through Joplin, Mo., on Sunday, killing nearly 100. The tornado was one of 68 reported across seven states this weekend. Unfortunately, this century will be a time when natural disasters and failures of human design go hand-in-hand. As Joel Achenbach explained earlier this month in the article reprinted below, we’ve engineered a planet that works well but is susceptible to catastrophic failures. 

This will be the century of disasters.

In the same way that the 20th century was the century of world wars, genocide, and grinding ideological conflict, the 21st will be the century of natural disasters and technological crises and unholy combinations of the two. It’ll be the century when the things that we count on to go right will, for whatever reason, go wrong.

Late last month, as the Mississippi River rose in what is destined to be the worst flood in decades, and as the residents of Alabama and other states rummaged through the debris of a historic tornado outbreak, physicists at a meeting in Anaheim, Calif., had a discussion about the dangers posed by the sun.

Solar flares, scientists believe, are a disaster waiting to happen. Thus one of the sessions at the American Physical Society’s annual meeting was devoted to discussing the hazard of electromagnetic pulses (EMPs) caused by solar flares or terrorist attacks. Such pulses could fry transformers and knock out the electrical grid over much of the nation. Last year the Oak Ridge National Laboratory released a study saying the damage might take years to fix and cost trillions of dollars.

But maybe even that’s not the disaster people should be worrying about. Maybe they should worry instead about the ARkStorm. That’s the name the U.S. Geological Survey’s Multihazards Demonstration Project gave to a hypothetical storm that would essentially turn much of California’s Central Valley into a bathtub. It has happened before, in 1861-62, when it rained for 45 straight days. The USGS explains: “The ARkStorm draws heat and moisture from the tropical Pacific, forming a series of Atmospheric Rivers (ARs) that approach the ferocity of hurricanes and then slam into the U.S. West Coast over several weeks.” The result, the USGS determined, could be a flood that would cost $725 billion in direct property losses and economic impact.

While pondering this, don’t forget the Cascadia subduction zone. That’s the plate boundary off the coast of the Pacific Northwest, one that could generate a tsunami much like the one that devastated Japan in March. The Cascadia subduction zone runs from Vancouver Island to northern California, and last ruptured in a major tsunami-spawning earthquake on January 26, 1700. It could break at any moment, with catastrophic consequences.

All of these things have the common feature of low probability and high consequence. They’re “black swan” events. They’re unpredictable in any practical sense. They’re also things that ordinary people probably should not worry about on a daily basis. You can’t fear the sun. You can’t worry that a rock will fall out of the sky and smash the earth, or that the ground will open up and swallow you like a vitamin. A key element of maintaining one’s sanity is knowing how to ignore risks that are highly improbable at any given point in time.

And yet in the coming century, these or other black swans will seem to occur with surprising frequency. There are several reasons for this. We have chosen to engineer the planet. We have built vast networks of technology. We have created systems that, in general, work very well, but are still vulnerable to catastrophic failures. It is harder and harder for any one person, institution, or agency to perceive all the interconnected elements of the technological society. Failures can cascade. There are unseen weak points in the network. Small failures can have broad consequences.

Most importantly: We have more people, and more stuff, standing in the way of calamity. We’re not suddenly having more earthquakes, but there are now 7 billion of us, a majority living in cities. In 1800, only Beijing could count a million inhabitants, but at last count there were 381 cities with at least 1 million people. Many are “megacities” in seismically hazardous places—Mexico City, Caracas, Tehran, and Kathmandu being among those with a lethal combination of weak infrastructure (unreinforced masonry buildings) and a shaky foundation.

Natural disasters will increasingly be accompanied by technological crises—and the other way around. In March, the Japan earthquake triggered the Fukushima Daiichi nuclear power plant meltdown. Last year, a technological failure on the Deepwater Horizon drilling rig in the Gulf of Mexico led to the environmental crisis of the oil spill. (I chronicle the Deepwater Horizon blowout and the ensuing crisis management in a new book: A Hole at the Bottom of the Sea: The Race to Kill the BP Oil Gusher.)

In both the Deepwater Horizon and Fukushima disasters, the safety systems weren’t nearly as robust as the industries believed. In these technological accidents, there are hidden pathways for the gremlins to infiltrate the operation. In the case of Deepwater Horizon, a series of decisions by BP and its contractors led to a loss of well control—the initial blowout. The massive blowout preventer on the sea floor was equipped with a pair of pinchers known as blind shear rams. They were supposed to cut the drillpipe and shear the well. The forensic investigation indicated that the initial eruption of gas buckled the pipe and prevented the blind shear rams from getting a clean bite on it. So the “backup” plan—cut the pipe—was effectively eliminated in the initial event, the loss of well control.

Fukushima also had a backup plan that wasn’t far enough back. The nuclear power plant had backup generators in case the grid went down. But the generators were on low ground, and were blasted by the tsunami. Without electricity, the power company had no way to cool the nuclear fuel rods. In a sense, it was a very simple problem: a power outage. Some modern reactors coming online have passive cooling systems for backups—they rely on gravity and evaporation to circulate the cooling water.

Charles Perrow, author of Normal Accidents, told me that computer infrastructure is a disaster in the making. “Watch out for failures in cloud computing,” he said by email. “They will have consequences for medical monitoring systems and much else.”

Technology also mitigates disasters, of course. Pandemics remain a threat, but modern medicine can help us stay a step ahead of evolving microbes. Satellites and computer models helped meteorologists anticipate the deadly storms of April 27 and warn people to find cover in advance of the twisters. Better building codes save lives in earthquakes. Chile, which has strict building codes, was hit with a powerful earthquake last year but suffered only a fraction of the fatalities and damage that impoverished Haiti endured just weeks earlier.

The current Mississippi flood is an example of technology at work for better and for worse. As I write, the Army Corps of Engineers is poised to open the Morganza spillway and flood much of the Atchafalaya basin. That’s not a “disaster” but a solution of sorts, since the alternative is the flooding of cities downstream and possible levee failure. Of course, the levees might still fail. We’ll see. But this is how the system is supposed to work.

On the other hand, the broader drainage system of the Mississippi River watershed is set up in a way that it makes floods more likely. The cornfields in parts of the upper Midwest, for example, have been “tiled” with pipes that carry excess rainwater rapidly to the rip-rapped streams and on down to rivers lined with levees. We gave up natural drainage decades ago. The Mississippi is like a catheter at this point. Had nature remained in charge, the river would have mitigated much of its downstream flooding by spreading into natural floodplains further upriver (and the main channel would have long ago switched to the Atchafalaya river basin—see John McPhee’s The Control of Nature—and New Orleans would no longer be a riverfront city).

One wild card for how disastrous this century will become is climate change. There’s been a robust debate on the blogs about whether the recent weather events (tornadoes, floods) can be attributed to climate change. It’s a briar patch of an issue and I’ll exercise my right to skip past it for the most part. But I think it’s clear that climate change will exacerbate natural disasters in general in coming years, and introduce a new element of risk and uncertainty into a future in which we have plenty of risks and uncertainties already. This, we don’t need.

And by the way: Any discussion of “geoengineering” as a solution to climate change needs to be examined with the understanding that engineering systems can and will fail. You don’t want to bet the future of the planet on an elaborate technological fix in which everything has to work perfectly. If failure isn’t an option, maybe you shouldn’t try it to begin with.

So if we can’t engineer our way out of our engineered disasters, and if natural disasters are going to keep pummeling us as they have since the dawn of time—what’s our strategy? Other than, you know, despair?

Well, that’s always worked for me, but here are a few more practical thoughts to throw in the mix. First, we might want to try some regulation by people with no skin in the game. That might mean, for example, government regulators who make as much money as the people they’re regulating. Or it could even mean a private-sector regulatory apparatus that polices the industry, cracking down on rogue operators. The point is, we don’t want every risky decision made by people with pecuniary interests.

Second, we need to keep things in perspective: The apparent onslaught of disasters doesn’t portend the end of the world. Beware disaster hysteria in the news media. The serial disasters of the 21st century will be, to some extent, a matter of perception. It’ll feel like we’re bouncing from disaster to disaster in part because of the shrinking of the world and the ubiquity of communications technology. Anderson Cooper and Sanjay Gupta are always in a disaster zone somewhere, demanding to know why the cavalry hasn’t showed up. 

Third, we should think in terms of how we can boost our societal “resilience.” This is the buzzword in the disaster-preparedness industry. Think of what you would do, and what your community would do, after a disaster. You can’t always dodge the disaster, but perhaps you can still figure out how to recover quickly. How would we communicate if we got flared by the sun and the grid went down over two-thirds of the country? How would we even know what was going on? Maybe we need to have the occasional “18th-century weekend” to see how people might get through a couple of days without the grid, the cell towers, the cable TV, the iTunes downloads—the full Hobbesian nightmare.

And make an emergency plan. Buy some batteries and jugs of water just for starters. Figure out how the things around you work. Learn about your community infrastructure. Read about science, technology, and engineering, and don’t worry if you don’t understand all the jargon. And then, having done that, go on about your lives, pursuing happiness on a planet that, though sometimes dangerous, is by far the best one we’ve got.