Discover, January 31, 1992
Even before the Kuwaiti oil fields began burning, researchers warned that we humans, busy with war, were about to play dice with the biosphere. Rising smoke from a devastated Kuwait, they said, might cause drastic changes in the planet’s weather.
On January 22, when the first oil wells were actually set on fire, Carl Sagan appeared on ABC’s Nightline. “We think the net effects will be very similar to the explosion of the Indonesian volcano Tambora in 1815, which resulted in the year 1816 being known as the year without a summer,” he said.
“There were massive agricultural failures in North America and in western Europe, and very serious human suffering, and in some cases starvation. Especially for South Asia, that seems to be in the cards, and perhaps for a significant fraction of the Northern Hemisphere as well.”
The scenario painted by Sagan and two fellow planetary scientists, Paul Crutzen of the Max Planck Institute for Chemistry in Germany and Richard Turco of UCLA, was the most frightening picture offered, but it was not the only one. Others ran the gamut, with predictions at the opposite end of the spectrum saying that the smoke’s effects would be marginal at worst. Now, with the benefit of nearly a year’s worth of hindsight, researchers are realizing that in one way or another, everyone was wrong.
Such a seemingly off-the-mark scientific response is particularly baffling when you consider that forecasters have had time to prepare for a soot-filled sky. Climate studies have been all the rage for the past decade, and the computers that handle raw data now ripple with number-crunching power. But even with these impressive tools, today’s climate modelers obviously still have their limits. Now that the smoke is clearing, many people may wonder how reliable such apocalyptic weathermen are, and if anyone should pay attention when they warn us of such other potential disasters as global warming.
A lot of the confusion last January came from the way scientists have to talk on television. All told, Sagan spoke for about two minutes on Nightline, and it’s difficult to suggest the complexity and ambiguity of computer models in that amount of time. “A conversation on 60 Minutes isn’t going to be scientifically profound,” says Turco. “You know it’s a sound-bite environment, and you do the best you can. But it seems as if all the contingencies get dropped out, and you get saddled with one prediction, and it’s hard to back out.”
Sagan, Turco, and Crutzen all based their predictions on the concept of nuclear winter, which they had first formulated in 1982. The scenario they originally modeled started with a massive nuclear exchange between the Soviet Union and the United States. They came up with estimates of how much would be set on fire (everything from asphalt roofing to trees around military bases), how much smoke would be generated, and how hot the fires would be.
The researchers realized that some soot would get attached to water droplets and fall back to the ground as dirty rain. But believing the old saw that oil and water don’t mix, they thought that 75 to 90 percent of the smoke would remain in the atmosphere. Meanwhile, they said, fire storms all over the planet would heat the air and create huge updrafts. As sunlight hit the rising black smoke, the plumes would heat up and rise even more, in a process called self-lofting.
If the smoke rose eight to ten miles, it would enter the stratosphere, high above any rain-producing clouds; the smoke would then linger in the stratosphere for months, spreading around the planet. Under its shadow, temperatures in the interior of the continents could cool as much as 70 degrees. Weather patterns would be mangled, whole ecosystems might die, and “nuclear winter” would ensue, possibly killing off far more people than the nuclear blasts or radioactivity.
From the start of the Gulf War, it was clear that the amount of smoke from the Kuwaiti oil fires would be far less than that of a full-scale nuclear war–less than 1 percent. “Comparisons with nuclear winter were irrelevant,” Turco says. But the methods he used to develop that scenario weren’t.
Using scaled-down estimates of how much oil might be burned, how much smoke this would make, and how much self-lofting would be produced, Turco found that a significant amount of smoke might rise into the upper troposphere, just below the stratosphere. It would eventually get cleansed from the air, but since it would take a long time to put out the fires–some early estimates were as long as three years–the purged smoke would be replaced. The Middle East could cool as much as 5 degrees on average, Turco concluded, and a thin layer of smoke would spread over the whole Northern Hemisphere. “You wouldn’t ever notice it in the sky,” he says.
As Sagan pointed out, the main concern was the monsoon. Every summer the Tibetan plateau heats up, causing the dry air above central Asia to rise. The air that moves in to replace it comes off the ocean, becoming heavy with water as it travels. When it hits land, it also heats up, dumping huge amounts of rain as it rises. Asian farmers from India to Thailand have to plan their crops accordingly. But the smoke from Kuwait, cooling the Tibetan plateau by even a couple of degrees, might have prevented the air over Tibet from rising, shutting off the rains and leaving hundreds of millions of people with nothing to harvest.
Other, less frightening studies came out. Richard Small, an atmospheric scientist at Pacific-Sierra Research, a California think tank, used lower estimates of smoke production and predicted that most of the smoke would go up only two miles. Most in fact did, although some rose as high as four miles. British and German teams, using more complex simulations to predict how the smoke would spread, found that it would cause local cooling but not affect the monsoon–a scenario close to what actually happened. In Kuwait temperatures were sometimes as much as 20 degrees below normal, and 150 miles away Bahrain recorded its coldest May on record. But the monsoons did not disappear.
Of course that’s small comfort for the people of Kuwait. One of the byproducts of burning oil is acid rain and acid deposition (when acidic particles drop out of the sky without the help of water). The latter has been severe in Kuwait. And for those who have respiratory problems or who have been weakened by other diseases, simply breathing the smoke-clogged air can be lethal. Despite the EPA’s cheerful prognosis in April that toxins from the smoke were “not at levels of concern,” public health experts attending a conference at Harvard in August estimated that the air pollution would kill 1,000 Kuwaitis over the next year.
Meanwhile the Gulf is still reeling from a spill of 250 million gallons of oil–more than 20 times larger than the Exxon Valdez spill and twice as large as the previous world record. Oil has soaked 35 miles of Saudi Arabia’s coastline, and because the Gulf has a shallow, sluggish circulation system, it will be many years before the waters clear. Even the unignited oil from sabotaged wells is harming the environment. It has formed vast pools and is sinking into the sand; it may damage plant and animal populations in the desert, and it could leach into the Gulf.
Bad as they are, these effects are not on the scale that Sagan, Crutzen, and Turco were projecting for the smoke’s effects. But the fact that the most dire predictions were wrong doesn’t make the other simulations right. Researchers who have been watching the smoke over the past few months have found far more things than were dreamed of in any of their climate models.
Take the color of the smoke, for example–a color it had seemed safe to assume would be black. At some wells the smoke was white, possibly laced with water vapor or salt. As a result, the soot was about half as dark as everyone expected, which cut down on self-lofting. The oil fires’ chemistry was also surprising: few hydrocarbons and less carbon monoxide were coming out of the fires than expected. Strangest of all was the discovery that the smoke particles actually mixed with water instead of repelling it.
“Even if you have a particle that’s ninety-nine percent one compound, if the other one percent is at the surface, that determines how it acts with water,” explains Randy Cofer of NASA’s Langley Research Center. This odd fact–still unexplained–probably made the smoke wash out of the air faster than researchers thought it would.
The fires themselves have also been put out faster than expected. Never having dealt with blazes of this proportion, the fire teams tried many different techniques before they settled on a few reliable ones. “In the beginning the techniques took a few weeks or months,” explains Henry Kendall, an MIT physicist and consultant to the firemen, “but they’ve just learned to do things better.” On November 6 the last of the 650 fires was snuffed.
Climate modelers, meanwhile, are just beginning to crank in the new data. In one after-the-fact simulation–done this summer by Thomas Sullivan, a meteorologist at Lawrence Livermore Laboratory–the smoke rising from Kuwait had the opposite effect on the monsoon from the one Sagan and company had expected. Sullivan ran a program originally developed for tracking radioactive clouds. He led actual weather readings into a computer and then tracked the spread of simulated smoke particles. In his model, smoke spread to parts of India where in reality there had been heavy rains. Even more disturbing, his simulated smoke showed up in Bangladesh in late April, which is when cyclones actually hit the country, killing 140,000 people. Could the real smoke from the fires somehow have triggered those events?
In both cases the strange attraction of the smoke to water may have been at work. In India it may have forced moisture to fall before the waterlogged air had a chance to reach regions farther north, where rainfall was 60 percent below normal. In Bangladesh the smoke wouldn’t have triggered the violent storms, but it could have aggravated them. In fact, what made the cyclones so deadly was the intensity of the accompanying rains, which were the most intense in decades. “Of course we’re talking about a simulation being run on the other side of the world from these events,” says Sullivan, “but still, it bears more investigation.”
With the eruption of Mount Pinatubo in June–an eruption that sent tons of volcanic “smoke” into the stratosphere–and the arrival of El Niño, a periodic atmospheric disturbance that warms the Pacific, the long-term effects of the Kuwaiti fires may never be completely known. A number of researchers, however, are already saying that climate predictions derived from the nuclear winter scenario will have to be overhauled, although the basic theory won’t change. In particular the figures Turco and his colleagues used for things like smoke production and self-lofting will have to be revised. They were originally based on small experimental fires, and now the Persian Gulf War has produced, as Small puts it, “a giant laboratory for the future.”
Climate modeling, like any science, is far from infallible, but as Turco points out, “it’s a self-correcting process.” Simulations of ozone destruction in the seventies were wildly inaccurate, but the basic principle held true, and now researchers have a good grasp on how it happens. Meanwhile, Kuwait’s oil still burns, only now the fire is back inside the world’s engines, and the emissions are warming the planet instead of cooling it. That means that the world still needs climate modelers–and that we had better listen to what they say, preferably for longer than two minutes.
Copyright 1992 Discover Magazine. Reprinted with permission.