The India-Pakistan doomsday scenario

The confounding circumstances of war, experienced by both sides, won't likely help, adds Kimball at the Arms Control Association. "The concern here has to do with the fog of war effect," he says. "The response times that the governments have will be relatively short; and in the background is the possibility that nuclear weapons may have been loaded, deployed on short- and medium-range delivery systems."

Still, Pakistan remains the more likely nuclear actor. Unlike India, Pakistan has never sworn off the possibility of a first strike. Its standing army is weaker than India's, so it relies more heavily on nuclear power. In fact, Pakistani officials often argue that nuclear weapons are the country's most viable defense. As Munir Akram, Pakistan's new ambassador to the United Nations, said Thursday, "India should not have the license to kill with conventional weapons while Pakistan's hands are tied regarding other means to defend itself."

There are also signs that high-ranking members of the Pakistani military favor an even more aggressive nuclear stance. For example, the secretary to ousted leader Benazir Bhutto, Brig. Gen. Amanullah, as quoted in the March issue of the Atlantic, said that "We should fire at them and take out a few of their cities -- Delhi, Bombay, Calcutta." Later in the article, he added, "Before I die, I hope I should see it."

What might the effects be, if Amanullah's death wish comes true? This week's American intelligence report is not impossible to believe, but, experts argue, a more likely nuclear scenario was mapped out in 1998 by Ramana, who is also a founding member of the Indian Coalition for Nuclear Disarmament and Peace.

Ramana's findings gain high marks from Makhijani, who holds a Ph.D. in physics from Berkeley, as well as from other nuclear experts, who defend the science behind Ramana's paper. And the story Ramana tells, while less grim than others, is morbid, depressing, dreadful. His report, which he wrote while studying at MIT's Security Studies Program, is titled "Bombing Bombay: Effects of Nuclear Weapons and a Case Study of a Hypothetical Explosion."

It assumes that a Hiroshima-size bomb explodes, as that bomb did, about 600 meters above the city, which, according to 1991 census figures, has a population of more than 9 million people. "If it is a civilian population that is being targeted, one might expect that it would be an airburst," Ramana says, noting that the Hiroshima bomb detonated in the air. "Surface bursts are preferred when attacking a hardened/buried target such as a missile silo or a command/control center," he says.

And ultimately, "depending on the population density in the part of the city that is targeted," Ramana concluded in his report, "the numbered deaths [caused by such a bomb] would range between 160,000 to 860,000." The use of nuclear weapons over any densely populated city in South Asia would result in similar casualty figures, according to Ramana.

Ramana starts by pointing out that the "prompt effects" include an initial flash of heat and light like "a thousand suns." This rush of energy, he argues, is hot enough to make sand "explode like popcorn" and to ignite all combustible materials within one to two miles around the point of explosion. Radiation would also immediately be released, leading to symptoms like nausea and bloody diarrhea for those who survive the initial explosion.

Almost immediately, a "blast wave" would arrive, hurricane-force winds of more than 65 mph that would blow for a distance of more than two miles. Everything not made of concrete within a circular radius of about three-quarters of a mile would collapse, including human bodies, while "missiles" ("physical objects propelled outward by the explosion") flew threw the air, killing others.

The delayed effects would start occurring within minutes. First, a massive "superfire" would form. A collection of individual fires, the flames would likely cover about two kilometers. Due to the fire's size, the fire zone would act as a colossal pump: It would draw in air from the surrounding areas; it would create winds with velocities as high as 30-50 mph. And the fire would be nearly impossible to put out, not just because temperatures would reach hundreds of degrees, but also because the shock wave would have destroyed water mains while collapsed buildings blocked streets.

Ramana's estimate of 160,000 to 860,000 dead, says Makhijani, is actually conservative.

"He's only talking about one nuclear weapon," Makhijani says. "If more than one were used, that kind of effect would reinforce itself. The dislocation of the fires would be worse. There'd be no exit for injured people, especially in a place like Bombay, which is an island, and which doesn't have sufficient medical facilities."

The only hope, Makhijani says, is for both parties to disengage. And as Rumsfeld prepares to depart for the region, Makhijani fears that discussion won't be enough. Perhaps, he suggests, the United Nations should create an international force that could rein in the Pakistani militants. Or perhaps the U.S. could help Musharraf's political standing by agreeing to lower trade barriers on textiles, which protect U.S. companies while making it harder for Pakistani companies to sell clothing in the U.S.

Other players and other solutions might also arise unexpectedly, says Smith at CDI.

"A wild card in the whole situation is China," he says. "Remember that China seized part of Kashmir from India (and Pakistan 'ceded' a section) in the 1960s. Also, China is a source of missile technology for Pakistan and might lose world standing if its client suffered a devastating blow. Would the Chinese do something militarily to divert India's focus from Pakistan, perhaps under the guise of acting to prevent escalation into a nuclear conflict whose radiation spillover would affect southern China?"

China's ties to Pakistan run deep, says author and international affairs columnist Eric Margolis. "Pakistan is China's most important ally, and vice versa. China and India have been in a long nuclear and conventional arms race for the last decade, which we in the West have completely ignored. China may not let Pakistan be crushed by India. And China feels that India is being built up by the United States as a counterweight to China. They [China and India] have been saber-rattling at each other, and it's very possible that there could be some form of Chinese military intervention."

All the questions, guesses and speculation may ultimately prove futile. War is not rational. No expert, no matter how experienced, can predict how or if the conflict will escalate. More to the point, "There's a certain surrealism to all of this because people frankly don't know what they're talking about," says Cordesman at CSIS.

So while the attempt to plot out a path to destruction may help the U.S. military prepare for a possible humanitarian response, the gaming out of a conflict is simply an irrational attempt to insert logic into an illogical world. The threat of nuclear war is real, but the goal should not be to explain how it can happen, Cordesman says, but rather to stop the process. "You don't prevent murder by committing colossal suicide," he says. "The apparent logic isn't real."

• View blog reactions