Mr. Plow eagerly awaits nuclear war

Step off, Al Gore. I, with a little help from an eager group of atmospheric scientists, have found a quick fix to global warming. All we need is a handful of nuclear weapons! They can even be small ones!
Youre probably thinking that the heat is messing with my mind. A slew of studies released in the past few months, though, has confirmed that using nuclear weapons could significantly -- perhaps even catastrophically - cool the planet.
This phenomenon was first studied towards the end of the Cold War, in the early 1980s. The idea was that the smoke and carbon particles released by fires (in turn caused by nuclear attacks on cities, where much of the world's fuel is stored) could have similar cooling effects to those known to be caused by the ash released in major volcanic eruptions - only worse (due to physical and chemical differences between ash and smoke). A seminal study in 1983, often called TTAPS (after its authors), confirmed this hypothesis and coined the term "nuclear winter."
Even using extremely crude modeling, TTAPS projected that a massive nuclear exchange between Russia and the U.S. could cause catastrophic cooling in the continental interiors - a change of as much as -35 degrees C (-63 degrees F). For comparison, the last global ice age, at its peak, saw average global cooling of only -5 degrees C (-9 degrees F) - though the cooling at continental interiors would have been more drastic. Later studies concluded that these changes would persist for around 3 years.
Nuclear winter studies continued until 1990 and then ceased abruptly (presumably the end of the Cold War sucked the urgency out of the issue). This fall, however, Alan Robock of Rutgers University and some of his colleagues have published several new studies on nuclear winter - the first such studies in almost 20 years.
Climate models today - and the computers to run them - are considerably more sophisticated than those of the early 1980s. Using these improved models, Robock et al. confirmed that the nuclear winter theory holds, in general. The temperature effects for a massive nuclear exchange should actually be slightly less extreme than originally predicted, but according to the new model they would last for over a decade, rather than just for a few years.
Taking a completely new approach, one study also examined a scenario no one bothered to consider during the Cold War: a regional nuclear conflict. They found that massive, superpower-style nuclear exchanges are not required to force major climate change. Even a relatively small nuclear exchange between, say, India and Pakistan, could cause average global surface cooling of over 1 degree C (1.8 degrees F) and peak cooling at continental interiors of around 4 degrees C (7 degrees F).
Interestingly, the studies found that the persistence of the climate changes did not depend on the size of the nuclear exchange. In other words, the climate effects from a regional nuclear war would last just as long as those from a global nuclear war, though they would be less extreme.
Recent modeling has also confirmed that nuclear exchanges will drastically reduce global precipitation, by as much as -45% for a massive superpower exchange and -10% for a regional exchange. In the former case, for instance, Northern Hemisphere monsoon seasons would disappear entirely.
These studies have weaknesses - for instance, they assume nuclear weapons will only target cities, where most smoke-generating fuel is gathered, rather than isolated military installations - but collectively they are a reasonable step towards updating the science of nuclear winter. After such a long hiatus, with nuclear proliferation looming in Asia and the Middle East, and even though nuclear winter itself is rather terrifying, I find it reassuring that long-neglected effects of nuclear weapons are being studied anew.
-- Eric Hundman
(Special thanks to Haninah for the illustration!)
UPDATE 7:10 PM: Russell Seitz says the whole nuclear winter thing has been oversold.
UPDATE 01/05/06 4:25 PM: Eric rebuts the rebuttal, here.