[Smudge]

When modeling was done for the Kuwait oil fires, there were two opposing expectations. One was that the fires would cause a kind of "year without a summer" similar to the Tambora eruption in 1815. The other was that the smoke would affect local weather conditions and be rained out within a week. In retrospect, we know that the latter proved most accurate.

But from what I understand, the crucial difference between the two models was how high the smoke would go, and as such the gap between the two models is actually a lot closer than the outcomes would imply. Had the conditions (weather, geography, etc) been ripe for more "self-lofting" of particulates, far more of them may have made it as far as the stratosphere, at which point they would stay in the atmosphere for far longer and would spread much further, without any weather patterns at that altitude to accelerate their return to the surface.