In the catastrophe at the end of the Cretaceous era, when the dinosaurs disappeared, a soot deposit has been found at the geological boundary that is equivalent to the combustion of at least 90% of the terrestrial biomass. The discovery of this soot layer, and its interpretation, was made by Ed Anders and colleagues at the University of Chicago, with several other researchers later confirming its existence and panglobal extent. Even prior to the Chicago group's discovery there was other evidence in hand of extensive fires at the time, charcoal having been identified in some locations.

There is plentiful evidence available for an impact's having occurred at the time of the Cretaccous/Tertiary (K/T) extinction in which the dinosaurs disappeared, and for similar impacts at the epochs of other mass extinction events. What is especially pertinent here is the soot layer: How was it produced? It seerns to indicate that the Earth went up in flames, or at least most of the plant life did. There are also paleontological data suggesting that although many marine creatures living in the top 10 meters of the oceans were wiped out, organisms living deeper down survived. One suggestion for this has been that the dust thrown out from the impact site (or sites, since it is now widely believed that multiple impacts occurred at the time of the K/T extinctions) could have caused the ignition of fires through induced lightning. It is well known that lightning may be set off by volcanic ash falls, the settling ash causing charge separations to occur in the atmosphere, which are the source of lightning bolts. With a huge impact, it is to be anticipated that there would be global fallout, and the iridium layer at the K/T boundary is evidence for just such an occurrence.

There are other reasons that one would expect the surface temperature of the Earth to be raised to an extremely high level quite likely well above the ignition point of even a lush green tropical jungle, in the aftermath of a large impact. The

on the Yucatan peninsula of Mexico is at least 180 kilometers across, and there are some arguments that a more realistic diameter might be 400 kilometers. The problem is that 65 million years have elapsed since it was formed and such a huge crater on Earth quickly slumps under gravity so that a series of concentric rings is created. It is not always clear which of the rings was formed by a crater that was evacuated and then slumped and which by the outgoing shock wave. Dating of the crater indicates that it was formed at a time indistinguishable from that of the K/T boundary, the probability of that occurring by chance being much less than 1%. If this crater were produced by an incoming asteroid about 10 kilometers in size (as originally estimated for the K/T projectile on the basis of the amount of iridium found) meeting the Earth at about 25 km/sec, then the energy released is equivalent to about 100 million Mt. Such an impactor, however, would form a crater only 60 to 100 kilometers in diameter. Thus it seems that in the case of Chicxulub, the pro- jectile was a lot larger-perhaps 20 to 40 kilometers in size--and the explosion much more energetic than originally visualized by Alvarez and colleagues.

The total volume of rock excavated from the crater was huge-at least 100,000 cubic kilometers and perhaps ten times that. Remember that the atmosphere of the Earth is very thin (in terms of thickness, as opposed to density). Atmospheric pressure drops to about one millionth that at the surface by the time you get to an altitude of 100 kilometers, a distance that is equivalent to half or less of the diameter of the crater. Thus the atmosphere is only a tenuous skin over the crater, and a large fraction of the pulverized rock ejected in the impact would be thrown up and out of the atmosphere on what is termed a ballistic trajectory. The rock is expelled upward and sideways, some of it moving swiftly enough to escape the Earth's gravitational pull, but most of it looping up above the atmosphere and reentering elsewhere around the globe. To take a conservative line, we will estimate that just 1% of the minimal 100,000 cubic kilometers of ejected rock is thrown above the atmosphere on ballistic trajectories, or about 1,000 cubic kilometers. Most will reenter in this way, the fraction escaping having implications for the concept of panspermia, the spreading of life from one planet to another. The sizes of the lumps of rock that will reenter range from that of sand grains to very large boulders, and most of these reentries would occur in the following few hours. It takes a weather satellite about 90 minutes to orbit the Earth at a low altitude, and the same laws of celestial mechanics apply to the ejected rock. For the Chicxulub impact, some rocks would have traveled just a few hundred kilometers from the impact site, landing in Haiti or Texas.2 Some would have crossed the Atlantic to fall in the region where Europe now sits, while others would have performed a half or even full orbit around the planet. The point is that a wave of debris cascaded down upon the atmosphere from above, producing one of the greatest global meteor showers ever, although it is unlikely that the inhabitants below appreciated it. As the ejecta reentered, it would have burned up, producing an essentially unbroken wave of meteors above the flora and fauna below. The whole sky would have been filled with burning meteoroids, which an hour or so before had been rocks deep below the surface of the ground in a corner of Mexico.

How much energy will be irradiated downward by this rock as it burns up? At reentry speeds varying from a few kilometers per second up to 11 km/sec, 1,000 cubic kilometers of rock will release energy equivalent to about a week's worth of solar energy spread over the whole planet. In many ways, one can imagine the situation as being analagous to a huge griller being located 50 to 100 kilometers above the surface, boosting the surface temperature to over 1,000 C. It is only to be ex- pected that under such circumstances the plant life of the continents would be rapidly desiccated and then ignited. No wonder there is a dense soot layer at the geological boundary at which the dinosaurs died. Computations by Owen Toon and colleagues at NASA-Ames Research Center show that the threshold for forest ignition globally would be reached with the heat produced by ballistic ejecta from an impact with an energy of about 100 million Mt, which is the minimum that seems to have been released at the time. An impact energy ten times lower would still produce ejecta capable of causing forest ignition over an area greater than that of the United States.

Much of the irradiation will come from the burn-up of smaller rocks and boulders, from millimeter to decameter sizes, at high altitude. We have already seen, however, that rocks as big as 50 to 100 meters in size will fragment and detonate in the atmosphere, as in the case, of Tunguska. Can tbe Tunguska event tell us anything about the ignition of forest fires? Any visitor to ground zero in Siberia is amazed not only by the radial tree falls (the trunks lying so as to point away from the epicenter), but also by the fact that the trees are largely charred (but only on one side, that being the side directed inward). It is easy to imagine what happened. The 10- to 20-Mt explosion produced a sufficiently large radiative flux so as to ignite the trees even in the soggy swamps of the region-no magnifying glass needed here to concentrate the Sun's rays and get a fire going. Nevertheless, we know that the fire did not consume the forest in its entirety. Within minutes the blast wave reached the burning trees and blew the flames out, leaving them charred on one side, but not burned through.

There are several mechanisms, then, that might be expected to lead to global fires in the aftermath of a large impact. As explained earlier, at the K/T boundary the layer of soot has a thickness indicative of plant materials equivalent to at least 90% of the current worldwide biomass being incinerated. For small impacts, such as Tunguska, we have unarguable evidence of forest ignition, with other perturbations of the environment. For example, the evenings following the Tunguska explosion (on June 30, 1908) were noted as being "White Nights" throughout Europe, with people able to read newspapers from the sky glow even after midnight. To my knowledge, it has never been explained properly whether this was due to sunlight scattered by high-altitude dust, some form of induced airglow, or some unknown physical cause. There have been claims that dust from the shattered projectile is detectable in Antarctic ice layers that were laid down around then, which would imply global (not just hemispherical) spreading of the dust, but this has been contradicted by others. Such dust identification would be important because there is weak coupling across the equator between the circulations of the atmosphere in the northern and the southern hemispheres. If dust from Tunguska (which is near latitude 61 0 North) were found in Antarctica, it would demonstrate that even small impacts are capable of causing global perturbation of the environment.

The maintenance of a high-altitude dust layer would lead to the scattering away of some significant fraction of the incoming sunlight and thus a cooling of the climate. Again, there have been claims that the Tunguska explosion, minor though it was, caused cooling of the northem hemisphere by a degree or so in the few years following 1908, but because normal fluctuations induced by circulation patterns on the Earth, solar cycle variations, and so on, can cause deviations of the same magnitude, it is not possible to do more than point the finger of suspicion at Tunguska. The climatic effects of singular occurrences like Tunguska are not of great concern, since terrestrial phenomena such as volcanic eruptions can cause similar climatic excursions through the dust that they release into the atmosphere. It is of interest, however, to understand the climatic deviation caused by Tunguska for scaling up to larger impacts, these causing perturbations far beyond those producible by mundane cataclysms such as volcanoes. .