Attribution: You must attribute the work in the manner specified by the author or licensor ( but no in any way that suggests that they endorse you or your use of the work).Noncommercial ‒ you may not use this work for commercial purpose.No Derivative works ‒ You may not alter, transform, or build upon this work.Sharing ‒ Individual scientists are hereby granted permission, without fees or further requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in other subsequent works and to make unlimited photo copies of items in this journal for noncommercial use in classrooms to further education and science.

Nearly 30 yr since Alvarez et al. (1980) detected an enrichment of iridium at the Cretaceous-Palaeogene (K-Pg) boundary (65 Ma), the events that lead to the demise of the dinosaurs have become a feature of considerable scientific and public debate. It is generally accepted that an extraterrestrial body collided with the Earth 65 m.y. ago and that the 200-km-wide Chicxulub Crater on the Yucatan Peninsula, Mexico, is the mark of this impact (Hildebrand et al., 1991). This impact blasted melted asteroidal and target rock debris across the planet, depositing the K-Pg boundary impact rock layers (Fig. 1). The environmental consequences of the impact, and particularly the amount of thermal radiation it delivered, remain the hottest topic of the K-Pg debate.

Several models have suggested that the thermal radiation released by the K-Pg impact ought to have been sufficient to have ignited wildfires locally, if not globally (e.g., Melosh et al., 1990; Kring and Durda, 2002). Melosh et al. (1990) suggested that the re-entering ejecta from the impact could have delivered enough thermal radiation to ignite forests. Their model produced an average global thermal pulse of around 50 kW m−2, which ought to have been sufficient to ignite vegetation, although they noted that the amount of thermal radiation estimated to be delivered to the ground is near the lower limit required for ignition of solid wood. Hildebrand (1993) argued that this global average was unrealistic and that close to the impact site (e.g., 1000–2500 km distance), the thermal pulse was likely two to three orders of magnitude greater. At distances greater than 10,000 km, the thermal pulse ought to have been an order of magnitude less, implying that it had a regional effect. This might have offered places of refuge for organisms at greater distances from the impact.

Kring and Durda (2002) explored such regional effects by modeling the trajectories of the low- and high-energy ejecta from the Chicxulub impact. They suggested that 12% of the high-energy ejecta would be lost to space, and 25% of the material would re-accrete within 2 h, 55% within 8 h, and 85% within 72 h. This means that shock heating of the atmosphere likely occurred as a series of pulses. Their model suggested that the K-Pg impact ejecta could have ignited wildfires on several continents around Earth. Colorado, in the United States, for example, ought to have received a thermal pulse in the order of 150 kW m−2, more than sufficient to ignite even wet vegetation in the area. While such results estimate that the direct radiation from the Chicxulub impact plume ignited wildfires across large areas of the planet, it has been suggested that global wildfires may have been ignited by lightning strikes initiated by charge separation in an atmosphere choked with dust particles, where dead forests would be a prime ignition target (Wolbach et al., 1990). Shuvalov and Artemieva (2002) discussed that it was more likely that forests were killed first and later burned. This theory would allow the period of wildfires to continue for several years, such that a large number of local wildfires in dead forest could have released large volumes of soot.

Observational paleontological data have been used to both support and dispute the numerical models. Wolbach et al. (1985, 1988, 1990) reported an enrichment of soot in the K-Pg boundary impact rocks that is not only isotopically uniform, suggesting a single source, but also bears an isotopic signature consistent with the burning of biomass. However, they note that some fossil carbon sources have the same isotopic range, and therefore the soot could also be sourced from combustion of hydrocarbons. Enhanced abundance of polycyclic aromatic hydrocarbons (PAHs) have been reported in K-Pg boundary impact rocks in Denmark, Italy, and New Zealand (Venkatesan and Dahl, 1989) and have been suggested to support the wildfire hypothesis.

In contrast, the record of fossil charcoal from K-Pg boundary impact rocks has largely provided arguments against the wildfire hypothesis. Fires have been shown to occur throughout non-marine K-Pg sequences, but the total amount of charred material in and around the K-Pg boundary itself is typically lower than the average amount recorded in either Cretaceous or Palaeogene strata. This pattern is consistent across the western interior of North America, from sites near the impact in the United States and up into Canada (Sweet and Cameron, 1991; Scott et al., 2000; Belcher et al., 2003, 2005). Moreover, multi-method and multi-proxy analyses have revealed that extensive K-Pg wildfires were unlikely (Belcher et al., 2003, 2005, 2009) and that the K-Pg soot and PAHs reveal a signature consistent with hydrocarbon combustion at the impact site (Belcher et al., 2005, 2009; Harvey et al., 2008). The lack of evidence for wildfires, along with the abundance of non-charred plant remains found in non-marine K-Pg boundary impact rocks, imply that the thermal radiation from the K-Pg impact produced ground temperatures on the order of 325 °C, with a short-lived peak of no more than 545 °C (Belcher et al., 2003). These temperatures require that the thermal radiation delivered by the impact cannot have exceeded 19 kW m−2 at any point, and not more than 6 kW m−2 for more than a few hours (Belcher et al., 2005). These figures contrast with estimates from numerical models that suggest values around 8 times greater for this geographic area (e.g., Kring and Durda, 2002).

A new numerical model by Goldin and Melosh (2009, p. 1135 in this issue of Geology) addresses the assumptions made in previous models and provides new estimates of the K-Pg thermal radiation. Goldin and Melosh have, for the first time in 25 yr since the wildfire hypothesis was posited, enabled estimates from numerical models and paleontological data to present a united front toward resolving this much- disputed burning question.

Goldin and Melosh have modeled the atmospheric re-entry of the K-Pg asteroid impact melt spherules arriving at distal sites. This new model includes both estimates of spherule opacity and also, importantly, the self-shielding effect of the spherules raining out through the atmosphere. As more and more spherules bombard and settle out of the atmosphere, they are thought to block an increasing portion of the downward thermal radiation emitted by later falling spherules. These two features (opacity and self shielding) have not been previously considered in K-Pg thermal radiation models. This not only highlights the relationship between the falling particles themselves and the thermal radiation but also considers their interaction with the atmosphere in order to understand how energy delivery to the upper atmosphere relates to transmission of thermal radiation to the Earth's surface.

The atmospheric re-entry of some 20,000 spherules per cm2 is shown to have heated the atmosphere on the order of 1600 K, in which each spherule loses kinetic energy due to increasing drag on entering an increasingly dense atmosphere. This heat is radiated at infrared wavelengths, heating the surrounding air. The spherules are estimated to reach terminal velocity at ~70 km altitude, which leads to an opaque cloud of spherules accumulating below, thus creating a self-shielding mechanism. Goldin and Melosh estimate that the flux of thermal radiation to the Earth's surface 5 min after initial spherule bombardment is on the order of 5 kW m−2 and remains similar thereafter. The asteroid impact is believed to have thrown large quantities of submicron-sized dust up into the atmosphere—Goldin and Melosh account for the interaction of this dust with the spherules and the thermal radiation, where a “hot cap” of uncondensed silicate droplets form a cloud “bank” of heat and may have increased the thermal flux to the ground surface to a maximum of 19 kW m−2, but not more than 10 kW m−2 for >20 min. These figures are in good agreement with observed paleontological data (absence of fires) that estimate a maximum delivery of 19 kW m−2, and not more than 6 kW m−2, for more than a few hours (Belcher et al., 2003, 2005). These data imply that surface temperatures did not exceed 325 °C (for more than an hour).

It is recognized that there was a major disruption to plant communities across the K-Pg boundary (Tshudy et al., 1984; Sweet, 2001; Nichols and Johnson, 2002). These new model-based results, taken together with the abundant literature on paleontological indicators of fire occurrence, suggest that extensive wildfires were not the cause. This model-data agreement does not eliminate the role of relatively high temperatures (on the order of a couple of hundred degrees centigrade) in some of the extinctions seen at this time; but does suggest that the thermal pulse component of the K-Pg impact was not as significant as has been previously thought. Additional mechanisms might be required to fully explain the K-Pg extinctions, in which the effects of an “impact winter” (Galeotti et al., 2004) followed by global warming (O'Keefe and Ahrens, 1989; Beerling et al., 2002), acid rain (Hildebrand and Boynton, 1989; Prinn and Fegley, 1987), and the additional stress of Deccan volcanism (Courtillot et al., 1988; Chenet et al., 2007) on the Earth system might be more fully investigated. It is clear that significant environmental perturbations would be expected following the K-Pg impact event, even in the absence of extensive K-Pg wildfires.