The currently accepted theory for P-T mass extinction is environmental change triggered by the Siberian Traps volcanic activity.

A paper(1) suggests that the Siberian Traps were caused by a hotspot created after an impact event on what was back then the antipode of Siberia: the Wilkes Land crater.

Thus, the impact may have triggered the “Great Dying” at the end of the Permian and contributed to the development of the hot spot that produced the Siberian Traps and now may underlie Iceland.

However, another more recent review paper(2) mostly discredits this theory:

Finally the recent report of a 500-km-diameter impact structure of possible Permian–Triassic age in the Antarctic(1), is based entirely on satellite gravity measurements, provides no geological context, and completely ignores the absence of any predictable regional and global geological effects from such a large impact event, e.g., the absence of any thick ejecta layer at nearby Antarctic Permian–Triassic boundary sections.

Leaving aside the discussion of whether such a thing caused the formation of the Siberian Traps and thus the P-T mass extinction, in theory, can an impact on one side of the Earth cause volcanism on its antipode?

Also very important, why would this happen? What's the supposed mechanism that converts short term seismic activity to volcanic activity derived from long term hot spots?

Note that I'm not asking about in situ volcanic activity (such as the case of Sudbury). The impact also has to be reasonable in magnitude: I don't consider the moon-forming event that destroyed the young Earth as a reasonable impact in magnitude.

(1) von Frese, R. R. B., L. V. Potts, S. B. Wells, T. E. Leftwich, H. R. Kim, J. W. Kim, A. V. Golynsky, O. Hernandez, and L. R. Gaya-Piqué (2009), GRACE gravity evidence for an impact basin in Wilkes Land, Antarctica, Geochem. Geophys. Geosyst., 10, Q02014, doi:10.1029/2008GC002149.

(2) Bevan M. French, Christian Koeberl (2010), The convincing identification of terrestrial meteorite impact structures: What works, what doesn't, and why, Earth-Science Reviews, 98, Issues 1–2, Pages 123-170, doi:10.1016/j.earscirev.2009.10.009.

edit 18/10/2015

There was a paper published recently in Science that suggests that this is exactly what happened in the K-Pg extinction event. The Chicxulub increased the intensity of the Deccan volcanism, on the other side of the planet. Here are some links:

Volcano-asteroid combo may have done in the dinosaurs

Could dino-killing asteroid have pumped up giant volcanoes?

edit 31/12/2016

Not everyone agrees! For example:

Triggering of the largest Deccan eruptions by the Chicxulub impact: Comment

  • $\begingroup$ Dynamic waves from impacts (just like earthquakes) can certainly act as trigger for a volcanic eruption on the other side. $\endgroup$
    – stali
    Dec 19, 2014 at 0:30
  • $\begingroup$ Probably. The Hellas Basin google.com/mars/… basin is opposite Tharsis Tholus google.com/mars/… Tholus on Mars $\endgroup$
    – mankoff
    Dec 22, 2014 at 3:28

1 Answer 1


Note: this is a part answer, based on recent research and modelling

There has been considerable research recently in what is known as antipodal disruption, in particular, a group from Princeton University have produced a model that overcomes the main deficiencies of commonly used models that have the Earth as a 'featureless sphere' (1)(2). Simulations using a symmetric spherical Earth produce a deceptively large peak at the antipode (2) - this is often cited as being some form of proof for the P-T example that you provided in the question and an antipodal relationship between the Deccan Traps and the Chicxulub crater (1).

However, the Princeton team's simulation (1), (2) show that there ought to be an antipodal effect, but when the elliptical shape and the heterogeneity of the surface is taken into account - where the waves end up and their displacement when they get there are quite different to the spherical models (1), to the order of 4 times less displacement and the antipode effects are not on a point source (2).

This results in a disagreement with the Deccan Traps-Chixculub relationship, as shown in the late Cretaceous model below, from (1):

enter image description here

This has implications for not only seismicity, but for volcanism and hot spot activity, due to the modelled impact seismic waves (across the surface and through the Earth) showing to be unfocused and scattered, this results in less severe ground displacement, tsunamis, and seismic and volcanic activity than previously theorized. Their models showed that the antipodal region had 'chimneys' of peak stress, strain and velocity.

Meschede et al. 2011 (2) conclude from their model that there would be greater seismicity in the region around the antipodal 'chimneys', this increase in seismicity can trigger volcanism, and even though the impact energy has been modelled to be insufficient (by several orders of magnitude) to induce melting to form an antipodal hotspot.

Some ongoing research by Hagstrum, 2005 (3) has suggested that oceanic impacts may result in hotspot/large igneous province antipodal pairs, this is due to, according to Hagstrum, oceanic crust having a greater seismic efficiency that continental crust. The mechanism is, according to Hagstrum,

an inherently antipodal mechanism of hotspot formation is proposed in which one hotspot forms at an oceanic large-body impact site and a second hotspot and possible LIP are formed from seismic energy focused in the lithosphere and upper asthenosphere at the impact’s antipode.

However, it is not clear if the non-spherical heterogenous nature of the Earth's lithosphere was taken into account, given that, as stated, the impact energy is insufficient to cause melting, Ivanov and Melosh (4) state that the seismic mechanism may contribute as a trigger for a "pregnant" hotspot to activate, not through melting, but through increased seismicity (fracturing rocks etc).

(Note, the use of "pregnant" is their words, not mine)


The models created by Meschede et al. (2) have found that it is theoretically possible for localised antipodal volcanism to occur, but this volcanism would not necessarily be the direct result of the impact itself, but due to the contribution of an increase in antipodal seismicity (increased earthquakes fracturing rocks in the lithosphere etc).

Meschede et al. (2) contend that their model indicates that the impact energy is very much insufficient to result in a sustained mantle plumes and hotspot volcanic regimes as in the previous models. However, there is still ongoing research about the impact energy contribution to the formation of hotspots and 'traps', particularly for oceanic impacts (3), but this is very much uncertain (2) and unlikely to be the direct cause (4).


(1) Kelly, 2011, Impact study: Princeton model shows fallout of a giant meteorite strike, News at Princeton, Princeton University

(2) Meschede et al. 2011, Antipodal focusing of seismic waves due to large meteorite impacts on Earth, Geophysical Journal International

(3) Hagstrum, 2005, Antipodal hotspots and bipolar catastrophes: Were oceanic large-body impacts the cause?, Earth and Planetary Science Letters

(4) Ivanov and Melosh, 2003, Large Scale Impacts and Triggered Volcanism, in Large Meteorite Impacts

  • $\begingroup$ One thing that I don't understand is why would a geologically very short transient of seismic activity cause long term hot spot and volcanic activity spanning at the least thousands of years? $\endgroup$
    – Gimelist
    Dec 20, 2014 at 8:24
  • $\begingroup$ The articles suggest that if the area was susceptible to it, this could be a final 'trigger'. $\endgroup$
    – user889
    Dec 20, 2014 at 8:26
  • $\begingroup$ I'm still not buying it. Then it would've happened anyway. But again - why would seismic activity cause melting? What's the link? $\endgroup$
    – Gimelist
    Dec 20, 2014 at 8:27
  • $\begingroup$ The impact energy is even though the impact energy has been modelled to be insufficient (by several orders of magnitude) to induce melting to form an antipodal hotspot., and as I mentioned in the conclusion, their models show that there is no definitive direct link to volcanism. $\endgroup$
    – user889
    Dec 20, 2014 at 8:32
  • $\begingroup$ okay, I can see your edited question has made my answer incomplete (even off topic) $\endgroup$
    – user889
    Dec 20, 2014 at 8:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.