The models Marshall and colleagues are studying begin with known characteristics of the physical world. For example, scientists know that the level of oxygenation of the deep ocean depends on the transport of oxygen-rich waters from the surface. But oceanic life consumes oxygen. Marshall's models suggest that if ocean circulation were weaker than it is now, consumption of oxygen might outstrip the supply of oxygen, leading to oxygen-poor (anoxic) deep oceans rich in dissolved organic carbon. Then, if a rapid change in ocean circulation were to flush the deep ocean—bringing abyssal waters to the surface—the rapid release of carbon dioxide to the atmosphere would have significant biological impacts, perhaps triggering extinctions.

But what might cause the deep ocean waters to become stagnant in the first place? "We find through computer-based experimentation that the answer depends on the strength of the atmospheric hydrological cycle, the pole-equator temperature gradient, and the geographical distribution of land and sea, amongst many other things," Marshall says.

Overturning stream function of late Permian thermal mode and haline mode circulation scenarios. In the thermal mode, much as in the modern climate, the abyss is ventilated by polar convection triggered by cooling, drawing oxygen down into the deep ocean. In the haline mode, saline-driven convection drives a shallow cell with sinking from the subtropics. The abyss is warm, stagnant, and anoxic. Depending on parameters, the circulation flips between the two modes every few thousand years.

Specifically, evidence suggests that by the end of the Permian period, the land masses of the Earth had aggregated into a single, hemispheric-scale supercontinent. Sea level was hundreds of meters below where it is today, the climate was warm and dry, and there were no polar ice caps. The scientists theorize, and simulate on the computer, that in this warm environment enhanced evaporation in the subtropics could have triggered haline convection, or the churning of seawaters caused by the increased density of water that comes with increased salinity. If this situation did occur, salt-heavy waters would have sunk to mid-depth, leaving the abyssal ocean stagnant, warm, and anoxic.

But the models also show that the haline mode is unstable and eventually flips to a mode reminiscent of the present climate. In this thermal mode, cooling triggers deep-reaching convection at the poles, flushing the deep ocean with oxygen-rich water. This could explain how the ocean became life-friendly once again.

Exactly when and how all this might have happened is under study. Currently the researchers are using their models to critically examine the conditions under which oceanic thermal and haline modes might be induced. "We find that, depending on the parameters, the ocean can remain for many thousands of years in one mode and then flip to the other mode," Marshall says. "The distribution of land and sea can favor one mode of circulation over another."

So how close are scientists to agreement on the causes of the late Permian extinction? "A consensus is not going to occur easily or often," Knoll replies. "I'm not sure how close we are to solving it. But we know lots more than we did even 10 years ago."

He adds, "It's predictive modeling like John's that's going to tell us what we should be looking for to answer these questions."

This research is supported by the National Science Foundation.

Team Members
Stephanie Dutkiewicz
Mick Follows
John Grotzinger
Chris Hill
John Marshall
Rong Zhang