Some 65 million years ago, the skies over India darkened as one of Earth’s biggest volcanic eruptions burbled from below. It rumbled on for millions of years, blocking out sunlight and casting a chill globally, to produce what we know today as the Deccan Traps.
Many believe the eruption sent the dinosaurs into severe demise before an asteroid collision finally finished them off. But just how the Earth produced such vast volumes of lava (covering an area greater than the Australian states of New South Wales and Victoria combined) has remained a bit of a mystery. Now a new study by a pair of geologists in Canada shows that the eruption may have been fed by not one, but two deep mantle plumes.
Like the hot air that rises to create a thundercloud, mantle plumes are thought to be narrow regions of convection that fast-track hot material all the way up from the core-mantle boundary and through the Earth’s 2,900-kilometre-thick layer of hot rock called the mantle.
There are thought to be a number of active mantle plumes today, some of which have created a chain of volcanic islands as the oceanic plate glides across the plume top. The Hawaiian-Emperor seamount chain, the Easter Islands and the Walvis Ridge (culminating in the island of Tristan da Cunha) are just a few examples.
By calculating past movements of tectonic plates, scientists have shown that the mantle plume currently underneath the Indian Ocean Island of Réunion was probably responsible for melting the mantle underneath the Deccan region 66 million years ago. But scientists have remained perplexed as to how one mantle plume could produce such a prodigious volume of melt.
Petar Glišović and Alessandro Forte from the University of Quebec in Montréal, Canada, decided to revisit the Deccan conundrum using a model of mantle convection and running it in reverse for 70 million years.
“This is a really hard problem as it is impossible to undo heat diffusion,” explains James Wookey, a geophysicist at the University of Bristol in the UK, who wasn’t involved with the study.
So the pair ran many iterations of their model, with each scenario starting 2.5 million years ago with a different mantle structure configuration, and run forwards until one produced current mantle conditions.
Taking the best fit and rewinding mantle dynamics by 70 million years, Glišović and Forte’s model showed that the Réunion mantle plume was situated underneath the Deccan region of India, as expected, but to their surprise there was also another mantle plume nearby at that time, responsible for feeding the volcanism on the East African island of Comoros today.
Publishing in Science, Glišović and Forte calculated that the combined heat of the Réunion and Comoros mantle plumes would have been sufficient to melt around 60 million cubic kilometres of mantle at the time of the eruption; more than enough to feed the Deccan Traps. “We see mantle plumes merging and splitting in our forward running models of mantle convection, so the idea that these two plumes merged in the past is certainly plausible,” says Wookey.
The model also shows that the Comoros plume had lost most of its heat by 40 million years ago, while the Réunion mantle plume ran out of steam around 20 million years ago. Today, both plumes are mere shadows of their former selves. But Wookey cautions against taking the findings too literally, adding: “the physics of the model is reasonable, but whether the mantle movements are precisely what the Earth actually did is another matter.”