The Day The Earth Nearly Died - programme summary

250 million years ago, long before dinosaurs roamed the Earth, the land and oceans teemed with life. This was the Permian, a golden era of biodiversity that was about to come to a crashing end. Within just a few thousand years, 95% of the lifeforms on the planet would be wiped out, in the biggest mass extinction Earth has ever known. What natural disaster could kill on such a massive scale? It is only in recent years that evidence has begun to emerge from rocks in Antarctica, Siberia and Greenland.

"At the end of the Permian you'd see virtually nothing alive"

Professor Peter Ward, University of Washington

The demise of the dinosaurs, 65 million years ago (at the so-called K/T boundary), was as nothing compared to the Permian mass extinction. The K/T event killed off 60% of life on Earth; the Permian event 95%. Geological data to explain the destruction have been hard to find, simply because the rocks are so old and therefore subject to all kinds of erosion processes. It seems plausible that some kind of catastrophic environmental change must have made life untenable across vast swathes of the planet.

"A volcanic eruption ten thousand times larger than man has ever seen"

Professor Vincent Courtillon, University of Paris

The world's biggest volcanoes
In the early 1990s, the hunt for evidence headed for a region of Siberia known as the Traps. Today it's a sub-Arctic wilderness but 250 million years ago, over 200,000km² of it was a blazing torrent of lava. The Siberian Traps were experiencing a 'flood basalt eruption', the biggest volcanic effect on Earth. Instead of isolated volcanoes spewing out lava, the crust split and curtains of lava were released. And the Siberian flood eruption lasted for millions of years. Could volcanic activity over such a long time alter the climate enough to kill off 95% of life on Earth?
Vincent Courtillon used a much smaller flood basalt eruption, in Iceland in 1783, as the basis for some calculations. Writing in the 18th century, Benjamin Franklin (then American Ambassador in Paris) described 1784 as a year without a summer. Ash from the eruption blacked out the sky and crops failed across Europe. Courtillon extrapolated the climatic impact of the Siberian Trap eruption from the records of the Icelandic event. He deduced that a 'nuclear winter' lasting decades would be followed by rapid global warming due to the increased level of greenhouse gases in the post-eruption atmosphere.

"It's the equivalent of a billion atomic bombs going off at the same place"

Dr Michael Rampino, New York University

Vincent believes the disruption of cooling followed by warming could cause the Permian extinction but other geologists disagree. Peter Ward returned to the Siberian Trap data to estimate the amount of carbon dioxide - and global warming - that could result. His worst case scenario is a temperature rise of 5°C, enough to kill off many species but not the 95% wipeout that ended the Permian.
If the Siberian eruptions were not deadly enough, what other effects might be at work? To try to answer that, Michael Rampino set out to establish an even more fundamental piece of data: how long did the extinction take? He studied rock sedimentation rates in the Alps and concluded that the Permian killer had stalked the planet for just 8,000-10,000 years, far less than had been thought. His mind turned to ways of causing such catastrophic destruction in - on geological timescales - the blink of an eye. He wanted to explore the possibility of a meteorite strike.
The hunt for meteor evidence
Meteor strikes that wipe out life may sound like sci-fi but it's generally accepted that an impact sparked the K/T extinction and the end of the dinosaurs. That meteorite was 10km wide and left a crater in what is now the Gulf of Mexico. The dust raised by such an impact could make global temperatures plummet overnight. How big would any Permian meteorite have to be? Rampino suggests one just 50% bigger could cause sufficient environmental change. There is one huge flaw in this argument: where is the crater?

"The original crater is completely drowned by lava"

Adrian Jones, University College London

Adrian Jones models the effects of impact on the Earth's geological crust. He has a hunch that meteorite crater hunters are looking for the wrong thing. After an impact, the crust rebounds to form a large shallow crater. If the meteorite if truly massive though, an extra process occurs. The combined heat of the impact and rebound is enough to melt the crust. Lava floods through and the crater disappears beneath new crust. If he's right, the Permian meteorite crater can't be found because it doesn't exist.

"When a meteorite wiped out the dinosaurs it left ample evidence in its wake"

Greg Rettaleck, University of Oregon

All of which serves to help proponents of the meteorite impact theory. Its detractors, though, point out that meteors leave several trails in their wake - fragments of minerals that have come from space. Greg Rettaleck mounted an expedition in the mid-1990s looking at Permian rock beds in the Antarctic. Some of the quartz grains looked like they had been fractured by a very energetic process - a meteorite?
Although this was evidence for a strike of some sort, there were unanswered questions as well. The K/T meteorite left a trail of iridium - characteristic of space materials - around the world. Yet there is no evidence the Permian strike did the same.

"No need to guess any more... the whole extinction from beginning to end"

Paul Wignall, University of Leeds

Paul Wignall is a British geologist who doubts a meteorite caused the mass extinction 250 million years ago. In the late 1990s he had a hunch of a way to prove his beliefs, a good idea of where to look for new evidence: Greenland. Permian rocks are hard to find because they are usually just thin layers, yet his trip yielded rock beds metres thick. This was more than just new evidence; it was the best he could have hoped to find.
Carbon copious
The Greenland rock told a very different story to that Michael Rampino had found in the Alps. Instead of a rapid event of under 10,000 years, the extinction beds Wignall examined lasted 80,000 years and showed three distinctive phases in the plant and animal fossils they contained. The extinction appeared to kill land and marine life selectively at different times. Such a long process contradicted the catastrophic meteorite theory but Wignall couldn't explain what had come close to killing all life on Earth. His best clue was the carbon isotope balance in the rock, which showed an increase in carbon-12 over time. The standard explanation - rotting vegetation - could not have caused such a marked effect. Wignall was curious what this could mean.
An answer came from geologist Gerry Dickens, who knew just how to get large amounts of carbon-12 rapidly, thanks to his work with offshore drilling companies in the USA. He had spent time helping them try to tap reserves of frozen methane hydrate from the seabed of the Gulf of Mexico. He knew methane hydrate is found around many of the world's coasts. Dickens wondered how large a rise in sea temperature was necessary to cause the solid chemical to gasify and ascend to the atmosphere. Experiments suggested a rise of 5°C would be sufficient. And he was amazed to see how much gas came from pieces of solid methane hydrate that were placed in water.

"The south of England would turn into the Sahara Desert"

Michael Benton, University of Bristol

When Paul Wignall learned of Dickens' findings, he used his carbon-12 data to estimate how much methane hydrate would have to be released to affect the isotope balance. Methane is one of the most potent greenhouse gases and he deduced that unlocking frozen methane hydrate would have caused a temperature rise of 4-5°C over time. Not enough to kill off 95% of life on Earth but he realised this was a compounded effect. A rise of about 5°C must already have occurred to prompt the frozen methane to melt. The combined temperature rise of 10°C is generally accepted as a figure able to cause truly mass extinction.
So it seems likely there were two Permian killers. The Siberian Traps did erupt, contributing first to a nuclear winter cooling effect (caused by dust) and and then to global warming (due to greenhouse gases). Over 40,000 years, some land animals gradually died out while life in the seas lived relatively calmly on, as the water temperature gently rose. Then the seas gave up their frozen methane. In just 5,000 years, there was massive loss of species from the world's oceans. In a third and final phase of the extinction, the Permian killer returned to stalk the land for another 35,000 years. By the end of that process, 95% of the Earth's species were extinct.