It all starts in 2025, as tensions between India and Pakistan escalate over the contested region of Kashmir. When a terrorist attacks a site in India, that country sends tanks rolling across the border with Pakistan. As a show of force against the invading army, Pakistan decides to detonate several small nuclear bombs.

The next day, India sets off its own atomic explosions and within days, the nations begin bombing dozens of military targets and then hundreds of cities. Tens of millions of people die in the blasts.

That horrifying scenario is just the beginning. Smoke from the incinerated cities rises high into the atmosphere, wrapping the planet in a blanket of soot that blocks the Sun’s rays. The planet plunges into a deep chill. For years, crops wither from California to China. Famine sets in around the globe.

This grim vision of a possible future comes from the latest studies about how nuclear war could alter world climate. They build on long-standing work about a ‘nuclear winter’ — severe global cooling that researchers predict would follow a major nuclear war, such as thousands of bombs flying between the United States and Russia. But much smaller nuclear conflicts, which are more likely to occur, could also have devastating effects around the world.

This week, researchers report that an India–Pakistan nuclear war could lead to crops failing in dozens of countries — devastating food supplies for more than one billion people¹. Other research reveals that a nuclear winter would dramatically alter the chemistry of the oceans, and probably decimate coral reefs and other marine ecosystems². These results spring from the most comprehensive effort yet to understand how a nuclear conflict would affect the entire Earth system, from the oceans to the atmosphere, to creatures on land and in the sea.

Scientists want to understand these matters because the nuclear menace is growing. From North Korea to Iran, nations are building up their nuclear capabilities. And some, including the United States, are withdrawing from arms-control efforts. Knowing the possible environmental consequences of a nuclear conflict can help policymakers to assess the threat, says Seth Baum, executive director of the Global Catastrophic Risk Institute in New York City, who has studied the risks of triggering a nuclear winter. “Fleshing out the details of ways in which it can be bad is valuable for helping inform decisions,” he says.

Cold-war forecasts

Nuclear-winter studies arose during the cold war, as the United States and the Soviet Union stockpiled tens of thousands of nuclear warheads in preparation for all-out assaults. Alarmed by leaders’ bellicose rhetoric, scientists in the 1980s began running simulations on how nuclear war might change the planet after the initial horrific deaths from the blasts^3,4. Researchers including the US planetary scientist and communicator Carl Sagan described how smoke from incinerated cities would block sunlight and plunge much of the planet into a deep freeze lasting for months, even in summer⁴. Later studies tempered the forecasts somewhat, finding slightly less-dramatic cooling⁵. Still, Soviet leader Mikail Gorbachev cited nuclear winter as one factor that prompted him to work towards drawing down the country’s nuclear arsenals.

After the Soviet Union collapsed in 1991, the world’s stockpiles of nuclear weapons continued to drop. But with many thousands of warheads still in existence, and with more nations becoming nuclear powers, some researchers have argued that nuclear war — and a nuclear winter — remain a threat. They have shifted to studying the consequences of nuclear wars that would be smaller than an all-out US–Soviet annihilation.

That includes the possibility of an India–Pakistan war, says Brian Toon, an atmospheric physicist at the University of Colorado Boulder who has worked on nuclear-winter studies since he was a student of Sagan’s. Both countries have around 150 nuclear warheads, and both are heavily invested in the disputed Kashmir border region, where a suicide bomber last year killed dozens of Indian troops. “It’s a precarious situation,” says Toon.

Both India and Pakistan tested nuclear weapons in 1998, highlighting growing geopolitical tensions. By the mid-2000s, Toon was exploring a scenario in which the countries set off 100 Hiroshima-size atomic bombs, killing around 21 million people. He also connected with Alan Robock, an atmospheric scientist at Rutgers University in New Brunswick, New Jersey, who studies how volcanic eruptions cool the climate in much the same way that a nuclear winter would. Using an advanced NASA climate model, the scientists calculated how soot rising from the incinerated cities would circle the planet. All around the dark, cold globe, agricultural crops would dwindle.

But after a burst of publications on the topic, Robock, Toon and their colleagues struggled to find funding to continue their research. Finally, in 2017, they landed a grant worth nearly US$3-million from the Open Philanthropy Project, a privately funded group in San Francisco that supports research into global catastrophic risks.

The goal was to analyse every step of nuclear winter — from the initial firestorm and the spread of its smoke, to agricultural and economic impacts. “We put all those pieces together for the first time,” says Robock.

The group looked at several scenarios. Those range from a US–Russia war involving much of the world’s nuclear arsenal, which would loft 150 million tonnes of soot into the atmosphere, down to the 100-warhead India–Pakistan conflict, which would generate 5 million tonnes of soot⁶. The soot turns out to be a key factor in how bad a nuclear winter would get; three years after the bombs explode, global temperatures would have plummeted by more than 10 °C in the first scenario — more than the cooling during the last ice age — but by a little more than 1 °C in the second.

Toon, Robock and their colleagues have used observations from major wildfires in British Columbia, Canada, in 2017 to estimate how high smoke from burning cities would rise into the atmosphere⁷. During the wildfires, sunlight heated the smoke and caused it to soar higher, and persist in the atmosphere longer, than scientists might otherwise expect. The same phenomenon might happen after a nuclear war, Robock says.

Raymond Jeanloz, a geophysicist and nuclear-weapons policy expert at the University of California, Berkeley, says that incorporating such estimates is a crucial step to understanding what would happen during a nuclear winter. “This is a great way of cross-checking the models,” he says.

Comparisons with giant wildfires could also help in resolving a controversy about the scale of the potential impacts. A team at Los Alamos National Laboratory in New Mexico argues that Robock’s group has overestimated how much soot burning cities would produce and how high the smoke would go⁸.

The Los Alamos group used its own models to simulate the climate impact of India and Pakistan setting off 100 Hiroshima-sized bombs. The scientists found that much less smoke would get into the upper atmosphere than Toon and Robock reported. With less soot to darken the skies, the Los Alamos team calculated a much milder change to the climate — and no nuclear winter.

The difference between the groups boils down to how they simulate the amount of fuel a firestorm consumes and how that fuel is converted into smoke. “After a nuclear weapon goes off, things are extremely complex,” says Jon Reisner, a physicist who leads the Los Alamos team. “We have the ability to model the source and we also understand the combustion process. I think we have a better feel about how much soot can potentially get produced.” Reisner is now also studying the Canadian wildfires, to see how well his models reproduce how much smoke gets into the atmosphere from an incinerating forest.

Robock and his colleagues have fired back in tit-for-tat journal responses⁹. Among other things, they say the Los Alamos team simulated burning of greener spaces rather than a densely populated city.

Dark seas

While that debate rages, Robock’s group has published results showing a wide variety of impacts from nuclear blasts.

That includes looking at ocean impacts, the first time this has been done, says team member Nicole Lovenduski, an oceanographer at the University of Colorado Boulder. When Toon first approached her to work on the project, she says, “I thought, ‘this sure seems like a bleak topic’.” But she was intrigued by how the research might unfold. She usually studies how oceans change in a gradually warming world, not the rapid cooling in a nuclear winter.