Tall and stony-faced, with a long and bitter history of fighting for democracy, Tongan leader Akilisi Pōhiva is not someone you'd expect to break down in tears at an intergovernmental summit.
At a meeting of Pacific leaders last month in the tiny island nation of Tuvalu, other attendees said Pōhiva was overcome with emotion as he tried to secure Australian Prime Minister Scott Morrison's support for a more forceful approach to tackling the climate crisis.
He was not the only one frustrated by Morrison's apparent lack of concern for the danger that smaller Pacific nations face as sea levels rise. "You are concerned with saving your economy," said Tuvalu's Prime Minister Enele Sopoaga to Morrison. "I'm concerned about saving my people." Fijian leader Frank Bainimarama later summarized the meeting as settling "for the status quo."
Climate change is not affecting the world equally or at the same pace. And the forum's failure to agree on stronger action was a pointed reminder that the countries most immediately endangered may not be able to count on others for a quick solution.
Increasingly, countries like Tonga and Tuvalu may consider going it alone by turning to geo-engineering -- a "deliberate large-scale intervention in the Earth's climate system" -- that can require less international consensus.
Geoengineering has long been seen as something out of science fiction, or a dangerous distraction from more practical solutions to climate change. But for the countries that are running out of time, it offers a last throw of the dice before they sink beneath the waves.
"(For these countries) there might be a risk of harm from doing this stuff, but you've got to balance this against the certain risks of unabated climate change," said Jeff McGee, a senior lecturer in climate change law at the University of Tasmania.
What is geoengineering?
There are two main categories of geoengineering: carbon dioxide removal, and more radical methods which seek to reflect sunlight and thus cool the Earth.
Some efforts to scrub carbon from the atmosphere are underway, but many methods -- apart from planting loads of trees -- remain energy intensive, making them difficult to do at scale while remaining carbon neutral.
"Pulling carbon from the atmosphere is very desirable if we could achieve it at scale," said Andy Parker, co-author of the Royal Society report and project director at the Solar Radiation Management Governance Initiative (SRMGI). "But based on the technologies we have at the moment it would be very slow and expensive."
For a long time, many in the climate community saw geoengineering as a potential distraction from much needed controls on emissions, providing an excuse for continued inaction, said Jan McDonald, a professor of environmental law at the University of Tasmania. However that inaction occurred regardless, and now geoengineering is beginning to seem like a necessary step as well as emissions mitigation.
"Back in the very early 1990s, nobody wanted to talk about adaptation because they were worried that would take away from the mitigation imperative. Everyone is willing to talk about it now," McDonald said. "We're on the cusp of that era now in terms of solar radiation management."
Solar radiation management, or SRM, is an attractive option for many because -- in theory -- it could provide a cost-effective way of reducing global temperatures and staving off some of the worst effects of climate change while the international community seeks a better solution.
"There have been big, big problems in decarbonizing enough to meet the 2 degree target," said John Moore, chief scientist at Beijing Normal University's College of Global Change and Earth System Science, one of the world's largest geoengineering research programs.
"People see (SRM) as a potential solution to avoid these overshoots in temperature where you might get tipping points."
How SRM works
All SRM methods focus on reflecting inbound sunlight back into space, reducing the amount of heat absorbed by the Earth and lowering global temperatures, or at least halting their rise.
The most effective solution appears to be stratospheric sulphate injection, whereby large amounts of reflective particles are sprayed into the upper atmosphere.
"That's not doable by anybody at the moment, the planes don't exist in sufficient numbers and sizes to do that effectively," Moore said.
Yet while research into exactly how to get the particles into the stratosphere cheaply is still underway, the science behind the method is solid. That's because we can trace similar drops in global temperatures during major volcano eruptions, when ash pours into the upper atmosphere, reflecting large amounts of sunlight.
The eruption of Mount Tambora in Indonesia in 1815 led to the "year without a summer," during which global temperatures temporarily dropped by as much as 3C and crops failed across Europe and North America. Not all eruptions are so drastic in their effect -- when Mount Pinatubo in the Philippines erupted in 1991 it caused global temperatures to drop between 0.1 and 0.6C during the almost two years the ash remained in the upper atmosphere.
While the theory behind spraying particles into the stratosphere may be somewhat strong, the consequences of doing this deliberately for an extended period are still unclear.
"If you want to quickly reduce global temperatures, then the only known method currently is solar geoengineering," said Parker, the SRMGI expert. "The problem is we don't know what the impacts would be."
According to a 2017 report by the Overseas Development Initiative, a UK-based think tank, "potential downsides of SRM include changes in regional weather patterns that could lead to droughts in Africa and Asia, damage to the ozone layer, continued ocean acidification, impacts on natural ecosystems and agricultural crops, impacts on tropospheric chemistry, diminished radiation for solar power, and the risk of human error."
More immediately practical and less risky methods include spraying seawater into the lower atmosphere to brighten clouds and enable them to reflect more sunlight, as well as thinning cirrus clouds -- which trap heat and prevent it from escaping to space.
Should developing countries take action?
So far, efforts to tackle the climate crisis have largely focused on international agreements to reduce emissions, ones that have been -- by and large -- profoundly unsuccessful in doing so.
While the most effective way of lowering global temperatures is to reduce emissions, this is something that requires a global response, unlike some geoengineering methods which could be carried out -- at least in theory -- by a single country or group of countries.
"Unfortunately, the most environmentally responsible way is also the most politically difficult," Ken Caldeira, a climate scientist at the Carnegie Institution for Science, said last year. "If there's a leader of a country whose people are starving, and they think by injecting some particles in the stratosphere they can feed their people and alleviate suffering, the political pressure to do that is going to be intense."
As research continues into geoengineering processes, they can be expected to become cheaper and more practical. Some have even suggested that research itself may be too risky, because it leads down the slippery slope of taking action.
But while the global community would arguably be irresponsible in focusing on radical, untested geoengineering approaches rather than finally acting to tackle emissions, individual nations -- faced with the reality of a worsening crisis -- may feel differently.
Such proposals "could have widespread impacts beyond the territory of a country that decides to use them, impacting regional precipitation and monsoon patterns," and would raise obligations under the "no harm" principle in international law, said Kerryn Brent, a legal expert at the University of Tasmania.
However, this is "not a rule to absolutely prevent harm, but a rule to take the absolute necessary tests," she said. "If they're ticking all the boxes and harm still occurs that doesn't necessarily mean they're in breach of the rule."
Her colleague McGee said that "in a crisis situation, where there's a necessity to do something ... it's probably going to be hard to get countries to come to court before they do it or even after they do it."
This could significantly limit the ability of the international community to rein in any unilateral action.
Moore, the Beijing Normal University expert, suggested developing nations "might want to use it as a threat."
"To say to countries that you have to help us financially to deal with the climate change you caused by your emissions, or we will try this geoengineering," he said.
However if the threat of climate disaster isn't enough to spark action, the less certain risk of geoengineering may not be any more motivating for countries responsible for the majority of emissions.
One major emitter is taking action in this regard: China. The country is both a major polluter and at major potential risk from climate change, and its leaders have invested heavily in geoengineering research alongside renewable energy and slowly moving away from China's dependency on coal.
"They don't want to be a leader in unilaterally doing it, they want to be a leader in understanding how it works and in a sense representing the developing world in geoengineering," Moore said.
China is more open to unusual solutions because of the country's own history of radically shaping its environment, he added.
"China is China as a powerful country historically because widespread irrigation allowed the population of China to triple," Moore said. "They're used to that idea that man and nature are working hand in hand."
Speaking at an event last year, Kim Stanley Robinson, the science fiction author and environmentalist, praised the way talking about geoengineering can open up mental space to consider other radical solutions.
"I like the idea of geoengineering because I think we're already doing it," Robinson said. "And once we admit to it and begin to try to take control of it for good, we are in a more honest relationship with the planet."
For many countries already experiencing the effects of how we have changed the planet for the worse, the drive to just try something -- even if it's dangerous -- to attempt to make it better may become increasingly strong.
"We are clearly reaching a moment of crossing one of those tipping points into a completely different planet, a jungle planet," Robinson said. "At that point, then you say we need to put the dust in the air."