On 7 May 2025, the UK Advanced Research and Invention Agency (ARIA) announced the list of projects to be funded under its £60 million solar geoengineering research initiative. Readers of this blog will already be familiar with solar geoengineering (also known as Solar Radiation Modification, or SRM for short), why it is a false response to the climate crisis, and why its implementation would put the Earth’s environment in a dangerously precarious state. There has been a disturbing upsurge in funding for SRM research and development in the past few years, almost exclusively by private philanthropy such as the Simons Foundation or Quadrature Climate Fund, but this is different. The ARIA programme constitutes a new escalation in the push to normalize this dangerous set of technologies.
The ARIA programme is the first time that a major government has put this scale of money into an SRM programme. Worse, unlike essentially all previous programmes funded at a significant level, it puts some of this money into outdoor field trials of the technology. That crosses a red line that has mostly held up until now. A limited amount of Australian government money, funnelled controversially through the Great Barrier Reef foundation (additional commentary here), has gone into cloud-brightening trials off the Australian coast. But the ARIA programme is a step change in the level and prominence of SRM development funding. To our knowledge, it is the first programme to allow field experiments in the stratosphere for development of stratospheric aerosol injection technology. The normalization of outdoor field experiments threatens to unleash a flood of funding for SRM technology development from governments worldwide. It is the height of folly to unleash SRM field experiments into a world devoid of either national or international governance of such experiments. ARIA describes these experiments as “small scale and controlled,” but once experiments become normalized, who is to say what “small scale” really means? The only governance is that voluntarily imposed by ARIA itself; it is a case of scientists being accountable to scientists.
The ARIA programme is built on false premises: it claims it will determine whether SRM can be “safely deployed” to head off climate tipping points. As one of us noted, it can do no such thing. In fact, no “small scale” experiments can determine what would happen in a sustained global scale deployment of any of the technologies proposed, or how the complex and imperfectly understood Earth system will respond over the multi-century (or more) time frame over which SRM interventions would need to be maintained. Nothing short of a full-scale deployment could reveal that, and even then it could take decades to determine the full effects on the planet of such deployment. This is one of the many fatal drawbacks of SRM as a response to the climate crises. The “tipping point” framing is also deeply flawed. Tipping points are poorly defined and involve aspects of the climate system that cannot be understood by “small scale” field experiments, and cannot be reliably simulated with current models, or those that could be developed over the time frame of ARIA funding. (Despite this, ARIA is placing another tranche of £60m into researching tipping points). What the ARIA programe can do, however, is to develop technological capabilities that some other entity could deploy — whether “safely” or not.
What is ARIA?
ARIA was the brainchild of Dominic Cummings, erstwhile sidekick of disgraced UK Prime Minister Boris Johnson, and architect of Brexit. ARIA is patterned on the US DARPA family of “moonshot” style freewheeling research organizations. The legislation that created it was passed in 2022, in the waning days of the Boris Johnson government. This legislation guaranteed that ARIA would remain in existence at least through 2032, with funding for the first five years set at an eye-watering £800 million. The appointment of its first director, Ilon Gur, was announced by Kwasi Kwarteng (then Secretary of State for Business, Energy and Industrial Strategy) on 19 July of that year. Gur was formerly a director of the US ARPA-E programme, aimed at commercializing new energy technologies. Similarly to the DARPA family, the operational conception of ARIA is to appoint programme directors and give them each a pot of money to essentially spend as they see fit, with carte blanche to formulate programmes and establish procedures for handing out grants. Directors have generally chosen to engage in some limited forms of consultation with the research community in formulating programmes, but this is largely discretionary. The legislation creating ARIA expressly exempts it from Freedom of Information requests, and most of the usual transparent contracting regulations. ARIA works in darkness, with the only light let out being what the directors choose to disclose. This in itself makes ARIA singularly unsuitable as a funding vehicle for such a contentious technology as SRM, with such enormous potentially adverse implications for society.
Besides that, if ARPA-E is the model (as suggested by Ilon Gur’s appointment) it is doubly inappropriate for SRM research, since ARPA-E was not aimed at addressing fundamental scientific questions, but rather bridging the gap between the development of novel technology and commercializing it. The presumption in ARPA-E is that if something could be done at a profit it should be done. That is far from the case for SRM. The appearance of the word “Invention” in the title of ARIA echoes the ARPA-E model, and indeed much of the SRM work they have funded is of a technology development nature, some of it even involving for-profit companies. In a commentary in Nature Climate Change, one of us has called into question the suitability of the ARIA “moonshot” model for climate change research of any kind.
The ARIA SRM programme emerged rapidly, with little public scrutiny. It had its genesis with the appointment of Mark Symes (an electrochemist with no background in climate science) as Programme Director in September 2023. Symes is the ARIA director who conceived the SRM project, under the heading of the “Climate Cooling” opportunity space. The first media attention we are aware of was in January 2024 with a story from the UK’s Research Professional News, in which one of us pointed out that Symes’ ideas were “naive and ill-informed”. A New York Times piece in September 2024 then sounded the alert that ARIA intended to fund outdoor SRM experimentation. Soon after that, a call for proposals was issued, with a closing date in December 2024. Despite efforts by many of us to gain media attention in the UK and the attention of the new Labour government, it was not until March 2025 that one of us managed to place an article on the ARIA programme in The Guardian. By then, it was too late to rein in the programme. Since the announcement of the projects funded, media attention has ramped up, but the U.K. government has been slow to take notice.
What is ARIA Funding?
Now let’s go into the specifics of some of the SRM projects ARIA has funded. Three grants, totalling £9.1 million (or up to £14.1 million with triggered matching funds) go to Marine Cloud Brightening (MCB) studies. Marine Cloud Brightening attempts to increase reflection of sunlight to space by near-surface marine clouds, through various schemes intended to reduce the cloud droplet size (which increases their reflectivity, since a lot of small drops reflect more light than the same mass distributed over larger drops). All of these grants are primarily aimed at technology development. One of them will fund expansion of the field trials ongoing over the Great Barrier Reef, led by Australia’s Southern Cross University. Another goes to the University of Manchester and five other organizations to develop sprayer technology for MCB interventions, initially with indoor experiments but with an eventual eye to moving the experiments outdoors. One of the partners is a for-profit firm Archipelago Technologies, underscoring ARIA’s potential commercialization ambition. The third grant goes to University of Reading and two other organizations, for development of electrostatic methods for MCB. This project, too, involves corporate partners.
The stratospheric outdoor experiment funding of £5.5 million goes to a project led by Prof. Hugh Hunt, of Cambridge University’s hubristically named Centre for Climate Repair, which has gotten its hands on a considerable array of funding for developing geoengineering technologies. While this project does not actually release particles or their precursors into the stratosphere, it still crosses a red line in that it trials geoengineering technologies (in this case particle design) in the actual stratosphere. This moves us further along the slippery slope to deployment without coming close to answering any of the most important questions about what deployment would actually do to our planet.
In a similar vein is the so-called Sea Ice Thickening project, a £9.9 million project led again by Cambridge’s Centre for Climate Repair. This involves two corporate partners (Real Ice and Arctic Reflections ) who aim to ultimately make money out of sea ice thickening. The idea of the scheme is to install pumps on Arctic sea ice floes to spray sea water on the ice surface in winter, where it will (hopefully) freeze and lead to thicker sea ice that can better survive the summer melt season. A good and unbiased summary of the concept can be found here, which is the source of most of the information discussed in the following.
Sea Ice Thickening
Sea ice thickening comes under the heading of “SRM” because sea ice reflects sunlight (though the global effect is muted because it only does so in the limited Arctic and Antarctic regions, and then only in the seasons where there is significant daylight). A bigger effect of sea ice is to insulate the atmosphere from the relatively warm ocean in winter, which is not strongly sensitive to sea ice thickness. The serious flaw in the scheme is that each pump would affect only a small area of sea ice (which becomes yet smaller if the floe fragments). It has been estimated very crudely that it would take millions of pumps at a cost of £50 billion per year (assuming that they could be adequately maintained) to increase sea ice thickness by just 1 meter over 10% of the Arctic. Further, there are likely environmental side effects, since thick multi-year sea ice in the Arctic does not occur via surface freezing. Ice initially thickens by freezing onto the bottom, but this process slows greatly as ice thickens. Thick multi-year ice instead thickens by sea ice dynamics, which compresses ice and piles floes one atop the other. Even in the unlikely event that it proved feasible at scale, this artificial thickening would have side effects; given that sea water rejects dense brine when it freezes, this thickening would increase brine rejection into the Arctic ocean.
But there are additional problems with the scheme, not least of which is maintenance of the pumps, and conflicts with nations that might not want thick ice floes drifting into their shipping lanes, or indigenous peoples who might prefer real solutions to the climate crisis (i.e. decarbonization) to techno-fixes that are disruptive but soon become ineffective. To avoid a termination shock from rapid melting, sea ice thickening (like all solar geoengineering methods) would need to be maintained essentially forever, but in the case of sea ice thickening even that (very) theoretical possibility is foreclosed, since it has been estimated that even optimal pumping strategies (if deployable) would not prevent the ultimate disappearance of Arctic summer sea ice under likely warming scenarios.
It has been argued that the huge cost of sea ice thickening is nonetheless worthwhile in comparison with the economic costs of sea ice loss, but this argument neglects the fact that sea ice will be lost eventually anyway. Moreover, there is the issue of how sea ice thickening’s cost would stack up against other ways the same amount of money could be deployed. While under operation, it has been estimated that sea ice thickening could perhaps offset 0.08 Watts per square meter of the roughly 4 Watts per square meter warming effect of doubling carbon dioxide. An investment of a similar amount of money (which in the end is sure to exceed the estimated £50 billion annually) in deployment of technologies like heat pumps, which reduce carbon dioxide emissions, would have an enduring (rather than temporary) effect and which would ultimately exceed what could be gained by a temporary fix on sea ice thickness.
Poor Judgement
Some of the projects ARIA has funded are almost equally troubling, even though they don’t involve outdoor experimentation. For example, £445,000 is being given to Cornell University for work designing future stratospheric aerosol injection experiments. Presumably, one would not spend money designing these things if one did not expect to eventually carry them out! Also, £400,000 is being given to a project led by the Planetary Sunshade Foundation to do engineering design work on spaceborne sunshades that would act to reduce the amount of solar radiation reaching the Earth. Like the sea ice thickening project, it is unlikely to come to pass given how much ‘cost-to-launch’ would have to come down in order to make spaceborne sunshades remotely feasible. But it is again indicative of a certain lack of judgement on the part of ARIA.
ARIA funding is not limited to the UK, and indeed most of the funded outdoor projects either go directly to foreign jurisdictions (as in the Australian field project) or to projects that involve activities outside the UK jurisdiction (e.g. sea ice thickening of necessity must conduct experiments outside the UK since the UK has no sea ice to thicken in its territorial waters). This puts many of the activities out of reach of the UK public that is funding them, and may well object to them being carried out. It raises difficult issues of which entities will be in charge of governance and compliance, and ARIA’s ability to enforce even its own mandated governance. While the jurisdictions involved in the projects funded so far have generally reasonable democratic governments, the general principle that ARIA can fund projects anywhere raises the temptation to shop for jurisdictions that limit public engagement or oversight.
A common fallacy in the justification for all of these SRM projects is that they “buy time” while fossil fuel emissions are being phased out, as if the SRM could be abandoned once that happens. This framing ignores some of the most fundamental and best understood aspects of climate physics. Because carbon dioxide is removed only very slowly by the oceans and by geological processes, the heating effect of the carbon dioxide we emit persists for millennia even after we stop emitting carbon dioxide. In essence, once carbon dioxide is emitted, it cannot be unemitted, barring development of removal technologies at a scale that are nowhere near on the horizon. The Cambridge Centre for Climate Repair is not the only SRM booster propagating this false framing, but they really ought to know better. Likewise for the Australian federal science foundation, CSIRO, which has made similar claims. CSIRO argues (optimistically) that MCB could buy 20 years of life for the Great Barrier Reef corals — but what happens then? This false “buying time” framing also appears frequently in media coverage of SRM.
Rein In ARIA
ARIA’s mission is “high risk high reward” research, and that’s good up to a point, but the “risk” shouldn’t involve the risk of developing technologies that imperil the whole environment which we rely on for survival. How much more damage will ARIA do if it is allowed to run out to 2032? Will we see yet more expansion of dangerous SRM trials? How much more of the UK public’s money is going to be squandered as ARIA goes into its second half? It is time for the Labour Government to sit up and take notice of this cuckoo’s egg planted in their midst by the former government, and rein it in.
—
Professor Raymond T. Pierrehumbert, University of Oxford
Professor Mike Hulme, University of Cambridge
Professor Peter Newell, University of Sussex
Professor Andy Stirling, University of Sussex