Climate-engineering research must have strong governance if it is to proceed safely, openly and responsibly1,2. But what this means in practice is not clear. The Stratospheric Particle Injection for Climate Engineering (SPICE) study demonstrates the difficult judgements involved. As chairman of the panel that supported decisions by the UK Engineering and Physical Sciences Research Council (EPSRC) as to whether and how this project should proceed (P.M.), and the architect of the project's governance process (R.O.), we draw lessons from these challenges.

In mid-September 2011, SPICE announced the go-ahead for the United Kingdom's first field trial of climate-engineering technology. SPICE aims to assess whether the injection of sulphur particles into the stratosphere would mimic the cooling effects of volcanic eruptions and provide a possible means to mitigate global warming. An equipment test — spraying water at a height of 1 kilometre — was proposed (see 'SPICE field trial'). No climate engineering would result from the test, but response to the announcement was dramatic, and the project was soon at the centre of a storm of criticism.

Careful review

On 26 September 2011, the EPSRC, one of the study's main funders, postponed the trial after a review. Later the same day, the council received a letter and open petition3, also sent to UK energy and climate-change secretary Chris Huhne and signed by more than 50 non-governmental organizations (NGOs) and civil-society organizations, demanding that the project be cancelled. The signatories saw the research as a first, unacceptable step towards a fix that would deflect political and scientific action away from reducing greenhouse-gas emissions. Others, by contrast, saw the research as urgently needed to find possible ways of coping with climate change4. The question at the heart of this debate was: should work in this controversial field proceed at all, and if so, under what conditions?

The strong feelings about the first test of SPICE's equipment show how important it is to have robust governance, and for scientists and funders to ensure that the public and other parties are consulted at the earliest opportunity. This is an unfamiliar and difficult process, but it is crucial for the evaluation of climate-engineering approaches.

SPICE was conceived in March 2010 at an EPSRC interdisciplinary workshop, at which researchers were invited to develop innovative geoengineering proposals. The project's funding incorporated field testing, but release of money was conditional upon it passing a 'stage-gate' review — a governance process in which funding for each phase of research and development is preceded by a decision point. To pass the review, SPICE scientists were required to reflect on the wider risks, uncertainties and impacts surrounding the test and the geoengineering technique to which it could lead — solar-radiation management.

On 15 June 2011, the stage-gate panel (including atmospheric scientists, engineers and social scientists, as well as an adviser to an environmental NGO) evaluated the SPICE team's response to five criteria for responsible innovation. These were that: the test-bed deployment was safe and principal risks had been identified, managed and deemed acceptable; the test-bed deployment was compliant with relevant regulations; the nature and purpose of SPICE would be clearly communicated to all relevant parties to inform and promote balanced discussion; future applications and impacts had been described, and mechanisms put in place to review these in the light of new information; and mechanisms had been identified to understand public and stakeholder views regarding the predicted applications and impacts.

Recognizing the efforts of the SPICE team, the panel concluded that although the first two criteria had been met, more was required on the remaining three. It asked the team to develop a revised communications plan to inform further public debate, a review of the risks and uncertainties of solar-radiation management — including social, ethical, legal and political dimensions — and a thorough process of engagement with stakeholders.

The test bed was delayed by EPSRC in September to allow the team to undertake these outstanding actions. When the panel reconvenes, it will independently assess a revised response; until then, the project remains under review.

Lessons learned

Aspects of SPICE's governance could have been improved. The framework should have been in place before the project's conception; the test date should not have been announced until the stage-gate criteria had been met; and the structures and resources to support the social research should have been in place earlier. Even now, the decision on whether to proceed will not be easy. There are few right or wrong answers to the many questions about climate engineering. But it is vital that we make space to listen to and discuss these questions, and that the debate transparently influences the decisions that are taken.

For geoengineering technology to progress, its developers must be mindful of wider impacts from the outset; and it must proceed under robust governance mechanisms. The SPICE responsible-innovation framework is one evolving approach to achieving it.