Scientists at Columbia University have warned that attempts to cool the planet by injecting sunlight-reflecting particles into the upper atmosphere could be far more complicated and risky than most research suggests.

The process, known as stratospheric aerosol injection (SAI), has been proposed as a way to temporarily offset global warming by dispersing fine particles that reflect solar radiation back into space.

Hundreds of studies have modelled how SAI might work, but the Columbia team says real-world conditions would make it much harder to control.

“Even when simulations are sophisticated, they’re idealised,” said Prof V. Faye McNeill, an atmospheric chemist at Columbia’s Climate School. “When you start to consider how far we are from that ideal situation, it reveals a lot of uncertainty.”

Writing in Scientific Reports, the researchers examined physical, economic and political barriers to large-scale SAI. They say where and when particles are released matters greatly, with potential side effects including disruption of monsoon systems and changes to global wind patterns.

Most models focus on sulfate aerosols similar to those produced by volcanic eruptions such as Mount Pinatubo in 1991, which cooled global temperatures but also damaged the ozone layer and reduced rainfall in South Asia.

Proposed alternatives such as calcium carbonate, alumina and titania may be less harmful but face practical limits of supply, cost and technical performance. At the microscopic size required, particles tend to clump together, reducing their effectiveness.

“It’s all about risk trade-offs,” said Prof Gernot Wagner, a co-author and climate economist at Columbia Business School. “Given the messy realities, it isn’t going to happen the way most models suggest.”