Verdict: The impact on the ocean environment is highly unpredictable. Also it is seen to be slow and costly. It can at best play a "moderate role in carbon sequestration."
Chemical air capture with carbon removal: Some scientists have been developing prototype devices that soak up ambient airborne carbon and fix it in various chemical or mineral compounds, which must then be transported and stored.
Verdict: It is seen to have lots of potential, but storage venues for the captured carbon must be identified. It's slow and expensive, but low-risk.
Cloud whitening: Cloud cover over the oceans can be increased by spraying tiny saltwater particles into the sky. This idea has received a lot of public attention and doesn't depend on sci-fi technology.
Verdict: Cloud-whitening efforts, carried out by special flotillas, can be turned off quickly in case of problems. The impact on marine ecosystems is not known.
Increasing Earth's surface reflectivity: Making Earth's surface more reflective could bring down temperatures. Measures include painting roofs, roads and other man-made surfaces white. Other ideas: large-scale desert reflectors.
Verdict: Surface whitening is straightforward, but it would take time to paint a sufficient area to achieve a result. The Royal Society estimates that all of the world's desert areas would have to be covered with reflectors to achieve a meaningful effect.
Stratospheric aerosols: Inspired by cooling after the 1991 eruption of Mount Pinatubo, some experts have proposed jettisoning various aerosols into the stratosphere to replicate the impact of a major volcanic explosion.
Verdict: The quantity of aerosols needed to reduce temperatures is estimated to weigh less than a 10th of the cargo shipped by air each year. But scientists don't yet understand the impact on the ozone layer and other climatic systems.
Space-based solar reflectors: Reflective or refractive surfaces - everything from orbiting mirrors to dust particles or trillions of metallic disks - could be deployed in Earth's orbit.
Verdict: Besides the cost, technical complexity and time involved (decades to deploy), the obvious question is what happens if something goes wrong after all that stuff is sent into orbit.
Source: Royal Society