climateprediction.net

climateprediction.net (CPDN) is the world's largest climate modeling experiment, harnessing volunteer computing to run thousands of slightly different climate model simulations in parallel. While conventional climate modeling centers might run a single model or a handful of variations, CPDN runs tens of thousands of ensemble members — each with subtly different parameter settings — to explore the full range of plausible climate futures and quantify the inherent uncertainty in climate projections.

Based at the University of Oxford and led by the Department of Physics, the project uses variants of the Met Office Unified Model — the same model used by the UK's national weather service for operational forecasting. By systematically varying parameters that represent poorly understood physical processes (such as cloud formation, aerosol effects, and ocean heat transport), CPDN produces probability distributions of climate outcomes rather than single deterministic predictions. This probabilistic approach is essential for policy-making, as it tells decision-makers not just what the most likely outcome is, but how bad things could plausibly get.

The project has been a major contributor to the Intergovernmental Panel on Climate Change (IPCC) assessment reports, providing crucial data on climate sensitivity — the amount of warming expected from a doubling of atmospheric CO₂. CPDN's work on weather event attribution has also pioneered the emerging science of determining whether specific extreme weather events (heatwaves, floods, storms) have been made more likely by human-caused climate change.

CPDN has produced results published in Nature, Science, and other leading journals. The project's "Weather@home" sub-project allows regional climate modeling at high resolution, giving communities and governments detailed projections of how climate change will affect their specific regions.