Climateprediction.net
[[File:{{#setmainimage:Clim.jpg}}|alt=climateprediction.net logo|center|frameless]]
climateprediction.net (CPDN) is a volunteer distributed computing project operating on the BOINC infrastructure, designed to improve understanding of climate change by running large ensembles of global climate model simulations. It is jointly coordinated by the University of Oxford and partner institutions including the UK Met Office and multiple UK research centres [1].
The project is one of the largest climate modelling experiments ever conducted in terms of computational scale and ensemble size, producing statistically significant distributions of future climate states rather than single deterministic projections [2].
Overview
climateprediction.net applies distributed computing to climate science by running thousands of perturbed climate simulations on volunteer computers worldwide. Each simulation varies uncertain physical parameters within plausible ranges, enabling probabilistic climate prediction through ensemble statistics.
This approach addresses a fundamental limitation in traditional climate modelling: uncertainty quantification in long-term climate projections.
Scientific Rationale
The project is based on the recognition that climate models contain structural and parametric uncertainty. Instead of producing a single forecast, CPDN explores a probability distribution of outcomes by varying:
- Cloud parameterizations
- Ocean heat uptake
- Aerosol forcing
- Radiative transfer parameters
This ensemble-based method allows estimation of climate sensitivity ranges consistent with observed climate data [3].
History
The concept for CPDN was introduced by Myles Allen in 1999 in the Nature commentary “Do-it-yourself climate prediction” [4].
Key milestones include:
- 1999 – Concept proposal for distributed climate modelling
- 2002 – UK research council funding secured
- 2003 – Project launch
- 2004 – Migration to BOINC platform
- 2006 – BBC Climate Change Experiment
- 2009 – BBC experiment concluded and data archived
Scientific Projects
CPDN runs multiple research streams grouped into active and completed experiments.
Current Projects
- DOCILE (Drives Of Change In mid-Latitude weather Events)
- TNC (The Nature Conservancy partnership studies)
- GOTHAM (Globally Observed Teleconnections and their role and representation in Hierarchies of Atmospheric Models)
- AFLAME (Attributing Amazon Forest fires from Land-use Alteration and Meteorological Extremes)
- HIASA (High Impact Atmospheric Simulation Analysis)
- EMBARK (ensemble-based climate risk modelling initiative)
- National Trust climate resilience modelling
These projects focus on attribution science, extreme event analysis, and climate risk quantification in regional systems [5].
Completed Projects
These completed studies provided foundational datasets for understanding climate variability, extreme weather attribution, and ensemble climate uncertainty.
Model Systems
CPDN has used multiple generations of climate models:
HadCM3-based Models
The HadCM3 coupled atmosphere-ocean general circulation model formed the basis of early CPDN experiments. It allowed long-duration simulations at relatively low computational cost.
Weather@Home System
An extension enabling regional downscaling of global climate projections, widely used in extreme weather attribution studies.
OpenIFS Models
Recent CPDN simulations use ECMWF OpenIFS configurations, enabling higher-resolution global modeling. Some configurations require multicore execution and up to tens of gigabytes of RAM per job [8].
BBC Climate Change Experiment
The BBC Climate Change Experiment (2006–2009) was a major public engagement initiative involving over 120,000 volunteers.
Participants ran simulations covering:
- 20th century climate (1920–2000)
- Future scenarios (2000–2080)
The experiment demonstrated the feasibility of large-scale public participation in climate modelling and contributed to early climate risk communication efforts [9].
Computational Methodology
Each CPDN work unit typically includes:
- Model initialization and spin-up
- Control simulation under baseline parameters
- Perturbed simulation with modified physical constants
Outputs are aggregated across thousands of independent runs to derive probabilistic climate response functions [3].
.jpg)
Scientific Output
CPDN data has contributed to numerous peer-reviewed studies in climate science, including:
- Stainforth et al. (2005) — climate sensitivity uncertainty distributions
- Murphy et al. (2004) — probabilistic climate modelling framework
- Sexton et al. (2012) — regional climate uncertainty analysis
- Philip et al. (2019) — extreme weather attribution studies
The BOINC scientific publication archive also lists CPDN-derived contributions across multiple climate research domains [10].
Project Team and Sponsors
The project is led by:
- Prof Myles R. Allen
- Andy Bowery
- Dr Neven S. Fuckar
- Dr Sihan Li
- Dr Friederike E. L. Otto
- Prof David Wallom
CPDN is based at the University of Oxford, primarily within:
- Environmental Change Institute
- Oxford e-Research Centre
- Atmospheric, Oceanic and Planetary Physics group
Impact
climateprediction.net has had significant scientific and societal impact:
- One of the largest ensemble climate datasets ever produced
- Early demonstration of probabilistic climate prediction
- Major contribution to extreme weather attribution science
- Public engagement through BBC collaboration
- Influence on IPCC uncertainty framing methodologies
Scientific Publications
A full list of CPDN-related publications is maintained at: https://climateprediction.net/publications/ [12]
Key peer-reviewed works include publications in Nature, Journal of Climate, and Environmental Research Letters.
See also
External links
References
- ↑ https://www.cpdn.org/about.php
- ↑ https://en.wikipedia.org/wiki/Climateprediction.net
- ↑ 3.0 3.1 Stainforth et al., Nature (2005)
- ↑ Allen, M. (1999), Nature
- ↑ https://climateprediction.net/projects/
- ↑ https://climateprediction.net/projects/completed-project/completed-weatherhome-projects/
- ↑ https://climateprediction.net/projects/completed-project/hadcm3-and-other-models/
- ↑ https://www.reddit.com/r/BOINC/comments/1mhrgkw
- ↑ https://www.bbc.co.uk/sn/climateexperiment
- ↑ https://boinc.berkeley.edu/pubs.php
- ↑ https://climateprediction.net/people
- ↑ https://climateprediction.net/publications/