climateprediction.net (CPDN) is a research project that uses Internet-connected computers to do research in climate science. You can participate by downloading and running a free program on your computer.

CPDN is based at the University of Oxford.

Join climateprediction.net (CPDN)

• Read our rules and policies
• This project uses BOINC. If you're already running BOINC, select Add Project. If not, download BOINC.
• When prompted, enter: https://climateprediction.net/
• If you're running a command-line version of BOINC, create an account first.
• If you have any problems, get help here.
  Join

  Already joined? Log in.

Happy New Year and update for 2026
We wish all the volunteers a Happy New Year and thank everyone for their contributions to CPDN to date.

CPDN looks forward to a busy 2026. We have recently been tidying our databases and systems. We will also be implementing badges soon, similar to other BOINC based projects.

There is more production work to come on the PAREST OpenIFS project (linux based). A small test batch went out at the end of last year. A larger production batch is now planned, though no release date as yet. There are also other projects being discussed. Plus there is continued development of our backend and workunit control software. It's likely we'll be sending out some test batches as beta applications in the near future as part of this development.

Thank you to all our volunteer number crunchers, our moderators and all our forum users for contributing to bring issues to our attention and solve issues for others.

The CPDN team.
19 Jan 2026, 11:05:43 UTC · Discuss

CPDN project downtime from Tuesday 3:30pm (UK time)
I hope you all had a good festive break. Just to let you know that the CPDN project will be going offline from tomorrow (Tuesday) at 3:30pm (UK time). This is in order to conduct some essential backend maintenance work. It is expected that this work may take several days to complete. Thank you in advance for your patience.

Update: this work has now been completed.
5 Jan 2026, 12:44:48 UTC · Discuss

Project PAREST - New batches
Project description: Exploring the role of circulation biases in limiting seasonal predictability

Recent work has shown that operational seasonal forecast models have deficient responses to predictable forcing (e.g. ENSO) on seasonal timescales. These appear to be linked to persistent circulation biases in the atmosphere – similar issues are also found in coupled climate models used to make projections on longer timescales. In this OpenIFS@home project we will create perturbed model parameter ensembles to generate a range of different atmospheric states and explore which atmospheric circulation biases act to limit circulation signals in the extratropics.

This project is led by scientists at the University of Reading, Meteorology Department.

For more details of the batches, please see the CPDN Forums.

Technical details:
App: OpenIFS 43r3 Multi-core Linear grid Tl319L91
Linux only.
4 physical cores required.
Expected task run-time: 2-7 days depending on hardware and boinc client settings.
Checkpoint files for this app are 4.5Gb.

This application has a peak memory usage of 26Gb. Only one task will be sent to each host at a time.

To contribute to this project you must enable the 'OpenIFS Tl319' option on the CPDN Project Preferences page on your CPDN web account and have enough free resource in the boinc client for the task.

Please ensure you have 'Leave non-GPU applications in memory whilst suspended' enabled. This prevents a restart from checkpoint files which is costly.

Please raise any issues/questions on our forums.
10 Nov 2025, 13:44:06 UTC · Discuss

Project DEPICT-SNOW: New batches
Project description:
While global mean snowfall is projected to decrease due to global warming, it remains unclear how the extreme snowfall will respond to global warming. This UK Research and Innovaton (UKRI) funded MSCA project aims to study the response of extreme snowfall in the upcoming decades to future global warming. The project will run three large-ensemble seasonal weather forecast experiments including one control and two sensitivity experiments to accomplish the project objectives. It is anticipated that the results will significantly improve our understanding of future extreme snowfall change.

Technical details:
App: OpenIFS 43r3 Multi-core Linear grid Tl319L91
Linux only.
4 physical cores required (not user changeable).
Expected task run-time: 4-6 days depending on hardware.
Checkpoint files for this app are 4.5Gb.
Each set of model output files is 22Mb.

Please note to contribute to this batch you must enable the 'OpenIFS Tl319' option on the CPDN Project Preferences page on your CPDN web account and have enough free resource in the boinc client for the task.

Please also ensure you have 'Leave non-GPU applications in memory whilst suspended' enabled. This prevents a restart from checkpoint files which is costly.

This application has a peak memory usage of 26Gb. Only one task will be sent to each host at a time.

Please raise any issues/questions on our forums.
13 Aug 2025, 15:25:43 UTC · Discuss

Release of a new multicore configuration of the OpenIFS application
A new multicore configuration of the OpenIFS application, known as Tl319, is being released by CPDN. This is the first time CPDN has deployed a multicore app and we believe the first time BOINC has been used for large memory applications.

As this configuration has a peak total memory requirement of 26Gb, we have introduced a new option under the 'CPDN Project Preferences' found in 'Your account' which must be selected in order to receive workunits using this configuration. Please ensure your machine has enough resources assigned to the boinc client even if this option is selected. CPDN will soon release batches for several projects using this configuration. For further information and discussion on this new app and batches, please use the forums.
22 Jul 2025, 10:29:24 UTC · Discuss

News is available as an RSS feed