Details of Award
NERC Reference : NE/C515747/1
Climateprediction.net: a practical platform for ensemble earth system modelling.
Grant Award
- Principal Investigator:
- Professor MR Allen, University of Oxford, Oxford Physics
- Grant held at:
- University of Oxford, Oxford Physics
- Science Area:
- Terrestrial
- Marine
- Atmospheric
- Overall Classification:
- Atmospheric
- ENRIs:
- Global Change
- Environmental Risks and Hazards
- Science Topics:
- Palaeoenvironments
- Regional & Extreme Weather
- Climate & Climate Change
- Abstract:
- The success of the climateprediction.net project, funded in the first NERC e-Science round, has demonstrated the potential of public resource distributed computing (PRDC) allied to grid-based data analysis and management for large ensemble simulation with even the most complex environmental models. Since launch in September 2003, climateprediction.net has recruited over 90,000 people in 130 countries and completed over 85,000 45-year simulations with a climate-resolution general circulation model. Successful migration to an open source platform for generic distributed computing solutions has consolidated our computing base. Results have confirmed the need for large ensembles to explore uncertainties in Earth System Models (ESMs), showing highly non-linear responses to parameter perturbations and a much wider range of sensitivities to increasing carbon dioxide than suggested by conventional modelling studies. The overarching scientific objective of this project remains quantifying uncertainty in forecasts of 21st century climate change. Specifically, we will be launching a 'flagship' coupled perturbed-physics ensemble simulation of 1950-2100 in 2005, coinciding with a concerted effort to introduce climateprediction.net into school and college science teaching. This timing will take advantage of the high political interest in this issue at the G8 and EU summits. Significant new challenges for the coupled experiment include a probabilistic treatment of non-greenhouse forcings, efficient representation of the information content of forcing fields and model states, and management of 'serially distributed' jobs being run on a succession of machines. Following recommendations from NERC, we are proposing a twelve month extension of the original two-year project on a pro-rata level-funding basis. Our aim is to launch and support the coupled model experiment. The unique requirements of ESMs, in particular the large volumes of data generated and the need to re-analyze data in the light of initial results, means that the PRDC platform needs to be integrated much more closely into conventional Data-Grid technologies than was the case for earlier distributed computing projects. We are working closely with our partners developing the Berkeley Open Infrastructure for Network Computing (BOINC) and NERC Data-Grid on this integration.
- Period of Award:
- 1 Jul 2005 - 31 Oct 2010
- Value:
- £283,628 Lead Split Award
Authorised funds only
- NERC Reference:
- NE/C515747/1
- Grant Stage:
- Completed
- Scheme:
- Directed Pre FEC
- Grant Status:
- Closed
- Programme:
- E-Science
This grant award has a total value of £283,628
FDAB - Financial Details (Award breakdown by headings)
| Total - Staff | Total - T&S | Total - Other Costs | Total - Equipment | Total - Indirect Costs |
|---|---|---|---|---|
| £180,119 | £7,094 | £10,034 | £3,525 | £82,855 |
If you need further help, please read the user guide.