ClimatePrediction.net
ClimatePrediction.net ( CPDN ) is a distributed computing project of the University of Oxford, which calculates the climate of the next 50 to 100 years and thus provides information about the extent of global warming are to be made.
Climate models make high demands on the performance of the computer hardware used and were traditionally reserved for the fastest supercomputers and clusters. The rapid increase in performance of commercially available PC hardware makes it possible that such models can also be run on relatively inexpensive conventional computers in a reasonable time today. The ClimatePrediction.net client uses unused computing capacity of a computer in order to attribute a climate model.
Project objectives
In addition to the direct target to predict by a large number of simulations, such as the global climate will develop in the next 50 to 100 years, in all probability, would ClimatePrediction.net also make a contribution to improve climate models used. In the predictions of climate models, there are large uncertainties on the one hand result from the fact that still not all relationships and interactions are known in the air, and on the other hand are based on the values of the parameters used materialize. Many physical values is known to have a significant impact on the climate. Which range of values is, however, possible and useful, it can be found only by observations and a large number of tests. Here is where ClimatePrediction.net. Participating in the project computer count each with slightly different parameterized climate models and hence lead to different results.
History
ClimatePrediction.net goes back to a comment by Myles Allen, entitled "Do- it-yourself climate prediction " which was published in Nature in 1999. He pointed to the possibility to modern climate models such as HadCM3 run on reasonably modern home computers to narrow the parameter space can. At that time, SETI @ home, one of the first distributed computing projects, especially around the mouth. Someone would be able to tell his grandchildren later, that it was he who had calculated the most accurate prediction of the global mean temperature in 2050 on a 1650 dollar PC.
By the year 2000, the project partners had found, including the Universities of Reading and Oxford as well as the UK Met Office. Originally, the project went under the name Casino -21 and later renamed ClimatePrediction.net. On 12 September 2003, the public launch took place with a client for Windows. On 26 August 2004, the project to a new infrastructure was converted based on the developed by the SETI @ home BOINC - team. At the same clients for Linux and Mac OS were published for the first time.
Project History
Within the ClimatePrediction.net project, several experiments are carried out.
The first experiment served to limit the parameter space. For this purpose the HadSM3 model of the Hadley Centre was used, which is also used for commercial weather forecasts in the UK ( but then with higher resolution). This model allows a detailed simulation of atmospheric processes, but has only a simplified ocean. Each participating computer works with this model three successive phases: In the first phase, the model is calibrated. Basically it involves giving to calculate a heat flow value, with which the " dummy Ocean " can be maintained at a constant temperature. It also aims to determine whether at all the selected parameters allow a stable climate. If it turned out at that stage that the climate becomes unstable ( the earth thus figuratively speaking either becomes an ice ball or gets into cooking), the simulation is terminated at this point. Otherwise, two additional phases are run in which the ocean can change its temperature, but the CO2 content of the atmosphere is kept constant. In the second phase with a pre-industrial CO2 levels, the global mean temperature should remain ideally stable. In the third phase is working with double the CO2 content. Here, the climate should stabilize at a new stable level.
In 2005, the second experiment was launched. Here, a new climate model with fully simulated ocean is used (initially HadCM3 ). Within this experiment, the climate for the years 1950 to 2000 is calculated. For this period is known from records how developed the air. There are combinations of initial conditions and parameters are now being sought for was the prediction closest to the actual climate. The models are divided depending on their results in a ranking.
The third experiment, active since February 2006, is an actual prediction for the years 2000 to 2100th The results of the different model combinations are in turn weighted by the rankings. The result should be a relatively accurate prediction of what will happen in the next 100 years according to the rules of probability to the climate. In addition, special models are in between a number of selected computer yet been sent, as a simulation of the sulfur cycle.
Side projects
In addition to the main project CPDN operates two side projects that also use the BOINC infrastructure.
In August 2005 the Seasonal Attribution Project was started which has as its goal to explore the influence of human-induced climate change on severe local storms. As an example of such a local extreme weather event of extremely rainy autumn of 2000 is used over England and Wales, who had led in many rural areas in Britain to flooding.
The second side-project was the BBC Climate Change Experiment, which was launched in cooperation with the British broadcaster BBC, and should simulate the entire world climate for the period 1920-2080. First results of the simulations were planned for May 2006 as a BBC documentary for publication.
First results
On 27 January 2005 the CPDN team published the first results in the scientific journal Nature. The results of 1148 "stable" simulations indicate a global warming of at best 2 to a maximum of 11.5 ° C with a doubling of CO2 concentration in the atmosphere. This breadth is much greater than the climate sensitivity from 2 to 4.5 ° C, which were predicted by the models used in the fourth Intergovernmental Panel on Climate Change (IPCC ) report. Analysis of the results suggests that it is, not to be ruled out entirely, but very improbable outliers in the CPDN - models, which have very high calculated temperature changes. However, they point to what is possible in principle. With the weather far more sensitive than previously calculated, was indeed held by various trials for possible but CPDN is the first project in which this possibility actually displays the results of the test ensembles.
The following studies carried out could specify that an increase in global average temperature by more than 4.5 ° C is observed for a doubling of CO2 concentration with a probability of more than 5 %.
Future developments
For the future it is planned to include other, not originating from the Hadley Centre climate models in the project or to improve the existing models or to operate with a higher resolution. Of the currently lying on the clients result data only a small part is being actually scientifically evaluated. For the future it is planned to continue with these valuable results further evaluations that are not directly related to the ClimatePrediction.net project.