Home
This Title All WIREs
WIREs RSS Feed
How to cite this WIREs title:
WIREs Clim Change
Impact Factor: 4.571

Ensemble modeling, uncertainty and robust predictions

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

Many studies of future climate change take an ensemble modeling approach in which simulations of future conditions are produced with multiple climate models (or model versions), rather than just one. These ensemble studies are of two main types—perturbed‐physics and multimodel—which investigate different sources of uncertainty about future climate change. Increasingly, methods are being applied which assign probabilities to future changes in climate on the basis of the set of projections (the ensemble) produced in a perturbed‐physics or multimodel study. This has prompted debate over both the appropriate interpretation of ensembles as well as how best to communicate uncertainty about future climate change to decision makers; such communication is a primary impetus for ensemble studies. The intuition persists that agreement among ensemble members about the extent of future climate change warrants increased confidence in the projected changes, but in practice the significance of this robustness is difficult to gauge. Priority topics for future research include how to design ensemble studies that take better account of structural uncertainty, how to weight ensemble members and how to improve the process by which ensemble studies are synthesized with other information in expert assessments. WIREs Clim Change 2013, 4:213–223. doi: 10.1002/wcc.220

Conflict of interest: The authors have declared no conflicts of interest for this article.

Figure 1.

Projected global warming under the A1B emission scenario, relative to 1961–1990 mean temperature. Results are from more than 2000 ensemble members from more than 800 model versions in the climateprediction.net BBC perturbed‐physics study. Darker shading indicates better fit with observational data. The black line indicates results from observational data. The dark blue lines indicate a ‘likely’ range (66% confidence interval) from the ensemble (see Ref 19 for details). Red bars indicate the ‘likely’ ranges (66% confidence ranges) specified by IPCC experts18 for around 2050 and 2080. (Reprinted with permission from Ref 24. Copyright 2012 Nature Publishing Group).

[ Normal View | Magnified View ]
Figure 2.

Changes in precipitation (%) at the 10, 50 and 90% probability levels in different UK regions by the 2050s as estimated by UKCP09. Changes are relative to the 1961‐1990 mean for each region, under a medium emission scenario. Wider ranges indicate the lowest and highest values of change seen across three emission scenarios and all three probability levels. (Reprinted with permission from Ref 25. Copyright 2009 Crown Copyright).

[ Normal View | Magnified View ]
Figure 3.

Projections and uncertainties for global mean temperature change in 2090–2099 relative to the 1980–1999 average, under several emission scenarios and from numerous studies. The 5–95% ranges from probabilistic studies are indicated by solid vertical lines. The IPCC ‘likely’ ranges are indicated by wide gray bars. (Reprinted with permission from Ref 18. Copyright 2007 Cambridge University Press).

[ Normal View | Magnified View ]

Related Articles

WIREs at 4ICESM

Browse by Topic

Climate Models and Modeling
Climate, History, Society, Culture > Ideas and Knowledge
Climate Models and Modeling > Knowledge Generation with Models

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts