Assessing the reliability of climate model ensembles using the Last Glacial Maximum and Mid Holocene
Abstract
The principal obstacle to assessing the reliability of projections from the multi-model ensembles is the long term nature of the forecast, which means that it is not possible to regularly confront the forecast system with new observational data. Thus, while encouraging results have been obtained assessing the CMIP3 ensemble on the global scale with respect to modern climate, this does not necessarily imply that the ensembles will remain reliable when run for future scenarios. We should, therefore, further assess and update our confidence in the ensemble by quantitative assessment of model performance for other climate states, that were not used during the model development process. For CMIP3 (and also CMIP5) the Last Glacial Maximum (LGM) and Mid Holocene qualify as a suitable climates from this point of view. Furthermore, the quality and quantity of available data at these periods is superior to other more distant periods with several gridded datasets having become available in recent years. We have analysed two LGM multi-model ensembles (the PMIP1 slab-ocean and PMIP2 coupled-ocean ensembles) with respect to the MARGO sea surface temperature dataset through the rank histogram method. Encouraging results, which add support to the hypothesis that the ensembles may reliably represent climate change, were obtained for both ensembles. The analysis also highlighted systematic differences between the two ensembles, consistent with the different model complexities. On the other hand, a single-model ensemble based on MIROC3.2 was found to be much less reliable. New results from continuing analyses incorporating terrestrial pollen datasets for the LGM and mid Holocene, alternative statistical assessments, and model runs from CMIP5 (as available) will also be presented.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFMGC53A..08H
- Keywords:
-
- 1626 GLOBAL CHANGE / Global climate models;
- 4928 PALEOCEANOGRAPHY / Global climate models;
- 1627 GLOBAL CHANGE / Coupled models of the climate system