Can Models Replicate Observed Temperature Trends Over the Past Decade? (Invited)
Abstract
The last 10 years of observational temperature records have received considerable public attention. It has been claimed that the relatively muted surface warming trend of the past decade constitutes 'evidence of absence' of a human effect on climate, and that model simulations of historical climate change are incapable of replicating the observed temperature behavior. We examine global-scale temperature changes in the troposphere and at the ocean surface. Rather than comparing temperature changes over a single time period, as in previous work, we select multiple analysis timescales, and then compare all possible simulated and observed trends on each timescale. We find that model simulations with anthropogenic and natural external forcing frequently produce 10-year periods with little or no warming of the surface and troposphere. On timescales longer than 20 years, model and observational datasets invariably show global-scale warming of the ocean and free troposphere. The observed warming on multi-decadal timescales is unlikely to be explained by model-based estimates of natural internal variability. There is no indication that model variability estimates are systematically smaller than observed. As our results clearly illustrate, signal-to-noise ratios are small on 10-year timescales. It is therefore not meaningful to search for a slowly-evolving human-caused climate change signal in a single, noisy 10-year period of observational temperature records.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2010
- Bibcode:
- 2010AGUFMED41D..02S
- Keywords:
-
- 1610 GLOBAL CHANGE / Atmosphere;
- 1626 GLOBAL CHANGE / Global climate models