Study: Models Not To Blame In Failure To Predict Pause28.01.2015
28.01.2015 18:00 Age: 2 yrs
Computer climate models are not fundamentally flawed and their apparent failure to predict the ongoing pause in global warming is due to the inherently unpredictable behaviour of the climate system, according to research to be published in Nature tomorrow. This research has important implications for the issue of climate sensitivity and suggests that climate models are not overestimating the long term warming trend.
Click to enlarge. Fifteen-year temperature trends. The solid black line is the time series of the global mean surface temperature, plotted as a departure (anomaly) from a baseline period 196190. The dashed black line is a smooth fit to this series, representing the long-term warming rate. The blue and red lines are linear trends for each 15-year segment running over 185064, 185165,
, 19992013. Each 15-year segment is shown in red if the trend rises faster than the long-term warming rate in the same 15-year period and blue if it rises more slowly. Marotzke and Forster show that these 15-year trends are dominated by natural (free) variations. The free variations drive the 15-year trends above and below the long-term warming rate as they ride along with it. Courtesy: Nature.
Click to enlarge. Piers Forster: No evidence that climate models overestimate sensitivity. Courtesy: UNFCCC
Click to enlarge. See below for caption.
Click to enlarge. Observed and simulated time series of the anomalies in annually averaged global-mean surface temperature (GMST), from1900 to 2012. All anomalies are differences from the 19611990 temporal mean of each individual time series. GMST is the globally averaged merged surface temperature (2m height over land and surface temperature over the ocean). The figure shows single simulations for the CMIP5 models (thin lines), the multimodel ensemble mean (thick red line) and the HadCRUT427 observations (thick black line). All model results have been subsampled using the HadCRUT4 observational data mask. a, (top) 114 realizations from the CMIP5 archive, obtained with 36 different models. b, (bottom) Subset of 75 realizations with the 18 different models for which information on ERF is available. The two model ensembles are nearly indistinguishable. Courtesy: authors and Nature.
Unpredictable random fluctuations in the climate system are to blame for the apparent failure of computer climate models to simulate the so called pause in global warming, according to new research.
This study finds that computer simulations of the climate work and that there is no evidence from recent observations to support the suggestion that models systematically overestimate the response of the climate system to greenhouse gas warming, according to Piers Forster of the University of Leeds, co-author of a new paper describing this research.
The failure of computer climate models to predict the pause in global warming that has continued, on some measures, since the mid 1990s is a key criticism of climate science made by climate sceptics and has prompted a number of climate scientists to suggest that computer models may overestimate the warming impact of greenhouse gases.
Natural variability in the climate system can give rise to global surface temperature trends similar to the global warming pause and climate models can reproduce such trends but the chaotic nature of the climate makes it impossible for any one climate model to predict with certainty the exact timing of such a pause, Forster told reportingclimatescience.com. He explained that the ability of computer models to predict long term trends is much more reliable.
Here is the text of a press release issued by Nature regarding this story:
Jochem Marotzke and Piers Forster analyse simulated trends in global mean surface temperature from 1900 to 2010 and identify factors that explain the differences between simulated and observed climate trends. They find that at any one time, the observed warming might be at the upper or lower limit of simulated rates, but there is no evidence of a systematic bias in model process. 15-year trends are dominated by internal variability, while 62-year trends are more influenced by uncertainties in time-varying inputs to the models, such as volcanic eruptions.
End of Nature news release.
Forcing, feedback and internal variability in global temperature trends by Jochem Marotzke & Piers M. Forster published in Nature, doi:10.1038/nature14117
Read the abstract and get the paper when it goes live here.
Our story on the recent Duke University and Scripps Institution of Oceanography research here.