Global Science Report is a feature from the Center for the Study of Science, where we highlight one or two important new items in the scientific literature or the popular media. For broader and more technical perspectives, consult our monthly “Current Wisdom.”
Our last post was a brief run-through of some items of interest from the recent scientific literature that buck the popular alarmist meme that human-caused climate change is always “worse than we thought.” But as we said in that post, finding coverage of such results in the dinosaur media is a fool’s errand. Instead, it thrives on “worse than we thought” stories, despite their becoming a detriment to science itself.
Not to disappoint, headlines from the first major climate change story of the new year claim “Climate change models underestimate likely temperature rise, report shows,” and it’s clearly Worse Than We Thought. In its January 5 (Sunday) paper, the editorial board of the Washington Post points to the new results as a call for action on climate change.
The trumpeted results appear in a paper published in the January 2nd 2014 issue of Nature magazine by a team led by University of New South Wales professor Steven Sherwood and colleagues which claims that the earth’s equilibrium climate sensitivity—how much the global average surface temperature will rise as a result of a doubling of the atmospheric carbon dioxide content—is being underestimated by most climate models. Sherwood’s team finds “a most likely climate sensitivity of about 4°C, with a lower limit of about 3°C.”
Sherwood’s most likely value of 4°C is about twice the value arrived at by a rather largish collection of other research published during the past 2-3 years and lies very close to the top of the likely range (1.5°C to 4.5°C) given in the new report from the U.N.’s Intergovernmental Panel on Climate Change (IPCC).
While there are a host of reasons as to why our understanding of the true value of the climate sensitivity is little better constrained now that it was some 20+ years ago (it was given as 1.5°C to 4.5°C in the IPCC’s first report issued, almost a quarter-century ago), it is widely recognized that our understanding of the role of clouds in a changing climate is central to the issue.
In describing the why climate models have such different climate sensitivity values, the IPCC writes, in the 2013 edition of it’s science compendium,
Sherwood and colleague set out to see if they could help nail down the specific cloud processes involved in the model spread and to see if recent observations could help better understand which models were handling processes related to cloud behavior better than others.
The rate of vertical mixing in the lower atmosphere has a direct role in the formation of clouds. In a broad sense, according to the authors, the more vertical mixing that takes place in the lower atmosphere, the more drying that occurs in the lowest levels of the atmosphere, and therefore cloud amounts must decrease.
Sherwood and colleagues then compared the amount of mixing simulated by a collection of climate models with some observations of the mixing rate derived from weather balloon observations and other observation/model hybrids (called “reanalysis” products). They found that the climate models which most closely match the observations turned out to be the climate models with the highest climate sensitivity. Climate models with low sensitivities largely failed to contain the observations at all.
Based on this general finding—that climate models with a greater sensitivity to carbon dioxide increases produce a better match to observations of low level mixing rates—Sherwood and colleagues conclude that future global warming is going to progress much faster than is generally accepted.
This is the EEBE (“everything else being equal”) trap in big print.
While Sherwood et al., and press coverage of their paper, emphasize model comparisons with the “real world” they fail to show the “real world” comparison that makes the most sense—how do the climate model projections of global temperature changes compare with observations of real world temperature changes?
If they aren’t strongly related to vertical motion changes, then everything else is most decidedly not equal.
Our Figure 1 below shows the observed global surface temperature history from 1951-2013 compared with the temperature evolution projected by the collection of models used in the latest IPCC report. We broke the climate models down into two groups—those which have a climate sensitivity greater than 3.0°C (as suggested by Sherwood et al.) and those with a climate sensitivity less than 3.0°C. Figure 1 shows that while neither model subset does a very good job is capturing evolution of global temperature during the past 15-20 years (the period with the highest human carbon dioxide emissions), the high sensitivity models do substantially worse than the lower sensitivity models.
How in God’s getting-greener earth did the reviewer boffins at Nature miss this? (Hint: it messes up the meme.)
Figure 1. Observed global average temperature evolution, 1951-2013, as compiled by the U.K’s Hadley Center (black line), and the average temperature change projected by a collection of climate models used in the IPCC Fifth Assessment Report which have a climate sensitivity greater than 3.0°C (red line) and a collection of models with climate sensitivities less than 3.0°C (blue line) (climate model data source: Climate Explorer).[/caption]
Sherwood et al. prefer models that better match their observations in one variable, but the same models actually do worse in the big picture than do models which lack the apparent accuracy in the processes that Sherwood et al. describe.
It’s Worse Than We Thought all right—but for the climate models, not the real world. The result can only mean that there must still be even bigger problems with other model processes which must more than counteract the effects of the processes described by Sherwood et al. After all, the overall model collective is still warming the world much faster than it actually is.
Predictably, such a conclusion is absent from popular coverage of these results and from call-to-action editorials based upon them.
Sherwood, S. C., S. Bony, and J-D. Dufresne, 2014. Spread in model climate sensitivity traced to atmospheric convective mixing. Nature, 505,37-42, doi:10.1038/nature12829.