Skip to content

New Paper: Models Continue To Show Too Much Recent Warming

About a month or so ago, Science magazine published a paper by Susan Solomon and colleagues that concluded that aerosols in the upper atmosphere that were unaccounted for in earlier estimations, have, over the past 10 years or so, acted to offset about 0.07°C of warming that would have otherwise occurred. In other words, we shouldn’t be so hard on the climate models for failing to anticipate the dearth of warming over the past 10-15 years.

Or should we?

It turns out, that what the paper really says, is that the amount of global warming that should have occurred over the past 10-15 years (that is, if the climate models were getting things correct) is about 25% greater than the model-expected warming from the combination of increases in greenhouse gases and lower atmospheric pollution alone. Which means that the observed warming during this same time—which has been close to nil—is even harder to explain and makes the models look even worse.

But, of course, that is not at all how the results were spun to the press.

First, a bit of background. When large volcanic eruptions occur, they can inject particulates into the upper atmosphere (stratosphere) where they are largely unaffected by the weather (which occurs in the lower level of the atmosphere called the troposphere) and thus can remain suspended for months to years, rather than days to weeks (the typical time that a particular aerosols particle remains suspended in the troposphere). These suspended particulates in the stratosphere primarily act as dirty little mirrors and reflect away some of the incoming radiation from the sun. This leads to a general cooling of the earth’s average temperature. The more aerosols in the stratosphere, the more cooling. Fairly major volcanic eruptions (such as El Chichon in 1982 and Mt. Pinatubo in 1991) can lead to perhaps a degree or so of cooling of the earth’s average surface temperature for maybe a year or so. The cooling effect attenuates as the aerosols eventually are removed from the stratosphere.

Generally large volcanic eruptions don’t go undetected and so several records of the “aerosol optical depth” (that is, how much stuff in the stratosphere) have been compiled that extend back for several centuries. Such records can be used as input to climate models to estimate how the earth’s temperature reacted to the eruptions, and when used alongside of greenhouse gas emissions, tropospheric aerosol emissions, and solar variability, the models are expected (by some optimistic folks) to reproduce the earth’s temperature evolution over the past century (or even longer). The veracity of the projections using such a scheme depend not only on how well climate model can model the physical interactions of these climate drivers, but also the degree to which other drivers of the earth’s climate (of which there are many) influence the climate over the same time periods (of course “false veracity” can be achieved by “tuning” the models to reproduce the observed temperature record, even when the physics in the models is incomplete, or even wrong).

What Solomon and colleagues found was that some reflective aerosols were getting in the stratosphere even without the benefit of strong volcanic eruptions—and thus have been exerting an unmodelled climate influence. Potential candidates responsible for the stratospheric injections of such aerosols include not-so-large volcanic eruptions and particulates from coal burning that somehow got lofted into the upper atmosphere.

Solomon and colleagues attempted to find out just how large a climate impact that temporal variations in this “background” stratospheric aerosol may be having. They did this by running a climate model several times with several different input scenarios. All the scenarios included the same derivation of the known histories of greenhouse gases and tropospheric aerosols, but differed on the evolution of stratospheric aerosols. One scenario included no stratospheric aerosols, another included only stratospheric aerosols from large volcanoes, and the third included the stratospheric aerosols from large volcanoes as well as the newly identified “background” aerosols.

Figure 1 shows what Solomon et al.’s climate model projected for changes in the earth’s average surface temperature from 1980 through 2010 using their various scenarios. The green line shows how the climate model projected that temperatures would rise just considering greenhouse gases and lower atmospheric aerosol pollution. The black line shows what happens when the influence of stratospheric aerosols from large volcanoes are included in the model, and the black line with the blue extension shows what happened when the new findings about the levels of “background” stratospheric aerosols are included as well. The two dips in the black/blue line are the resulting cooling from the eruptions of El Chichon and Mt. Pinatubo.

Figure 1. Climate model projections of the evolution of the global average temperature from 1980 through 2010 based on several different scenarios of stratospheric aerosols. The scenarios include: no stratospheric aerosol forcing (solid green line); only background aerosol forcing with no volcanoes (dash-dotted green line); stratospheric aerosol forcing transitioning to no stratospheric aerosol forcing after 2000 (black line); stratospheric aerosol forcing from 1998, then including “background: aerosols (black followed by blue lines)(adapted from Solomon et al., 2011; post 2010 results omitted).

Now the spin. From the results depicted in Figure 1, Solomon and colleagues conclude that “the observed increase in stratospheric aerosol since the late 1990s caused a global cooling of about –0.07°C as compared with a case in which near-zero radiative forcing is assumed after year 2000, as in the forcing data sets often used in global climate models. Figure [1] shows that stratospheric aerosol changes have caused recent warming rates to be slower than they otherwise would have been.”

True enough. It seems that the inclusion of the “background” stratospheric aerosols does act to reduce the rate of warming from what it “otherwise would have been” from 1998-2010 (the blue line in Figure 1 rises slower than the black line in Figure 1 from 1998 through 2010).

So where’s the spin, then?

Check out the first paragraph of the press release that was issued by the National Oceanic and Atmospheric Administration (NOAA) to announce the findings of Solomon et al.:

A recent increase in the abundance of particles high in the atmosphere has offset about a third of the current climate warming influence of carbon dioxide (CO2) change during the past decade, according to a new study led by NOAA and published today in the online edition of Science.

Huh? It did no such thing! For that comparison, you have to compare the blue line with the green line in Figure 1.

The variation of stratospheric aerosols has acted to increase the rate of global warming during the past decade over and above that expected from carbon dioxide (and lower atmospheric aerosols). The press release has spun the results 180 degrees from what they actually are. And in doing so, has sparked a bunch of media coverage proclaiming that we now know part of the reason why the earth’s average temperature has risen so little during the past 10-15 years despite rapidly rising atmospheric levels of greenhouse gases.

Figure 2 shows what we mean. In Figure 2, we show only the portion of Figure 1 from 1998 through 2010 and we add trend lines through the model projections for each scenario (we omit the “background only scenario” (the dashed green line in Figure 1) as the trend is virtually the same as the trend in the “no stratospheric aerosols” scenario). Notice that the trend in the no stratopsherice aerosol scenario (the one with only greenhouse gases and tropospheric aerosols; green line) over the period 1998-2010 is projected to be 0.17°C/decade. When stratospheric aerosols from large volcanic eruptions are included, the trend increases to 0.27°C/decade and when large volcanic eruptions and “background” aerosols are included, the trend becomes 0.21°C/dec. In other words, even when Solomon’s “background” stratospheric aerosols are included, the model-projected rate of global warming is about 25% greater than had there been no volcanic eruptions/background aerosols in the first place. This is because in Solomon et al.’s climate model, the earth’s surface temperature is still recovering from the combined cooling effects of the combination of the El Chichon and Mt. Pinatubo eruptions—a recovery which as hastened the recent rate of warming.

Figure 2. Same as Figure 1 except only showing the period from 1998-2010. trends have been added (adapted from Solomon et al., 2011).

So instead of identifying why there has been less global warming over the past 10-15 years than climate models have projected, Solomon et al. have found that climate models have missed the mark even further—for had they properly included the effect of volcanoes and background stratospheric aerosols, they would have projected even a greater rate of warming than they already do. And since observations show that there has been relatively little, if any, warming over this same time period, the results from Solomon et al. means that climate models are doing worse than even the model lovers have realized.

But, you wouldn’t know this unless you read the pages of World Climate Report!


Solomon, S., et al., 2011. The persistently variable “background” stratospheric aerosol layer and global climate change. Science, 333, 866-870, DOI: 10.1126/science.1206027.

World Climate Report, 6 September 2011