1 Detecting humanity’s fingerprint in the atmosphere on Sun Dec 01, 2013 11:18 pm
The June 12, 1991 eruption column from Mount Pinatubo taken from Clark Air Base. The eruption of Mount Pinatubo is one of the relatively few large volcanic eruptions to affect the Earth's atmosphere between 1979 and 2005.
As policymakers and armchair critics gear up for the impending release of the next report from the Intergovernmental Panel on Climate Change, the research community has been busy pushing out new results. One of the new papers that arrived this week is what's termed an attribution study, which attempts to understand the causes behind changing trends in the climate.
We recently covered the extensive efforts made to ensure that climate models accurately reflect the processes that drive our planet's climate. The new paper shows how the resulting models can be applied to understand what's actually happening to our planet. In this regard, the models are just like the ones that are used in astrophysics and plate tectonics: they help us understand the processes that drive events that we can't control in the lab, but on which we can gather data.
In this case, a large number of factors, from stochastic processes to natural cycles in the ocean currents, create short-term variability in the climate system—we can view those as noise. Superimposed on those factors are signals, significant changes that drive long-term changes in the climate. These are often termed forcings, since they push the climate into a new equilibrium state. Decades of study have identified forcings, such as changes in solar activity, volcanic eruptions, and alterations in the level of greenhouse gasses. An attribution study such as this simply looks at all of these factors in order to determine which of them are causing recent changes in the climate system.
Attributions work because the different forcings affect the atmosphere in distinct ways—in the language of the paper, they leave fingerprints. So, for example, the depletion of ozone in the stratosphere cools it, since ozone is a greenhouse gas that can trap heat in this layer of the atmosphere. In contrast, large volcanic eruptions add lots of material to the stratosphere; some of it absorbs light, leading to stratospheric heating. But volcanoes aren't simply the converse of ozone loss; some of those same particles reflect light, keeping it from reaching the lower atmosphere (called the troposphere), thereby cooling the lower atmosphere. These distinctive fingerprints allow us to attribute the changes we see to different forcings.
In this case, the authors focused on matching the satellite record, which provides a three-dimensional view of the atmosphere, with both depth and pole-to-pole details about the temperature. They compare this record to the output of a suite of climate models, run under various conditions. These included things like only solar forcings, only volcanic forcings, and all human influences (such as greenhouse gasses, aerosol emissions, and ozone depletion). They also did runs with only natural and only human influences, and then everything combined.
By tracking the trends for a large chunk of satellite data (January 1979 to December 2005), they were able to show that solar forcings were largely neutral (which makes sense, given that solar activity hasn't changed that much during the period). With the exception of Mount Pinatubo, there haven't been any major volcanic eruptions during the period, either. As a result, the net natural influences were pretty minor: a very slight cooling of the stratosphere outside the poles, and an equally slight warming of the lower atmosphere.
That's an absolutely awful match for the actual satellite data. The data show a very strong stratospheric cooling, extending somewhat into the upper troposphere. In contrast, the lower troposphere has warmed considerably, with a very strong bias toward the North Pole (in contrast, the air over Antarctica has only warmed slightly during that period).
That's a much, much better match to the model runs that used human influences alone. Those showed strong stratospheric cooling over both poles and weaker cooling elsewhere. It also showed lower atmospheric warming, also biased toward the Arctic. Combining all the natural and human influences increased the match further. Overall, the authors conclude that there's a clear human fingerprint on the atmosphere.
This isn't to say the match is perfect. The models underestimate the cooling of the stratosphere and overestimate the amount of warming in the lower atmosphere. There are also some spatial issues, as the real world data are more strongly biased toward Arctic warming than the output of the models.
Which raises the question of whether the human fingerprint is significant. Measured as a statistical fit, the answer is clearly yes. But the authors go further, quantifying the noise in the system as represented by shorter term variability. And they show that, with the exception of a short period around the eruption of Mount Pinatubo, the signal of human influences rises clearly above the noise of natural variability.
The authors conclude the paper by saying "Our results are robust to current uncertainties in models and observations and underscore the dominant role human activities have played in recent climate change." This isn't exactly new, and the new paper is largely an incremental improvement on past analyses of the same type. But, as the impending IPCC report release focuses the public's attention on climate change, the paper is a good reminder of the sort of work that led to the conclusions that will appear in that report.]