Head in a Cloud

troposphere and stratosphere meet blogosphere

Head in a Cloud random header image

Stratospheric Effects in IPCC Projections?

June 28th, 2007 by tanya · 4 Comments

Recently an interesting article found its way to my desk. It is titled ‘How Will the Stratosphere Affect Climate Change?’ and is in the June 15th 2007 issue of Science Magazine (Volume 316, Page 1576). This article discusses the importance of stratospheric effects in tropospheric climate models considered by the Intergovernmental Panel on Climate Change (IPCC). The stratosphere (10-50km) can affect the energy balance in the troposphere through the ozone layer. In addition, changes in stratospheric circulations affect weather and climate in the troposphere. The question the authors pose is: Why is it so difficult to include stratospheric effects in IPCC projections?

The guts of the article elaborate on the connection between the stratosphere and troposphere, then use that foundation to emphasize the importance of including stratospheric effects in models. A summary of the main points are given below.
1). Greenhouse Gases (including ozone) can heat or cool the atmosphere depending on the balance between absorption and emission. This balance depends on altitude and temperature.
2). Overall cooling in stratosphere due to carbon dioxide and ozone depletion (ozone is primary culprit in lower stratosphere). Lower stratosphere radiative changes are mainly latitude dependent, thus cooling at poles and warming in tropics.
3). Latitudinal dependence = change in north-south temperature gradient, thus a change in the lower stratospheric wind structure.
4). A change in wind structure will modify atmospheric Rossby waves (which propagate from the troposphere into the stratosphere). These changes in turn affect weather and climate at the Earth’s surface.

Then the article gives some examples of connections between the stratosphere and troposphere to further drive the point home. Next models and their stratospheric limitations are discussed. Again, a summary is given below.
1). Coupled atmosphere-ocean models – many include radiative effects of ozone depletion and ozone depleting substances but do not include changes in the ozone layer or the dynamics of troposphere/stratosphere coupling.
2). IPCC models – most have a fixed stratospheric ozone forcing constant, thus dynamical responses to stratospheric radiative changes are not likely captured.
3). Climate models with well-represented stratospheres – accurately account for stratospheric circulations changes due to climate change but fail to correctly propagate these variations downwards into the troposphere. Most damp out tropospheric responses by using prescribed ocean-surface temperatures.
4). Some coupled chemistry-climate models can simulate ozone changes and how that couples to climate change. According to these models ozone recovery will be accelerated because of the stratosphere cooling due to increasing greenhouse gases. A cooler atmosphere will slow down chemical reactions which destroy ozone. Pre-1980 levels should be reached by the middle of this century and become thicker beyond 2050 as the stratosphere cools.

Now we have a list of ways the stratosphere influences the troposphere, thus should be included in climate models. So the stratosphere is changed by and changes the meridional temperature gradient, which then affect ozone levels and circulations in the stratosphere. Eventually these changes propagate down into the troposphere, where they are not always accurately accounted for.

Coupling models that represent Earth’s different spheres (bio, hydro, atmo, etc …. ) is important, but lofty. The difficulty in such a feat is immense and the scientific community is inching toward that everyday, but limits do exists due to computing power. Highlights given in this article of model inadequacies is helpful, but I imagine those involved in the various model developments would gladly expand, if computationally possible. As to the models used by the IPCC and their helpfulness. First, are there better out there, I am unsure. It seems there as a model gains features, say better chemistry, other bells and whistles are lost due to, again, computing constraints. Am I wrong in this?

Lastly, the authors do point out that a complete inclusion of the stratosphere into tropospheric models would not significantly alter average surface warming but could substantially influence the evolution of polar climate. Part of me wants to stop and say ‘wait, wouldn’t an inclusion of the stratosphere change projected averages in a noticeable way’. This statement could be unwarranted and a result of my lack of understanding in the area of modeling, but it is my gut reaction to the statement.

Either way, the stratosphere is important to tropospheric climate and is a challenge the scientific community is correcting as quickly as possible. I know I am sitting on the edge of my sit!

Tags: climate · stratosphere

4 responses so far ↓

  • 1 Robb // Jun 30, 2007 at 11:13 am

    I may not fully understand what the authors meant by “Lower stratosphere radiative changes are mainly latitude dependent, thus cooling at poles and warming in tropics.” I’m not sure I’ve seen warming in the tropical stratosphere in any observations, am I wrong here?

    Currently observation and models are not in agreement in the understanding of atmospheric amplification (more warming in the troposphere than surface) in the tropics or in the MSU era for globally averaged data as pointed out by CCSP “Temperature trends in the lower atmosphere: Steps for understanding and reconciling differences” [Karl et al., 2006]

    This is due to the understanding of temporal changes in temperature, forcings etc profiles not being well known. This leads me to wonder why the authors would be so confident in knowing what the outcome would actually be if a complete inclusion of stratospheric changes were in the tropospheric models. This challenge and movement are, to me, the most necessary next step in understanding how and why the atmosphere is changing.

    Robb

  • 2 Tanya Phillips // Jul 5, 2007 at 9:47 am

    Currently there aren’t any observations, at least that I know of, which show a warming in the tropics. Thus the authors should have been careful with that statement. There are some models that predict a warming in the tropics and cooling at the poles, but no observations. One paper they cited shows a warming in the tropics in their model output. However, they do state that current (up to 2005) observations do not show this, so they must be missing some mechanism. According to them this could be due to the indirect effect of ozone depletion overpowering the direct effect of halocarbens, but as ozone recovers a warming could be seen in the tropics.
    So, a warming in the tropics could be seen in the future, but is not currently occur. They weren’t to clear on this in the paper and needed to be since they use it to highlight the importance of including stratospheric effects in IPCC projections.
    The authors do seem pretty confident on the outcome of complete inclusion of stratospheric changes in tropospheric models, but it is a broad statement they are making. So I am going to assume that they don’t it to be taken it as gospel. However, this is in Science Magazine (read by the general public) and they should be careful.
    Thanks for pointing these out.

  • 3 Hank Roberts // Jul 23, 2007 at 1:55 pm

    Possibly relevant (just reading and speculating)
    I noticed in the most recent Geophysical Abstracts a note that the location — latitude — of emissions of sulfate from coal plants affects what becomes of it.

    http://www.agu.org/pubs/crossref/2007/2006GL028668.shtml

  • 4 Tanya Phillips // Jul 24, 2007 at 12:25 pm

    Looked at the agu paper you listed. Its relevance is indirect since the authors of the posted ‘Stratosphere’ article did not specifically mention sulfate of sulfur, but talked about aerosol and greenhouse gases in general.
    Emission location for such a gas is important and should be incorporate into models, if it is not already. I imagine there are issues with incorporating too much detail in a model.
    Enjoyed the article, thanks for the link.

Leave a Comment

*