Wednesday, September 13, 2006

More on climate change and tropical ocean temperatures

There's yet another new study about the contribution of human-induced climate change to ocean warming in the regions of hurricane development ("tropical cyclogenesis regions"). This one, by Ben Santer and colleagues in the Proceedings of the National Academy of Sciences (check Sept 12), uses the results from a number of different climate model simulations conducted for the upcoming IPCC assessment to estimate the amount of warming caused by natural or background climate variability and by external (human) forcing.

Media coverage of observed climate warming, especially related to hurricanes, tends to suggest that are two possible, but distinct, causes: forcing from the increase in greenhouse gases, or natural climate variability. Critics or skeptics of climate change science are often interviewed and say the observed warming is due to natural variability. This claim sets up a straw man: an assumption that climate scientists think that observed warming is entirely due to greenhouse gases or that the models that climate scientists use do not consider the effect of natural variability.

Nothing could be further from the truth. Much of the energy in the climate change science community is devoted to using climate models to examine exactly this problem: what fraction of observed warming over the past century can be attributed to natural variability, or to the increase in greenhouse gas concentrations? This "climate change attribution" is done not just by running climate models with, and without, human forcings (GHG emissions) and then comparing the results to the historical temperature record, but by also analyzing "control runs" of the models. In a control run, the model is run for hundreds or thousands of "years" to estimate the background climate variability (in the model). The control run is contrasted with observations to gauge ability of the model to represent different "modes" of variability (ie. year-to-year, decadal, multi-decadal). With knowledge of the models' strengths and shortcomings, detailed statistical analysis can then used to tease out the role of natural variability and of greenhouse gas emissions in the observed warming.

If your work involves climate science, or you are just interested in the subject of climate change attribution or specifically in the hurricane issue, I recommend reading the Santer et al paper itself to get a sense of the strengths of this form of analysis. You can also read the nice summary of the subject on Realclimate.

2 comments:

tim said...

"The forcings that drive long-term climate change are not known with an accuracy sufficient to define future climate change." -- James Hansen, "Climate forcings in the Industrial era", PNAS, Vol. 95, Issue 22, 12753-12758, October 27, 1998.

Yeah that is the guy who is a father of modern GCC movement

Simon Donner said...

First, by 'define', Hansen was referring to the degree of warming, not whether future climate change will happen or not (with continued increases in GHGs).

Second, notice the paper is dated 1998. While there is still uncertainty, there's been immense advances in our understanding of the natural variability of the climate system, and the forcing caused by greenhouse gaes. Again, I'll suggest reading the Santer paper, and citations therein, or the post on Realclimate.