Single realizations vs. means of ensemble projections

[h/t Quark Soup via AFTIC]

Marlo Lewis of CEI (yes, again) writing over at Planet Gore has a post purporting that not only has there been “no warming in the 21st century” (all ~8 years of it? Gosh!), but that “no climate model predicted it”. His evidence is an abysmal graphic allegedly provided to him by John Christy (who along with Roy Spencer produces UAH’s MSU troposphere temperature record):

I say “allegedly” because although together Spencer and Christy have gotten some things wrong before, and Spencer has made some outlandish, unscientific claims on his own (no need to revisit those here), so far I am not aware of any behavior or statement on Christy’s part that suggests he is capable of this degree of misrepresentation or error. The chart presented is ostensibly Figure TS26 from the IPCC AR4 Technical Summary combined with HadCRUT (3v?) and UAH Lower Troposphere temperatures. The first thing that should jump immediately out at anyone seeing this should be what are being purported to be the “model predictions” here.

The averaged ensemble projections for the SRES emissions scenarios are not predictions of individual realizations of the climate (i.e. how any given year plays out)- many “runs” or realizations of multiple models are averaged not only to reduce potential model biases, but also to reduce internal variability, in order to emphasize the “forced” component of the climate (IPCC AR4 WG1 10.3). The noisy natural amplifying or effects of “weather” are therefore smoothed out, not explicitly “predicted” for these various emissions scenarios over time- the mean projections are intentionally not meant to capture the kind of variability that any individual realization, be it reality or of a specific prediction or forecast, would attempt to capture.

That being said, does that mean that “no climate model” realization that went into the mean projections of the forced component of climate over time reflect anything like what Lewis and presumably Christy claim? All one has to do is look at a single emissions scenario (A1B for example) and examine the individual runs to illustrate how specious Lewis’ claim and ‘Christy’s’ graphic are:

IPCC AR4 A1B Individual Realizations

Image courtesy of RealClimate

If all of this sounds a bit familiar, it’s likely that you remember the absolute butchering of Keenlyside et al., 2008 across the denialosphere as well as by a certain “non-skeptical heretic”.

Presumably someone like Christy would be a bit more aware of the distinction between individual realizations of climate and projections of the forced component of climate over time. Which begs the question- is Marlo Lewis misrepresenting John Christy? If not, why hasn’t Christy objected publicly to being used as a part of such a ham-handed attempt to discredit the modeling community, and what exactly was Christy attempting to convey with such a graphic in the first place? Attempting to splice the average of the different emissions scenario runs and two temperature records together with no indication of error bars for either doesn’t seem to serve any legitimate purpose I can think of off hand.

Any ideas?


3 responses to “Single realizations vs. means of ensemble projections

  1. > Possibly related posts:
    > (automatically generated)
    > * Apparently not…

    Ya know, it might be worthwhile to consider deleting that “automatic” line from the template, since to a casual reader it looks like you’re recommending …. well, look at his blogroll …

    [I’m not sure how to go about that in WordPress even if I wanted to. That doesn’t seem to be a widget or design option. -TB]

  2. I know it’s removable — Tamino’s posts lack the autogenerated ones, and he uses WordPress.

    On topic, I think you’re overthinking this to some extent. At a glance, without looking deeply, this is sort of like plotting two curves and playing with the Y-axes such that they perfectly overlap — showing agreement or disagreement where there is none. The run-of-the-mill inactivist isn’t interested in deep fact checking or data analysis (obviously; only the more sophisticated leaders of the pack bother with that, and often to mislead). The quick, take-home at-a-glance message is “the models don’t agree with reality!” — more ‘evidence’ for the inactivists to use to decry climate science. This could also explain why Christy is cited as the source — q.v. pseudoepigraphy.

    All of this is rampant speculation, of course, but it’s all I’ve got at the moment.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s