Nice Rorschach blot, Mr. Pielke.

I agree wholeheartedly with statements 1, 2 and 3. As for 4, tell me what “this” means.

If it means that we have no way to ascertain in advance what the temperature will precisely be, I`d say Ms. Greene-Raine is right – besides the effects of greenhouse gases, we don`t know what people will do over 50 – 100 years to further alter the environment, and even with perfect knowledge of future human behavior can`t precisely know the output of the tremendously complicated climate system, and don`t know whether it might be interrupted by exogenous variables.

But if it means that she thinks we don`t know that pumping GHGs into the atmosphere without finding ways to remove them (so that concentrations remain at today`s levels or increase) will, other things being equal, lead to higher mean global temperatures, then I`d say she is sadly misinformed.

Thanks for the exchange. Just one last point for now:

You write: “it is obvious that one learns nothing about the reliability of a climate model at 100 years from the reliability of a weather model at 7 days.”

From the standpoint of common sense, one actually learns a lot, best captured by Niels Bohr, “Prediction is difficult, especially about the future.”

For more on this than you’d probably ever want to see, have a look at this volume, which has chapters on weather and climate prediction, and 8 other case studies:

Sarewitz, D., R.A. Pielke, Jr., and R. Byerly, Jr., (eds.) 2000: Prediction: Science, decision making and the future of nature, Island Press, Washington, DC.
http://sciencepolicy.colorado.edu/about_us/meet_us/roger_pielke/prediction_book/

Thanks!

Coby Beck responded, “Whatever else it is, controversial, contrary to consensus, etc, it is really just disappointingly ignorant.”

Well, Coby, I agree with Nancy Greene Raine. So Nancy Greene Raines–and I—say that it’s “a bit of a stretch” to say that “in 50 or 100 years, the temperature is going to do this.” And you say it’s not a “bit of a stretch,” and we’re both ignorant for saying it is.

Fortunately, the dispute is resolvable (though only on a pretty long time scale). Why don’t you predict the 3-year average surface temperature anomaly, centered about 2010, 2020, 2030, 2040, 2050, 2060, 2070, 2080, 2090, and 2100?

Here is the location for NASA GISS surface temperature anomalies for the last ~120 years. To give you an example of what I mean, the anomalies for 2003, 2004 and 2005 were 0.67, 0.60 and 0.77 degrees Celsius, respectively. So the 3-year average centered around 2004 was 0.68 degrees Celsius. (Note: The anomalies are based around a temperature of 14 deg C, so the estimated average surface temperature for the 3 years centered around 2004 was 14.68 deg C.)

http://data.giss.nasa.gov/gistemp/tabledata/GLB.Ts.txt

You should be able to predict the anomalies for 2010, 2020, 2030, 2040, 2050, 2060, 2070, 2080, 2090, and 2100 to within less than +/- 1.0 degree Celsius (roughly the difference between coldest part of the Little Ice Age and the present), if your predictions aren’t really a “bit of a stretch.”

Or…do you agree that, “To say in 50 or 100 years, the temperature is going to do this, is a bit of a stretch”?

“I agree that climate modelling also has elements of the first kind, but it is fundamentally a boundary value problem so the analogy still holds”,

recent research clearly shows that climate prediction on yearly, decadal, and century time scales is an initial value problem. I will have a weblog posting on this issue on Climate Science in the next few days on this subject, but have extracted part of my posting below.

” Climate models include weather processes as a subset of the model… to characterize climate as a boundary value problem ignores peer reviewed papers which illustrate that climate predition is very much an initial value problem. Just one example is

Claussen, M., C. Kubatzki, V. Brovkin, A. Ganopolski, P. Hoelzmann, H.-J. Pachur, Simulation of an abrupt change in Saharan vegetation in the mid-Holocene, Geophys. Res. Lett., 26(14), 2037-2040, 10.1029/1999GL900494, 1999.

Further examples are discussed in

Rial, J., R.A. Pielke Sr., M. Beniston, M. Claussen, J. Canadell, P. Cox, H. Held, N. de Noblet-Ducoudre, R. Prinn, J. Reynolds, and J.D. Salas, 2004: Nonlinearities, feedbacks and critical thresholds within the Earth’s climate system. Climatic Change, 65, 11-38.

and

Pielke, R.A., 1998: Climate prediction as an initial value problem. Bull. Amer. Meteor. Soc., 79, 2743-2746.”

The mindset of assuming multi-decadal global model simulations is a boundary value problem is based on treating the Earth’s surface as a precribed boundary for use in atmospheric general circulation models. In reality, there are a wide range on nonlinear interfacial fluxes between the different components of the climate system such that the climate system response very much depends on the initial conditions throughout the climate system.

One last comment on this topic. I skimmed through the Palmer article you linked (thanks!) and he seems to support the idea that climate modelling is best characterized as a boundary value problem and weather is an initial value problem, so I’m not sure how you can use his views to support your argument.

“forecasts which are not dependent on initial conditions, for example predicting changes in the statistics of climate as a result of some prescribed imposed perturbation, would constitute a ‘prediction of the second kind’.”

“estimating the effects on climate of a prescribed volcanic emission, prescribed variations in Earth’s orbit (thought to cause ice ages) or prescribed anthropogenic changes in atmospheric composition, would constitute a climate prediction of the second kind.”

I agree that climate modelling also has elements of the first kind, but it is fundamentally a boundary value problem so the analogy still holds. Again consider the difference between exactly and similar or primarily/mostly vs only/absolute…

Thanks for the additional information, I have learned something, which is why I come here.

It does not however address the point of my very simple and straightforward argument that predicting wave/weather events is not “exactly analogous” to predicting tides/climate change (nor is it close enough to infer reliability of predicting one from the reliability of predicting the other). I did not have in mind boundary vs initial value problems when I wrote it or referenced it, it is an appeal to much less technical properties of the systems in question.

As your additional information demonstrates, it is easy to over-intellectualize a simple point as a way of avoiding obvious conclusions (it is obvious that one learns nothing about the reliability of a climate model at 100 years from the reliability of a weather model at 7 days).

Unless you think that Nancy Green Raine was thinking about boundary vs initial value modeling problems and was aware of the papers and concepts you point out above then I think the charge of arguing from ignorance holds water. My choice of the words “inexcusably ignorant” has as much to do with her job title and public prominence as it does with her thought process and I would not apply such a label to everyone using that argument.

I do, however, have no reservations about saying one must be willfully ignorant to hang on to that position (bad long term weather forecasts –> unreliable climate models) after having the basic concepts explained to them.

Once more for the record, I do not say that predicting climate change is trivial, just that it is too dissimilar from predicting weather to make useful parralells in policy debates.

Cheers.

>>At what point should the climate predictions, er sorry, scenarios start showing up in the weather? At what point does it become acceptable to question the con[s]ensus of opinion?<<

Prediction is science, scenario is management.

Our institutions haven’t done a good job at educating the public on scenarios, so the error is understandable.

But as Roger’s dad endeavors to point out, one metric is not enough for management, nor is it enough to assess the state of the planetary ecosystem, as there are regional changes going on all around us. For example, you’ll want to think about regional changes and their affect on species.

Or, you can think about man’s effects on regional ecosystems in this way:

http://news.yahoo.com/s/ap/20061209/ap_on_sc/warmer_world_african_lakes%26printer=1

Where we have climate/ecosystem change refugee problems looming on the horizon. Will our institutions be able to deal with them, and react better after the Katrina exercise?

The subtlety of these socioenvironmental interactions, you can see, isn’t captured by the GAST metric you use for your argument above.

As for Kamloops State, I’d much rather have some no-account backcountry town put itself on the map like this, instead of having some 15-year-old shoot up his classroom. This of it as progress.

Can you explain to me how “misrepresenting science” and the ’scientization of politics’ are mutually exclusive and distinct? Because IMO the two are not mutually exclusive — Greene has misrepresented the science by claiming that there have been ’scientific discoveries’ which rendered the KP out of date. What studies is she referring to? But you are also correct that she is ’scientizing politics’ (as I understand the term) to the extent that she is invoking science to discredit a political agreement (KP).

On the other hand your claim that science cannot dictate the worth of the KP is equally dubious. If sensitivity to 2XCO2 is miracously found to be near zero, then the ‘worth’ of KP is clearly called into question (and this seems to be what Green is implying by her statement). If it is determined to be 5-6 C, then clearly it may be of more worth (at least insofar as it represents the only active global effort to reduce GHGs in a significant way).

Re the tides question, all you said in your first response to Coby was:

“And if you want to get into the differences between weather and climate forecasts, within the scientific community this issue is not so “black and white” as you suggest. Predicting the future behavior of the climate is not like predicting tides.”

But as Coby and I have pointed out it is very similar to predicting tides to the extent that both are boundary value problems, and given your ambiguous initial response, I think it’s reasonable that I interpreted your criticism that way (although your follow-up did help clarify things).

More generally, pointing out the ways in which analogies do not apply does not detract from the ways that they do apply. It’s the difference between ’similar’ and ‘exactly’.

Thanks for your comments. A few replies;

You are in fact cherry picking (or misrepresenting) the science of climate prediction. Consider this point of view from Tim Palmer:

“Compared with weather and seasonal forecasts, where verifying data is reasonably plentiful, it is difficult to assess quantitatively the predictions of uncertainty in forecasts of climate change.”
http://ej.iop.org/links/r1UzFhli1/eDHamx6K2xGHjdGqav5vpA/r002r1.pdf

One reason for this uncertainty is the degree to which long-term climate can be treated solely as an boundary value problem. There is in fact a range of views within the climate science community on this subject.

Consider the following perspective of F. Giorgi, (2005, Climatic Change) which includes your perspective, but is also broader:

“. . . we can still predict the statistical behavior of the climate system in response to external forcings. This is the realm of predictability of the second kind. Climate change prediction at time scales of multi-decadal to centennial thus has a predictability component of the second kind. We cannot predict the evolution of specific weather events years or decades into the future but we can address questions such as: How will climate statistics such as mean precipitation, precipitation variability or frequency of extreme events change over the next decades in response to changes of GHG concentration?

On the other hand, because of the long time scales involved in ocean, cryosphere and biosphere processes a first kind predictability component also arises. The slower components of the climate system (e.g. the ocean and biosphere) affect the statistics of climate variables (e.g. precipitation) and since they may feel the influence of their initial state at multi decadal time scales, it is possible that climate changes also depend on the initial state of the climate system (e.g. Collins, 2002; Pielke, 1998). For example, the evolution of the THC in response to GHG forcing can depend on the THC initial state, and this evolution will in general affect the full climate system. As a result, the climate change prediction problem has components of both first [initial] and second [boundary] kind which are deeply intertwined.”
http://www.springerlink.com/content/y1511t4413357484/fulltext.pdf

You might want to rethink the confidence that you express in the following sentence as the basis for labeling others as “inexcusably ignorant”: “Weather modeling is an initial conditions problem, where as climate modeling is a boundary conditions problem.”

And for the record, I do not think that Greene was invoking initial vs. boundary conditions, but even if she had, she (like Coby) would be guilty of cherry picking the science for convenience sake. Her statement is of course underdetermined …

Thanks!