Image
© Unknown
Guest post by Indur M. Goklany

Although it is encouraging that the Royal Society now acknowledges that climate science may not be as settled as it previously implied, the Society's new report still stands as an embarrassment to science because it fails to offer justifications based on science (and policy analysis) for a number of its (politically correct) statements.

First, it claims in its opening sentence, "Changes in climate have significant implications for present lives, for future generations and for ecosystems on which humanity depends." But two paragraphs later it acknowledges that, "[T]he impacts of climate change ... are not considered here." Hence, the RS has no scientific (or other basis) for this claim. At most it could say, "ALTHOUGH WE DID NOT CONSIDER THEIR IMPACTS, changes in climate COULD have significant implications for present lives..., IF SUCH CHANGES ARE VERY LARGE." [Suggested INSERTIONS in the RS's original language are in UPPERCASE letters.]

For the same reason, the RS's statement in the very last paragraph (number 59), "However, the potential impacts of climate change are sufficiently serious that important decisions will need to be made", is unsupported by any evidence.

Equally embarrassing are statements regarding the cause (or attribution) for recent warming. In paragraph 2, it states, "There is strong evidence that the warming of the Earth over the last half-century has been caused largely by human activity." But we are talking climate, not weather, and half a century doesn't even span a one full cycle of the AMO or PDO, nor does it span the length of the Little Ice Age, Medieval Warm Period, or other historical periods of climatic change.

Moreover, what is the "strong evidence" referred to in the preceding quote that allows the RS to claim that warming has been "caused largely by human activity"? This "strong evidence" comes down to the RS's acceptance of the methodology underlying the IPCC's claim of attribution (see paragraphs 37 to 39). But as noted in another WUWT post, this is an "argument from ignorance" - certainly not what one would expect from as august a scientific body as the Royal Society.

In addition, this argument assumes the validity of models, even though they have never been validated using "out-of-sample" observational data, and are, moreover, unable to simultaneously provide reliable estimates for surface temperature AND precipitation at less than continental scales (see Appendix A). The inability of models to generate reliable estimates for surface temperature at such scales is noted in paragraph 50, but the RS is conspicuously silent on their inability to do any better with precipitation. These issues (and others) are noted on the annotated PDF of the RS report here:

RS_ClimateChange_SummaryofScience (PDF)

Third, it is disappointing to see the RS use the term "predict" in conjunction with model results. The IPCC does not "predict", it "projects". As Kevin Trenberth has noted, the IPCC makes "no predictions ... instead [it] proffers 'what if' projections of future climate that correspond to certain emissions scenarios."

APPENDIX A: IPCC Models Have Not Been Validated

IPCC models have not been validated using out-of-sample data under conditions of high greenhouse gas concentrations.

It is insufficient for a climate model to accurately reproduce the spatial and temporal pattern for one climatic variable; it should be able to do so for the ensemble of variables that have a significant effect on impacts. This includes not just temperature but, perhaps more importantly, precipitation. But little confidence can be placed in the IPCC model results to simultaneously reproduce results for both temperature and precipitation even when "in sample" data are used, let alone when out-of sample" data are utilized.

As noted elsewhere (Goklany 2009, pp 12-13, at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1548711

[F]or a climate model to be valid, it should be able to simultaneously forecast with reasonable accuracy the spatial and temporal variations in a wide variety of climatic variables including temperature, pressure and precipitation, as well as endogenously produce the patterns and rhythms of ocean circulation (among other things). But we know from the AR4WG1 that models are unable to do this even for "in sample" data. As it states, "Difficulties remain in reliably simulating and attributing observed temperature changes at smaller [that is, less than continental] scales" [AR4WG1: 10.] And this [is] what it says about projections of climate change:
" There is considerable confidence that climate models provide credible quantitative estimates of future climate change, particularly at continental scales and above. This confidence comes from the foundation of the models in accepted physical principles and from their ability to reproduce observed features of current climate and past climate changes. Confidence in model estimates is higher for some climate variables (e.g., temperature) than for others (e.g., precipitation)." (AR4WG1: 600, emphasis added)
This tacitly acknowledges that confidence is low for model projections of temperature at less than continental scales, and is even lower for precipitation - perhaps even at the continental scale. Notably, it doesn't provide any quantitative estimate of the confidence that should be attached to projections of temperatures at the subcontinental scale. This lack of confidence in temperature and precipitation results at such scales is reaffirmed by recent reports from the US Climate Change Science Program (CCSP):
"Climate model simulation of precipitation has improved over time but is still problematic. Correlation between models and observations is 50 to 60% for seasonal means on scales of a few hundred kilometers." (CCSP 2008:3).

In summary, modern AOGCMs generally simulate continental and larger-scale mean surface temperature and precipitation with considerable accuracy, but the models often are not reliable for smaller regions, particularly for precipitation." (CCSP 2008: 52).
The IPCC does not say that "all" features of current climate or past climate changes can be reproduced, as a good model of climate change ought to be able to do endogenously. In fact, it notes:
"Model global temperature projections made over the last two decades have also been in overall agreement with subsequent observations over that period (Chapter 1). "Nevertheless, models still show significant errors. Although these are generally greater at smaller scales, important large scale problems also remain. For example, deficiencies remain in the simulation of tropical precipitation, the El Niño-Southern Oscillation and the Madden-Julian Oscillation (an observed variation in tropical winds and rainfall with a time scale of 30 to 90 days)." (AR4WG1: 601; emphasis added).