The Next Battlefield
Posted by Jeff Id on February 21, 2010
Guest post — John Pittman
The Next Battlefield
After the climategate emails, the failure of Copenhagen, and the IPCC Gate-du-Jour, a common thread is starting to appear on the AGW proponents’ blogs. The thread is that even if something is found from climategate, it won’t matter because… That because is usually the models. But do such claims agree with what the IPCC 4AR stated? A few articles with different aspects of why it doesn’t matter. http://www.newsweek.com/id/225778 http://www.huffingtonpost.com/2009/12/02/climategate-the-7-biggest_n_371223.html http://articles.latimes.com/2009/dec/09/opinion/la-oe-macdonald9-2009dec09?pg=2
First, there are a bunch of red herrings out there on both sides. One of the common ones is that it has been getting warmer and ALL THE EVIDENCE shows it. The real fight is over CO2 induced large temperature increases. Even all the modeling done about the potential effects are a side issue compared to the real question. The real question is CO2 guilty, and how was it proven. WRT the emails, do they give reason to doubt this proof? To answer, we go to IPCC 4AR.
We start with some relevant methodology from Chapter 8.
From Section 8.1.1 WG1 IPCC 4AR http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch8.html Over time, statistics can be accumulated that give information on the performance of a particular model or forecast system. In climate change simulations, on the other hand, models are used to make projections of possible future changes over time scales of many decades and for which there are no precise past analogues. Confidence in a model can be gained through simulations of the historical record, or of palaeoclimate, but such opportunities are much more limited than are those available through weather prediction.
This is pretty straight forward. If we want to determine if a model is reasonable in under a 100 to 130 years for a 100 year prediction, we need historical and paleoclimate simulations, which means a good historical record and good reconstructions. http://rsta.royalsocietypublishing.org/content/365/1857/2053.full where it is stated
And even if the scenario were to be followed, waiting decades for a single verification dataset is clearly not an effective verification strategy. This might sound obvious, but it is important to note that climate projections, decades or longer in the future by definition, cannot be validated directly through observed changes. Our confidence in climate models must therefore come from other sources. The judgement of whether a climate model is skilful or not does not come from its prediction of the future, but from its ability to replicate the mean climatic conditions, climate variability and transient changes for which we have observations, and from its ability to simulate well-understood climate processes. For example, climate models are evaluated on how well they simulate the present-day mean climate (e.g. atmospheric temperature, precipitation, pressure, vertical profiles, ocean temperature and salinity, ocean circulation, sea ice distributions, vegetation, etc.), the seasonal cycle and climate variability on various time scales (e.g. the North Atlantic oscillation, ENSO, etc.). Their response to specified forcing is compared to the observed warming over the industrial period. They are evaluated against proxy data from past climate states, e.g. the Last Glacial Maximum, the Mid-Holocene, the last Interglacial period, or even further back in time.
The next quote confirms the need of good observations
From 8.1.2 A climate model is a very complex system, with many components. The model must of course be tested at the system level, that is, by running the full model and comparing the results with observations.
The next section gives the basic approach. However, the Royal Society publication by Tebaldi and Knutti linked above is a better, more complete read with history and updates. The important point is that projections’ skillfulness cannot be tested directly or how good a test of skill of a model’s result cannot be tested directly.
18.104.22.168 What does the accuracy of a climate model’s simulation of past or contemporary climate say about the accuracy of its projections of climate change? This question is just beginning to be addressed, exploiting the newly available ensembles of models. A number of different observationally based metrics have been used to weight the reliability of contributing models when making probabilistic projections (see Section 10.5.4)….. For any given metric, it is important to assess how good a test it is of model results for making projections of future climate change. This cannot be tested directly, since there are no observed periods with forcing changes exactly analogous to those expected over the 21st century. However, relationships between observable metrics and the predicted quantity of interest (e.g., climate sensitivity) can be explored across model ensembles.
Now to a series of quotes that go to the heart of whether the emails matter. In this section, the use of “present climate” is an important part of the evaluation. Note that the models run with fixed radiative forcing and try to simulate known or assumed conditions including pre-industrial. The use of aerosol forcing is allowed to vary within its range of variability.
22.214.171.124 Testing models’ ability to simulate ‘present climate’ (including variability and extremes) is an important part of model evaluation (see Sections 8.3 to 8.5, and Chapter 11 for specific regional evaluations). In doing this, certain practical choices are needed, for example, between a long time series or mean from a ‘control’ run with fixed radiative forcing (often pre-industrial rather than present day), or a shorter, transient time series from a ‘20th-century’ simulation including historical variations in forcing…Models have been extensively used to simulate observed climate change during the 20th century. Since forcing changes are not perfectly known over that period (see Chapter 2), such tests do not fully constrain future response to forcing changes. Knutti et al. (2002) showed that in a perturbed physics ensemble of Earth System Models of Intermediate Complexity (EMICs), simulations from models with a range of climate sensitivities are consistent with the observed surface air temperature and ocean heat content records, if aerosol forcing is allowed to vary within its range of uncertainty. Despite this fundamental limitation, testing of 20th-century simulations against historical observations does place some constraints on future climate response (e.g., Knutti et al., 2002). These topics are discussed in detail in Chapter 9.
Section 126.96.36.199 states why the emails are important in its first sentence. Next follows a discussion of problems or potential problems that highlight why the emails are important and those who claim it can’t or won’t matter are wrong. First, models, by themselves, with just the modern period are not sufficient for determination. Using aerosol forcings which in its full range of uncertainty goes from a slightly positive to a negative forcing, increases uncertainty not decreases it since even the sign can either vary or is unknown. The next part of the section discusses limitations and uncertainties. But it is important to note “20th-century climate variations have been small compared with the anticipated future changes under forcing scenarios.” Then states the issues are discussed in depth in Chapter 6, which is where we want to go. The last part quoted is the use of initial conditions being used in a test. The inappropriateness of these sorts of tests can be found here http://climateaudit.org/2007/02/11/exponential-growth-in-physical-systems/ . This is an important consideration when considering some of the arguments made in support of using a “model” only approach. The work by Dr. Browning indicates that this is a questionable practice with undefined (for the modelers) pitfalls. http://climateaudit.org/2006/05/15/gerry-browning-numerical-climate-models/ is a must read for understanding that running to the model only as presented in the IPCC AR4 is not methodologically correct to recent publications that have NOT been incorporated into the AR4. Using models that have a compromise, such as a hyper viscous layer in an initial value and boundary condition model does not mean the models can’t be useful. It does mean that the math and physics are not correct. Take using such a layer, it violates one of the assumptions necessary to pose the PDE’s namely it is a continuum. But IF and ONLY IF it can be shown to be useful can such be used. The methodology is to use “modern” records and especially the last 1000 years to justify skill.
188.8.131.52 Simulations of climate states from the more distant past allow models to be evaluated in regimes that are significantly different from the present. Such tests complement the ‘present climate’ and ‘instrumental period climate’ evaluations, since 20th-century climate variations have been small compared with the anticipated future changes under forcing scenarios derived from the IPCC Special Report on Emission Scenarios (SRES). The limitations of palaeoclimate tests are that uncertainties in both forcing and actual climate variables (usually derived from proxies) tend to be greater than in the instrumental period, and that the number of climate variables for which there are good palaeo-proxies is limited. Further, climate states may have been so different (e.g., ice sheets at last glacial maximum) that processes determining quantities such as climate sensitivity were different from those likely to operate in the 21st century. Finally, the time scales of change were so long that there are difficulties in experimental design, at least for General Circulation Models (GCMs). These issues are discussed in depth in Chapter 6… Climate models can be tested through forecasts based on initial conditions. Climate models are closely related to the models that are used routinely for numerical weather prediction, and increasingly for extended range forecasting on seasonal to interannual time scales.
The rest of this is a repeat for those familiar with the methodology, so I have tried to limit the discussion to certain sections that highlight why the emails are important. The first section has most, if not all the key words, as to why the emails are so important. Replication, independent, cross-verification, confidence, inferences, and of course, the all time champ ROBUST. Each of these keywords has been called into question by the emails. Probably the best source, unless you want to do it yourself, is Climategate: The CRUtape Letters by Mosher and Fuller. This is not to say that it has been proven, but that the evidence is that confirmation bias should be assumed to have occurred unless it is shown that it has not. The evidence indicates that until the work is redone in an open environment, it should NOT be trusted. Please note and understand the methodological claim of depends heavily on replication and cross-verification from independent sources. This is a keystone to understanding whether or not the emails matter. They do.
Section 184.108.40.206 The field of palaeoclimatology depends heavily on replication and cross-verification between palaeoclimate records from independent sources in order to build confidence in inferences about past climate variability and change. In this chapter, the most weight is placed on those inferences that have been made with particularly robust or replicated methodologies.
In case there is any doubt, this section links models, and past climate changes which are stated to be “key to testing physical hypotheses.” Two points 1.) the initial value and scale problems in Dr. Browning’s work supported by Sylvie Gravel show that without some empirical justification, models cannot be supported as developed; 2.) Section 220.127.116.11 underscores that as one goes further into the past the utility for explaining current conditions, forcings, or relationships become more and more suspect i.e. their utility becomes progressively limited. This translates to the best time period to use is the one in question: the medieval warm period to the present. I have noted in discussions by AGW proponents that far past CO2/Temperature relations are being offered. A reading of the quoted Chapter 6 sections show that the conditions may be so different that such a claim as being currently applicable is unlikely. If it is unlikely, then these AGW proponents are actually laying claim to that the certainty indicated in AR4 and AR3 is false or the IPCC has it wrong. Take your pick.
6.2.2 Climate models are used to simulate episodes of past climate (e.g., the Last Glacial Maximum, the last interglacial period or abrupt climate events) to help understand the mechanisms of past climate changes. Models are key to testing physical hypotheses, such as the Milankovitch theory (Section 6.4, Box 6.1), quantitatively. Models allow the linkage of cause and effect in past climate change to be investigated…At the same time, palaeoclimate reconstructions offer the possibility of testing climate models, particularly if the climate forcing can be appropriately specified, and the response is sufficiently well constrained. For earlier climates (i.e., before the current ‘Holocene’ interglacial), forcing and responses cover a much larger range, but data are more sparse and uncertain, whereas for recent millennia more records are available, but forcing and response are much smaller. Testing models with palaeoclimatic data is important, as not all aspects of climate models can be tested against instrumental climate data. For example, good performance for present climate is not a conclusive test for a realistic sensitivity to CO2 – to test this, simulation of a climate with a very different CO2 level can be used. In addition, many parameterizations describing sub-grid scale processes (e.g., cloud parameters, turbulent mixing) have been developed using present-day observations; hence climate states not used in model development provide an independent benchmark for testing models. Palaeoclimate data are key to evaluating the ability of climate models to simulate realistic climate change.
For other considerations as to why the emails matter, there have been a series of posts on The Air Vent and other sites examining specific issues and emails. Using the keyword climategate will yield a number of relevant posts on both sides of the issue.