The Intergovernmental Panel on Climate Change
The IPCC (Intergovernmental Panel on Climate Change) has become the primary agency, although certainly not the only agency, for advising governments around the world on strategies and actions that recognize climate change as a factor that can affect economies, health, safety, and much more. The IPCC is a governmental organization (not really a scientific organization) although it makes extensive use of scientific and other materials and information to make its recommendations through a series of reports known as Assessment Reports. There have been four such and a fifth (AR5) is due to be completed in 2013 and 2014.
The IPCC does not carry out any original research and does not monitor or oversee any original research. It’s role is to summarize and discuss the results of climate change research and its implications. IPCC does attempt to ensure that the work it does is of a high standard: “The AR5, summarizing the state of scientific knowledge about climate change, is going through an elaborate system of drafting, review by experts and governments, and revision to ensure that it meets the highest standards, is comprehensive and reflects the published literature and a range of scientific viewpoints.”
In fact, the IPCC does much more than and much less than summarize the state of scientific knowledge about climate change. It deals extensively and importantly with all the implications of a changing climate and how the world might adapt to that change or even attempt to manage or control the change. On the other hand, it does not attempt to critique the science behind climate change other than to insist, in the main, that the publications used in the physical sciences behind climate change are reviewed by peer scientists. Implications arising from the results of climate change science cited in IPCC assessment reports include weather extremes, sea level rise, and temperature increases among other things. The subsequent reports deal with how to adapt to or mitigate these implied areas of concern.
The fifth report will be prepared using the same internal process as the previous reports. The IPCC has three “Working Groups” to prepare the reports. WG I deals with the “Physical Science Basis” of climate change (due mid September 2013). WG II deals with “Impacts, Adaptation and Vulnerability” that are implied by the results of WG I (due mid March 2014) and WG III deals with “Mitigation of Climate Change” based on the results of WGII (due early April 2014). Finally a synopsis of all the WG reports will be pulled together in the “AR5 Synthesis Report” (due October 2014).
In the course of preparing the reports, workshops are held on specific topics. For example, one was held dealing with ocean acidification which resulted in a report: IPCC, 2011: Workshop Report of the Intergovernmental Panel on Climate Change Workshop on Impacts of Ocean Acidification on Marine Biology and Ecosystems [Field, C.B., V. Barros, T.F. Stocker, D. Qin, K.J. Mach, G.-K. Plattner, M.D. Mastrandrea, M. Tignor and K.L. Ebi (eds.)]. IPCC Working Group II Technical Support Unit, Carnegie Institution, Stanford, California, United States of America, pp. 164. All such workshop reports are not necessarily considered by the IPCC Panel, but because they have members of the various working groups on them, the material is submitted for consideration and forms part of the working groups’ background information. In each case however, the reports carry the following statements: “This workshop was agreed in advance as part of the IPCC workplan, but this does not imply working group or panel endorsement or approval of the proceedings or any recommendations or conclusions contained herein. Supporting material prepared for consideration by the Intergovernmental Panel on Climate Change. This material has not been subjected to formal IPCC review processes.”
In the case of WG 1 dealing with the physical basis of climate change the membership and contributors are primarily working scientists. The reports of the scientists that are used to prepare the first draft of the AR5 chapters must first have been published or be accepted for publication in a reviewed journal. These papers are not under the influence of any specific body other than the institute to which the author belongs and the editorial policies of the journal in which the work is published.
The “AR” reports of the IPCC are essentially three groups of summaries of the scientific papers termed “First-Order Drafts” prepared by scientists. The initial drafts are then subject to review, additions, and alteration by a combined panel of experts and governmental representatives (more than 130 governments are represented in IPCC but only 30 to 60 are usually active on any one report). Thus, this Second-Order Draft has some political dimensions to it. Finally the Third-Order Draft is once again reviewed, but this time by government representatives only. The final draft is summarized in language designed to helpful for policy makers. Both the full final draft report and the summary for policy makers then goes to the IPCC plenary for acceptance and approval.
For example, here is the Table of Contents for: Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, 2007, Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.) Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.
Preface and Foreword
Summary for Policymakers
Frequently Asked Questions
Historical Overview of Climate Change Science
Changes in Atmospheric Constituents and Radiative Forcing
Observations: Atmospheric Surface and Climate Change
Observations: Changes in Snow, Ice and Frozen Ground
Observations: Ocean Climate Change and Sea Level
Coupling Between Changes in the Climate System and Biogeochemistry
Climate Models and their Evaluation
Understanding and Attributing Climate Change
Global Climate Projections
Regional Climate Projections
Thus, the reports of the IPCC, as is standard for any intergovernmental organization, represent many-voiced statements. The voices are the scientists, experts and government-appointed representatives ranging from diplomats to business people. The hoped-for result is one where the foundation is based on evidence and scientific predictions from that evidence. The recommendations are a balanced meld of the science, the implications of the science, and the political strategies that could be undertaken to deal with the implications. A quick glance at the actual wording in this report based on the work of the scientists is that it remains an evidence-based discussion. The entire report is available online or as a pdf.
Over the course of the life of the IPCC, many improvements in the scientific process have been undertaken. The number of participating scientists and climate modellers has increased, as well as the computing power and database sharing. AR4, the most recent completed report, had experts from more than 130 countries, more than 450 lead authors, more than 800 contributing authors, and an additional 2,500 experts reviewed the draft documents. So there are many voices included in these reports.
The overall data set used is archived by the Data distribution Center (DDC) which was set up to: “facilitate the timely distribution of a consistent set of up-to-date scenarios of changes in climate and related environmental and socio-economic factors for use in climate impacts assessments. The intention is that these new assessments can feed into the review process of the IPCC.”
The DDC provides four main types of data and guidance (quoted from DDC website):
1. Observed Climate Data Sets: The climate observations comprise 1961-1990 mean monthly data over global land areas for nine variables on a 0.5º latitude/longitude grid, together with decadal anomalies from this mean for the period 1901-1995. This data set is currently being updated to 2000 and interpolated to a finer resolution (10 x 10 arc minutes). Pointers are provided to other relevant global climatologies.
2. Global Climate Model Data: Global climate model data is available as monthly means or as climatologies. Data is held for climate model projections used as input to the Second, Third and Fourth IPCC Assessment Reports. Full details of the variables held can be found on the relevant data pages. The climatologies of climate model projections can also be viewed through the DDC visualisation service.
Daily fields are not provided by the DDC, but are available from PCMDI or from the respective modelling centres (see list below).
3. Socio-economic data and scenarios: Socio-economic data and scenarios are required for describing socioeconomic development and adaptation capacity. The reference data include country and regional level indicators of socio-economic and resource variables. The scenario data supplied extend to 2100 and are based on the assumptions underlying the new set of emissions scenarios developed for the IPCC Special Report on Emissions Scenarios, SRES, as well as the six IS92 emissions scenarios prepared by the IPCC in 1992 (see previous post on climate change models). There is also detailed guidance on the use of DDC data to develop socio-economic and adaptation scenarios as well as links to related guidance material developed by other agencies.
4. Data and scenarios for other environmental changes: Some data and information for other environmental changes are also included in the site. These include data on global mean CO2 concentration, global and regional sea-level rise, regional ground-level ozone concentration, sulphate aerosol concentration and sulphur deposition. All of these scenarios were developed for the IPCC Third Assessment Report based on the SRES emissions scenarios. Detailed documentation and guidance is also provided for the use of these data.
Much of this information, or at least the definition of the variables, is freely available in summary or tabular format. For qualified users of climate models that conform to the standard coupled models, it is possible to download the actual data and run simulations that test their climate models’ match to the data. There is an enormous amount of data archived — close to 400 terabytes — so unless an interested user is part of a university or institution that has access to huge data storage areas and a supercomputer to process the data, it is not of any real interest to try to examine it in detail. For those interested in the results of observations, many are available in tabular form as low resolution data or even global or hemispheric averages. A very simple example that may be of interest to those following the graphing of global temperature (the lowest resolution data) from 1880 to the present is available here.
The IPCC and the basic science underlying the climate change models have been challenged in many quarters, including some scientific challenges. Some such challenges have been valid criticisms of the approach to the modelling. To address those concerns, the Working Group I responsible for basic physical science has embarked on a very large combined effort called CMIP5.
By the way, for readers uncertain about what a mathematical model is like, here is a simple example. I want to figure out how much sand I can put in a box. Here is the “Mathematical model”: length (L) times width (W) times height (H) equals the volume (V) of the box. L, W, and H, are “variables”. They are all correlated so if I change one of them, the value of V changes as well. Suppose we also want the “model” to estimate the weight of the sand. The weight (Wt) will be the specific gravity of sand (SS) times the same volume (V). So in mathematical terms it looks like this: LxWxH = V and V x SS = Wt. Putting the two together and simplifying we get LxWxHxSS = Wt. So that is the model. The model has no numbers to start with. If we do an experiment to predict the answer we would make a pretend box that was let’s say .5m x 1m x .25m on the inside dimensions, and let’s fill it with pretend dry sand that has a specific density of 1602 kg/cu.m. The predicted answer is 200.25 Kilos. To test the model, we make a real box that is the same interior dimensions and fill it with real sand. When it is full, we can pour out the sand and measure its weight on a scale. If we do not get 200.25kg, we need to figure out if we made a mistake in the model or if we made a mistake in the measurements of the real box and real sand. While the SS of the sand is a real variable, it is hard to measure, so instead we chose a constant value. But if the sand is a little wet or even very wet, the Weight Wt changes even if the volume V stays the same. We classify the wettness in groups of values. (dry, damp, wet, very wet) and choose in between values. This is called a parameter (variable constant). The process of deciding what values to use in each situation is called “parameterization” and is often used in climate models for variables that are difficult to measure — like clouds.
Climate modellers do exactly the same thing when they create their models. They design a mathematical model (it’s more complicated than the one I just gave as an example!), run a simulation or prediction to see how well the model fits reality. If it isn’t perfect, then they have to figure out if the model is wrong or if the measurements are wrong. Depending on what they find, they re-measure or find a different way to measure. If the data stays the same, the problem is with the model. If the data need to be revised, the same model is re-tested. If it is still wrong, the problem is most likely with the model. So it is a long and often tedious process to iron out all the potential wrinkles.
CMIP5 (Coupled Model Intercomparison Project Phase 5)
CMIP5 is an enormous undertaking with literally thousands of scientists from all over the world contributing information and published reports. It is one of the largest international scientific projects ever undertaken and the information (both input and output, as well as the original observed values to be used as the definition of “reality”) used in the project is available to see. In the scientific literature, there is a great deal of discussion about climate change uncertainties, the values to be applied to variables, the mathematical values of the effects of feedbacks, and so on. Some variables, such as clouds which although they can be modelled, are too small to be included as ordinary variables in the overall models. Instead they are subject to “parameterization”. The idea here is to model them at the correct scale, but to do many runs and with each run introduce very small changes to the values in the factors that cause clouds. These are run many times in groups (called ensembles). As the many answers are assembled, the values tend to settle around averages, so the means of these averages are used as the values of the parameters. Parameters are often characterized as variable constants and this is why. The clouds are valid variables to use in the models, but at least with present-day supercomputers, it is not feasible to include them as full variables. So the answer is to create a fixed number to use instead, but to use a number that is adjusted to the conditions as appropriate (a variable constant) as determined by ensemble runs. Most climate modellers agree that clouds are an important variable that is not handled very well, so in this next series of simulations, special emphasis will be placed on clouds.
The most fundamental summaries provided by the IPCC are projections of climate change based on the work of many active climate scientists using many climate models, all of which are essentially based on one approach to climate modelling: AOGCM (Fully coupled atmosphere-ocean model of the three-dimensional global climate). These in turn are based on a General Circulation Model (GCM), a three-dimensional model. When the GCM is of the atmosphere only, it often is dubbed an AGCM. A similar circulation model of the ocean is also a GCM, often dubbed OGCM. When the two are coupled together so that they interact as in the real world, they are dubbed the AOGCM (Atmosphere-Ocean General Circulation Model).
The basic variables are the temperature, humidity, liquid/ice water content and atmospheric mass. Factors considered are many of course, but typically include at least, advection, radiation calculations, surface fluxes (latent, sensible heat etc.), convection, turbulence and clouds. More elaborate Earth System models often consider atmospheric chemistry and aerosols (including dust and sea salt).
Over time the resolution of these models has increased with the computing power of the computers (an increase of about 1,000,000 times since the 1970’s). In this figure (from AR4 IPCC Climate Change 2007: Working Group I: The Physical Science Basis)
“Geographic resolution characteristic of the generations of climate models used in the IPCC Assessment Reports: FAR (IPCC, 1990), SAR (IPCC, 1996), TAR (IPCC, 2001a), and AR4 (2007). The figures above show how successive generations of these global models increasingly resolved northern Europe. These illustrations are representative of the [currently] most detailed horizontal resolution used for short-term climate simulations. The century-long simulations cited in IPCC Assessment Reports after the FAR [First Assessment Report] were typically run with the previous generation’s resolution. Vertical resolution in both atmosphere and ocean models is not shown, but it has increased comparably with the horizontal resolution, beginning typically with a single-layer slab ocean and ten atmospheric layers in the FAR and progressing to about thirty levels in both atmosphere and ocean.” (Quote from IPCC AR4)
One of the problems faced by climate scientists is attempting to bring the resolution of the time dimension (as opposed to the spatial dimensions) down so that a greater degree of fit to the observable variations on a human scale is possible. In AR5 in Working Group I, the ambition is to move from the current 30 year to 10 year (decadal) periods in projections. Weather predictions themselves suffer from the same problem — that is most of the Earth-based systems are non-linear or chaotic. They are “chaotic” but not random (they are subject to knowable forces). They do not normally follow smooth changes, and that is why we see rapid changes in weather conditions and in some cases, rapid changes in climate conditions.
While the comprehensive models are huge, many smaller models are used to handle subsets of the large models. Nonetheless, to handle just the physical science basis of the climate modelling, the amount of information to be handled, coordinated, and integrated, is mind boggling.
For ordinary citizens, it is also possible to participate in at least one experimental simulation by allowing one agency to run simulations in the background on your computer when it is not being used for other things. “Climateprediction.net is a distributed computing project to produce predictions of the Earth’s climate up to 2100 and to test the accuracy of climate models. To do this, we need people around the world to give us time on their computers – time when they have their computers switched on, but are not using them to their full capacity.”
Most scientific predictions however will be run on supercomputers. To help ensure a cohesive and efficient system, over 20 organizations that employ many climate models and sub-models have agreed to share both the models and the data from those models as well as to participate in a rigorously defined protocol so that inter-model comparisons can be made accurately.
In “A Summary of the CMIP5 Experiment Design; Taylor, Stouffer, Meehl, 2009 (2011)” the authors state that “CMIP5 promotes a standard set of model simulations in order to:
• evaluate how realistic the models are in simulating the recent past,
• provide projections of future climate change on two time scales, near term (out to
about 2035) and long term (out to 2100 and beyond), and
• understand some of the factors responsible for differences in model projections,
including quantifying some key feedbacks such as those involving clouds and the
CMIP5 does not specify the nature of the models, only the tests that they should undertake. The reason for this is to be able to make comparisons to provide a multi-model context for
1) assessing the mechanisms responsible for model differences in poorly understood feedbacks associated with the carbon cycle and with clouds,
2) examining climate “predictability” and exploring the ability of models to predict climate on decadal time scales, and, more generally,
3) determining why similarly forced models produce a range of responses.”
It will be interesting to see the increased accuracy or increased failure rate when new attempts to model clouds and carbon cycles are added to the models. Many look forward to seeing the increased resolution to a 10 year time scale. That may increase the ability to both hindcast and capture more of the variation as well as improve (or not) the short- and long-range forecasts. An emphasis on testing the predictive accuracy of models over periods where the data is well known will add to the confidence. At present there is no intent to restrict shared data to models that have a specific level of accuracy, but of course the results will be readily available for comparison.
CMIP5 acts as a coordinating body to compare experiments, especially simulations, that will be used in the reports of the IPCC. CMIP5 recognizes that it cannot be comprehensive; it cannot possibly include all the different model intercomparison activities that might be of value, so it expects that various groups and interested parties will develop additional experiments that might build on and augment the experiments.
The modeling centers make their data available either by directly publishing on their own ESG (Earth System Grid) Data Node or by sending it to a designated core data node. The model output is published to the Data Node, which makes the data visible to a Gateway and enables its delivery to end-users who can search, discover, and request data and data products. A quick glance at the completed products with data available from CMIP5 (not the entire data set for ESGF) illustrates that today the project is a long way from complete. This chart is constantly updated, so if you look at the chart a year from now, it will be quite different with many more blanks filled in.
On the other hand, one of the most impressive aspects of the fifth reports of IPCC will be that the scientific base will include a much broader scan of climate models combined with rigorous testing using variables that are very tightly defined in a set of simulations where the differences among the models will be immediately obvious. How well the models will be able to predict the recent past by testing the predicted data points against the against actual observed data points should give both a better method of measuring the confidence of any given model and the range of variation amongst models using at least the same minimum number of well-defined variables.
List of Main Climate Modelling Centres Contributing to CMIP5
Beijing Climate Center, China Meteorological Administration
Canadian Centre for Climate Modelling and Analysis
Centro Euro-Mediterraneo per I Cambiamenti Climatici
Centre National de Recherches Meteorologiques / Centre Europeen de Recherche et Formation Avancees en Calcul Scientifique
CSIRO (Commonwealth Scientific and Industrial Research Organisation, Australia), and BOM (Bureau of Meteorology, Australia)
Commonwealth Scientific and Industrial Research Organisation in collaboration with the Queensland Climate Change Centre of Excellence
The First Institute of Oceanography, SOA, China
College of Global Change and Earth System Science, Beijing Normal University
Institute for Numerical Mathematics
Institut Pierre-Simon Laplace
Institute of Atmospheric Physics, Chinese Academy of Sciences; and CESS, Tsinghua University
Institute of Atmospheric Physics, Chinese Academy of Sciences
Atmosphere and Ocean Research Institute (The University of Tokyo), National Institute for Environmental Studies, and Japan Agency for Marine-Earth Science and Technology
Japan Agency for Marine-Earth Science and Technology, Atmosphere and Ocean Research Institute (The University of Tokyo), and National Institute for Environmental Studies
Met Office Hadley Centre
Max Planck Institute for Meteorology (MPI-M)
Meteorological Research Institute
NASA Goddard Institute for Space Studies
NASA Global Modeling and Assimilation Office
National Center for Atmospheric Research
Norwegian Climate Centre
National Centers for Environmental Prediction
Nonhydrostatic Icosahedral Atmospheric Model Group
National Institute of Meteorological Research/Korea Meteorological Administration
Geophysical Fluid Dynamics Laboratory
National Science Foundation, Department of Energy, National Center for Atmospheric Research