Global Warming: Experts’ Opinions versus Scientific Forecasts

Global Warming | Policy Reports

No. 308
Friday, February 01, 2008
by Kesten C. Green and J. Scott Armstrong

An Audit of the IPCC Report

To what extent have those who have made climate forecasts used scientifically tested forecasting procedures?  Using forecasting audit software, the authors of this study independently assessed the extent to which IPCC procedures conformed to or violated forecasting principles.

Climate Forecasters’ Use of the Scientific Literature on Forecasting Methods.24 There is little use of forecasting principles in environmental research generally; apparently they are not used at all in climate research.25 An Internet search found no Web sites or papers on climate change that referenced forecasting methodology.26 Neither the IPCC Report’s Chapter 8, “Climate models and their evaluation,” nor any of the 788 referenced works therein, refer to forecasting methodology or established forecasting principles.27; The same was true of Chapter 9, “Understanding and attributing climate change,” and its 535 references.  A survey of climate scientists (described below) did not yield references to any relevant papers on forecasting.28

Forecasting principles have been derived from all known empirical evidence on estimating the as yet unknown. The principles are therefore scientific.  Evidence comes from all disciplines that have produced relevant evidence and the principles are applicable to all forecasting problems — from weather to company sales, from the spread of non-native species to investment strategy, and from war fighting to egg hatching rates.  [See the sidebar on Principles of Forecasting.]

Forecasting Audit Results.  Of the 140 forecasting principles, the audit found that 127 forecasting principles are relevant to the procedures used to arrive at the climate projections in the IPCC report.  Of these 127, the methods described in the report definitely violate 60 principles, 12 appear to be violated and there is insufficient information to assess the use of 38.

“Official climate reports do not use scientific forecasting principles.”

For example, “Make sure forecasts are independent of politics (Principle 1.3),” is one of the 60 principles the IPCC process clearly violated.  David Henderson, a former Head of Economics and Statistics at the OECD, has given a detailed account of how political considerations influence all stages of the IPCC process.29 For example, the “Summary for Policymakers” that accompanies each of the IPCC’s assessment reports is released with much media and public fanfare.  The summary is written in negotiation with the explicit input of legislators, policymakers and/or diplomatic appointees.  Most recently, chapter by chapter drafts of the Fourth Assessment Report’s “Summary for Policymakers” were released months in advance of the final version of the full report, with the directive that the final version of the chapters in the report be expressly written to reflect the language negotiated by the lead authors with the participating political appointees to the IPCC.

Read Article as PDF