Fix The Mistakes That Killed The Climate Change Campaign

  • Date: 03/06/18
  • Larry Kummer, Fabius Maximus

Summary: Tracing the misrepresentation and misuse of RCP8.5 from a new paper back to its earliest days. How a big mistake and a small one combined to put much of climate science on a dead-end road. How climate scientists refusal to recognize these mistakes has kept it on this road.

Forecasting with models

After three decades, the campaign for public policy action to fight climate change has failed to produce substantial results in the US. With vast investments of work and money, plus support of many powerful institutions, it ranks among the biggest such failures in US history. Let’s trace one reason for its failure, starting with a new paper and looking back through time to the original error. Remember, it is never too late to learn and change course.

William Nordhaus is a professor of economics at Yale and creator of the Dynamic Integrated Climate-Economy model (DICE), one of the most widely cited integrated assessment models providing guidance to policy makers about climate change. His latest paper goes to the heart of climate policy debate.

Global Melting? The Economics of Disintegration of the Greenland Ice Sheet

NBER working paper by William D. Nordhaus, May 2018.

Opening of the abstract. Red emphasis added.

“Concerns about the impact on large-scale earth systems have taken center stage in the scientific and economic analysis of climate change. The present study analyzes the economic impact of a potential disintegration of the Greenland Ice Sheet (GIS). The method is to combine a small geophysical model of the GIS with the DICE integrated assessment model. The result shows that the GIS is likely to disappear over the next millennium or so without climate policy, but an active climate policy may prevent the GIS from crossing the threshold of irreversibility. …”

This is a well-constructed paper. It shares two typical characteristics of its genre. One is fun, the other has had awful effects.

Predictions over absurdly long horizons

“The ice-sheet decline is slow, with a GIS half-life of approximately eight centuries in the baseline path …”

What was high tech 800 years in the past? The wheelbarrow, rudder, and windmill were new inventions. The chimney would soon be invented. The hourglass and paper were a century away. Tech progress was slow. It accelerated in the 12th century, sped up again roughly 300 years ago, and reached incredible rates of progress in the late 19th century. Guessing what will possible even one century in the future is wild speculation.

Also, multi-century predictions of climate change should not be taken seriously. They are far beyond the state of the art.

Misrepresenting the future

“The arrow is the range of model estimates for a high warming scenario (RCP 8.5) from IPCC (2013) p. 1191, and has comparable forcings as the DICE-GIS baseline run.”

RCP8.5 is not just a “high warming scenario” but the worst-case scenario of the four used in AR5, aka IPCC 2013 (details here). As a worst-case scenario should, it assumes ugly changes in important long-term trends – such as tech (from progress to stagnation) and fertility (slow or stopping the decline in emerging nations). Each is unlikely; the combination is very unlikely.

Also, researchers have long questioned if the world has sufficient economically recoverable coal to fuel the 21st century – as RCP8.5 assumes. See two recent papers here (ungated copy here) and here.

If DICE-GIS has similar forcings, it must have similarly improbable assumptions. Therefore, it provides no basis for Nordhaus’ statement that “GIS is likely to disappear over the next millennium or so without climate policy.” He should have said that “under severe but unlikely circumstances, the GIS is likely to disappear over the next millennium.” But what’s the fun in that?

Full post

Recent Popular Articles


We use cookies to help give you the best experience on our website. By continuing without changing your cookie settings, we assume you agree to this. Please read our privacy policy to find out more.