NIPCC Report on Species Extinctions due to Climate Change

The NIPCC – Interim Report 2011 updates their last 2009 Report, with an overview of the research on climate change that the IPCC did not see fit to print. Its published by the Heartland Institute with lead authors Craig D. Idso, Australian Robert Carter, and S. Fred Singer with a number of other significant contributions.

I am grateful for inclusion of some of my work in Chapter 6 on the uncertainty of the range-shift method for modeling biodiversity under climate change.

The controversy centered on a paper by Thomas et.al 2004 called “Extinction Risk from Climate Change“, that received exceptional worldwide media attention for its claims of potentially massive extinctions from global warming.

Briefly, the idea is to simulate the change in the range of a species under climate change by ‘shifting’ the range using a presumed climate change scenario.

Daniel Botkin said of the Thomas et.al. 2004 study

Yes, unfortunately, I do consider it to be the worst paper I have ever read in a major scientific journal. There are some close rivals, of course. I class this paper as I do for two reasons, which are explained more fully in the recent article in BioScience:

… written by 17 scientists from a range of fields and myself (here).

While there are many problems with this paper, the most amazing, as I see it, is the way they used changes in the size of species-ranges to determine extinctions. Its generally believed that contracting a species-range increases the probability of extinction.

Consider the case of a species that disperses freely under climate change. While the range-size of individuals change, the average range-size should stay the same, unless there is a major obstruction like an ocean or mountain range. Species whose range size decreases are balanced by species whose range size increases. Overall, the net rate of extinction should be unchanged.

However, Thomas et.al. 2004 simply deleted all species whose range expanded. A massive increase in extinctions was therefore a foregone conclusion, even assuming free dispersion.

There are a number of other ways a bias towards range-reduction can be introduced, such as edge effects and over-fitting assumptions, that I show in my book “Niche Modeling“. In a normal science this would have been a cautionary tale of the dangers of ad-hoc methodologies.

It’s an example of the intellectual bankruptcy of the IPCC report that the uncertainties of Thomas et.al. 2004 and other similar studies were ignored by Working Group II. For example, in Impacts, Adaption and Vulnerability, 13.4.1 Natural ecosystems

Modelling studies show that the ranges occupied by many species will become unsuitable for them as the climate changes (IUCN, 2004). Using modelling projections of species distributions for future climate scenarios, Thomas et al. (2004) show, for the year 2050 and for a mid-range climate change scenario, that species extinction in Mexico could sharply increase: mammals 8% or 26% loss of species (with or without dispersal), birds 5% or 8% loss of species (with or without dispersal), and butterflies 7% or 19% loss of species (with or without dispersal).

And in 19.3.4 Ecosystems and biodiversity:

… up to 30% of known species being committed to extinction * (Chapter 4 Section 4.4.11 and Table 4.1; Thomas et al., 2004;

And in other summaries Table 4.1

Clearly the major difficulty with all this work, something that turned me off it but few acknowledge, is that the lack of skill of simulations of climate change renders fraudulent any claim to skill at the species habitat scale. Only now is the broader climate community finally starting to accept this about multi-decadal climate model predictions, such as contained in the 2007 IPCC WG1 the climate assessments. The NIPCC illustrates the broader opinion which should have been integral to the IPCC process from the beginning, IMHO.