Response to “Romm Critique of COP21 Impact Deeply Flawed”

Dr. John Sterman, MIT Sloan School of Management
Andrew Jones, Climate Interactive
Dr. Lori Siegel, Climate Interactive


The blog post by Bjorn Lomborg, “Romm Critique of COP21 Impact Deeply Flawed”, includes multiple errors. The most significant ones include:

  • Criticizing Climate Interactive’s baseline scenario (RCP 8.5) as “highly exaggerated”,yet using the same baseline in the Global Policy article that supports the critique.
  • Confusing CO2 and CO2e, with significant implications for conclusions.
  • Erroneously comparing CI’s “business as usual” (i.e., unabated emissions) scenario with scenarios that include policy interventions that reduce emissions, again with significant implications for conclusions.

The details of these errors plus other specific comments are below.

Recap: Primary Deficiencies of the Analysis in “Global Policy”

A Climate Progress blog post by Joe Romm makes the following points regarding Bjorn Lomborg’s paper “Impact of Current Proposals”, published in Global Policy:

  1. The analysis ignores China’s pledge to peak emissions by 2030.
  2. The analysis does not adequately include the INDCs offered by nations that together constitute 53% of global emissions, including Canada, Japan, Australia, South Korea, Mexico, and Brazil.
  3. In the “Pessimistic” case, the analysis assumes that after the European Union and United States meet their 2030/2025 climate pledge, they will immediately reverse decades of policy-driven reductions with the result being rapid emissions growth that returns the emissions path to the reference scenario. Details and graphs are shown later.

Specific responses to statements in the blog post follow.

Reference or “Business as Usual” Scenarios

The “Romm Critique…” blog post states that [CI’s analysis] “depends almost entirely on a highly exaggerated baseline unsupported by mainstream analysis”


“The Climate Interactive analysis relies on an incredibly far-fetched ‘no action’ scenario where the planet will emit huge amounts of carbon”

The Global Policypaper itself (see Figure 11) uses the same baseline – both are both calibrated to the highly reputable, widely accepted IPCC AR5 RCP8.5 scenario.

US Analysis

“Indeed, about 70% of the suggested temperature reduction identified by Climate Interactive is completely spurious. We can see the size of the difference in the following figure”


The Global Policy paper makes two fundamental errors that renders the comparison in this graph erroneous.

First, the EMF scenarios report CO2 alone while the Climate Interactive scenario shows CO2e (CO2 equivalent) emissions, that is, the impact of CO2 and all other greenhouse gases (GHGs), including methane, N2O, and the f-gases. The EMF paper notes the distinction on page 37: “Covered CO2 is used here as the emissions variable for measuring cumulative reductions because several models do not include the non-CO2 gases.”


Second, unlike the Climate Interactive scenario, the EMF scenarios are not “business as usual” scenarios – thus their flat shape through the century. The EMF paper notes that the scenarios already include policies similar to the ones that would deliver upon the INDCs. They include “different baseline assumptions about recent and anticipated non-climate related regulatory policy changes.”

To check, we downloaded the EMF22 suite of scenarios for the US, just focusing on CO2 alone and reference or “business as usual” scenarios alone and found the CI scenario (“C-ROADS”) to fall within the EMF spread. See figure to the right.

The corrected graph shows how the blue line now lands amongst the other colored lines within the spread of scenarios.

Long Term Assumption

[The analysis in Global Policy] “extends its Paris promise every year for another 70 years.”

CI provides a set of scenarios that include BAU (no policy action), “INDC Strict” (full implementation of the INDCs through 2030, with no actions that nations have not yet pledged after that), and a set of scenarios with greater ambition.  These are fully documented here.


In contrast, the analysis in “Global Policy” does not truly “extend its Paris promise.” It assumes, to the contrary, in the main scenario reported (“Pessimistic”), that emissions snap back up to the original business as usual scenario of rapid emissions growth, ignoring decades of policy action, as seen in the orange line on the graph for the US, to the left (from the “Global Policy” paper, Figure 3).

And the red line on the graph for the EU, below (Figure 7).


In the “Optimistic” case emissions for the EU and US are also assumed to rise (the blue lines).


[CI uses] “an unbelievably high ‘no-action’ scenario”


The C-ROADS model uses historical data from the highly credible and widely cited sources, CDIAC and EDGAR. These sources tend to be slightly higher for recent years than the “National Communications” data and FAO data used in other analyses. Recent news that China emissions of coal are approximately 1 GT higher than previously reported supports this choice.  Further, as shown in the diagram to the left, recent global emissions as measured by CDIAC/EDGAR track most closely to the RCP8.5 scenario, which CI uses as a BAU case.

“Chinese ‘peaking’ promise is inconsistent with a robust analysis of Paris 2016-2030 promises”


China’s official INDC includes: “To achieve the peaking of carbon dioxide emissions around 2030.” An accurate assessment of INDCs needs to include all INDCs.  The CI “INDC Strict” case assumes that all INDCs will be fully implemented, but does not credit countries with emissions reductions they have not formally pledged to make.

[Climate Interactive research] “doesn’t include” [peaking]

The Climate Interactive analysis does include China peaking CO2 as per its INDC (as shown in red in the image to the right), but does not include peaking of the other greenhouse gases because they were not included in the INDC. Thus total emissions rise slowly post 2030, as shown in blue in the image to the right.

“Even if Chinese ‘peaking’ promise were included, results would not change significantly.”

This statement is false. Rerunning the INDC Strict scenario but assuming each country resumes BAU growth rates beyond the 2025-2030 target years for the INDCs yields a temperature change of 3.7° instead of 3.5°. If China simply cuts it emissions to 15% below BAU in 2030 (approximately equivalent to a 65% reduction in the carbon intensity of the economy) and then resumes BAU, the result is warming of 3.9°C.  Thus China’s commitment to achieving peak emissions by 2030 by itself reduces expected warming by 0.2 to 0.4° by 2100, even assuming that China’s emissions remain constant at the peak level throughout the remainder of the century.


“not have been published in a peer-reviewed journal”

The structure, assumptions and behavior of the C-ROADS model are described in this peer reviewed article:  Sterman, J., et al. (2013) Management Flight Simulators to Support Climate Negotiations. Environmental Modelling and Software 44: 122-135.

In addition, the C-ROADS model was assessed by a peer review panel, chaired by former IPCC head, Sir Robert Watson.  The full report of the panel is available online.