The following animation is for a "below 1.5°C" pathway, or radiative forcing of 1.9W/m² in 2100. This scenario has more cumulative negative emissions than the original tweet, as I have linked more closely to actual scenario data.

Approximate pathways to ~1.7°C in 2100

The "well below 2°C" pathway has a radiative forcing of 2.6W/m² in 2100, and actually has a median temperature of about 1.7°C in 2100 (reference, data). One can debate if this is "well below 2°C", but I am happy to argue it is.

To go from 1.4°C to 1.7°C substantially increases the carbon budget and decreases the need for negative emissions. It appears as though there is little variation in short-term pathway, but the percentage reduction in 2040 varies from 40% to 60% of 2017 levels. Relatively small in some senses, but quite some effort!

Approximate pathways to ~2.2°C in 2100

The pathway to 2.2°C in 2100, with temperature still rising in 2100, is consistent with a radiative forcing of 3.4W/m² in 2100 (reference, data). This scenario is not so far over 2°C, and in comparison to 1.4°C and 1.7°C looks like a "walk in the park". This is a point I tried to make in an earlier blog, a 0.5°C difference in mitigation is a huge difference in the carbon budget.

Many know that I am a tad sceptical if we can keep "well below 2°C", but hey, 2.2°C no problem!

It is important to reflect on how hard each extra 0.1°C is in terms of mitigation! Using the median Transient Climate Response to cumulative carbon Emissions (TCRE) of 1.65°C/3667GtCO₂, gives 200GtCO₂ for each extra 0.1°C (assuming constant non-CO₂ emissions).

Though, do note, temperatures are still rising in this pathway, albeit slowly. We don’t need much negative emissions, if any, for a reasonable shot at 2.2°C in 2100. And, given all the constraints in the world (poverty reduction, geopolitical tensions, greed, etc), then we should be happy to keep temperatures around 2.2°C with the option to go lower still open?

And, here are all the figures together for 1.4°C, 1.7°C, and 2.2°C to allow an easier comparison of the differences. (Download GIF here)

Method (only read if you are into details)

The challenge is how to distribute a cumulative emissions budget over time, when net-negative emissions are allowed. This is to extend a weakness of an earlier study, that did not allow net negative emissions (though, did allow gross negative emissions that were less than gross positive emissions).

The general approach is to work independently with the gross positive and gross negative emissions. By definition, cumulative gross emissions, whether positive or negative, can only grow over time: they are monotonically increasing (or decreasing). The global burden can then be distributed for the gross positive and gross negative emissions, and the sum of the two leads to net emissions, which may be either positive or negative at different points in time.

The distribution of the allocated budgets over time requires the development of stylised functions of gross positive and negative emissions.

For gross positive emissions, we follow our earlier workdistributing positive emissions, but it is extended to allow ‘residual’ positive emissions in 2100. These are emissions that cannot, except at presumed prohibitive cost (in whatever way cost might be measured), be reduced to zero, for example certain industrial or transport emissions (residual emissions or difficult-to-eliminate emissions).

For the pathway of gross negative emissions, we use a scaled cosine function, growing from zero following an S-shape typical in technology deployment, reaching a steady state in 2100, indicating some form of technological maturity or saturation plateau.

The functional forms that we have used are stylised. Real pathways will certainly not be so smooth or monotonic, but stylised functions are nonetheless simple and therefore useful. It is straightforward to test the sensitivity of the analysis to different functional forms and different parameters (e.g., residual emissions) or to. The purpose of this analysis is to demonstrate and explore a framework, and stylised functions give more flexibility.

The following figure demonstrates the global pathways with the same cumulative gross positive and negative emissions to 2100 as the actual IMAGE scenario but using the stylized functions. The pathways for gross positive emissions are very similar, indicating that the IMAGE model’s mitigation pathway for this scenario leads to a pathway close to an exponential decline with residual emissions, the basis of our method.

The pathways for gross negative emissions differ somewhat, with IMAGE in this scenario allowing much more rapid deployment of BECCS, reaching almost full deployment in only 20 years (from 2030 to 2050). This rapid deployment means that more CO₂ is removed from the atmosphere in the first half of the century, relative to our pathway, such that less is required in the latter half.

The results for the MESSAGE model, based on Rockström et al (2017), are remarkably similar. It would be hard pressed to convince anyone we have not done a curve fit here, but we haven’t. All we do is take the cumulative budgets from the integrated assessment model, and distribute over time with stylised functions!

While the integrated assessment model results are based on an internally consistent modelling framework, the stylised functional forms used here allow rapid analysis and greater flexibility. Using the methodology developed here, it is straightforward to quickly explore the consequences of differences choices of negative emissions, while retaining the same carbon budget for a given temperature level.

See a presentation of some of this work on SlideShare.

Allocating negative emissions to countries from Glen Peters