UNEP EMISSIONS GAP REPORT
UNEP EMISSIONS GAP REPORT

 

Also in this chapter:

2. Which emission pathways are consistent with a 2°C or 1.5°C temperature limit?

Lead authors: William Hare; Jason Lowe; Joeri Rogelj; Elizabeth Sawin; Detlef van Vuuren
Contributing authors: Valentina Bosetti; Tatsuya Hanaoka; Jiang Kejun; Ben Matthews; Brian O’Neill; Nicola Ranger; Keywan Riahi

2.1 INTRODUCTION

This chapter identifies future emission pathways that are consistent with a 2° C or 1.5° C temperature limit. Many scenarios and pathways for annual global emissions of greenhouse gases have been published in the scientific literature to explore possible long-term trends in climate change. This literature has been used in this report to understand the kind of pathways consistent with the goal of limiting global temperature increase to less than 2° C or 1.5° C above pre-industrial levels.

Among the different studies of future emission pathways, two main types can be identified. The first type is produced by integrated assessment models (IAM), which simulate both future climate and future socio-economic systems, including the emissions of greenhouse gases from industry and power generation, agriculture, forestry and other land use activities (see for example Clarke et al. 2009, Edenhofer et al. 2010, van Vuuren et al. 2007). IAMs take into account assumptions about technological and economic constraints and so, to some extent, provide a view on what are “feasible” emission reductions. The second type of pathway, described here as “stylized”, explores more directly the relationship between emissions and temperature, for example by making assumptions about the timing and magnitude of peak emissions and rates of reduction23 following the peak. These are pathways produced by models that do not explicitly simulate change in the energy system or feasibility of emission reduction rates.Stylized pathways are designed to better understand the temperature outcomes resulting from emission pathways computed by carbon cycle and climate models, without making assumptions about how those emissions are produced (see for example Lowe et al. 2009, Meinshausen et al. 2009).

Although both approaches provide important insights and findings, only results from IAMs are used here for quantitative analysis, unless otherwise stated.

Scenarios published by IAMs in the literature mostly look into optimal pathways to achieve a certain long-term target and not into the question of what emission range in 2020 would achieve a temperature limit. For this reason, we have assembled a large set of scenarios computed with various objectives in mind, and have tested them to see if they are consistent with temperature limits. The combination of these scenarios provides insight into the full range of 2020 emissions consistent with long-term temperature limits. It is possible that other feasible pathways will be identified by modelling groups, once they begin to run their models to explore the full 2020 emissions range.

Although IAM studies have paid little explicit attention to the question of the range of 2020 emissions consistent with temperature limits, there are some studies of stylized pathways that have done this (Bowen and Ranger 2009, Meinshausen et al. 2009).

In our quantitative assessment of IAM results we have attempted to take the differences between studies (in terms of uncertainties of various input assumptions and different approaches) into account by re-analysing the results of these studies using a common set of assumptions about base year emissions, coverage of non-CO2 gases, carbon cycle assumptions and interpretation of climate goals (as explained in Box 2a ). These re-analysed pathways have been evaluated in terms of their consistency with a 2° C and 1.5° C limit. An important factor here is that projections of the future climate all contain uncertainty (Meehl et al. 2007). This means that when discussing the possibility of satisfying a particular temperature limit, it is necessary to express the result in terms of a probability. As explained in Box 2a, the MAGICC model (Meinshausen et al. 2008) has been used here to take into account some of this uncertainty.