The construction of new fossil-fuel energy infrastructure implies a commitment to burn fossil fuels and therefore produce CO2 emissions for several decades into the future. The recent letter by Davis and Socolow (2014 Environ. Res. Lett. 9 084018) highlights the current and growing commitment to future emissions, and argues that this emission commitment should be accounted for at the time of new construction. The idea of accounting for future committed emissions associated with current energy policy decisions is compelling and could equally be applied to other aspects of the fossil-fuel supply chain, such as investing in the development of new fossil-fuel reserves. There is evidence, for example, that oil reserves are growing faster that the rate of extraction, implying a growing future emissions commitment that is likely incompatible with climate mitigation targets.

There is growing evidence that global temperature change responds approximately linearly to cumulative CO2 emissions. Consequently, we can define a quantity of allowable cumulative emissions (a global carbon-emissions budget) for a given climate target [2-7]. An important implication of this idea is that while the temperature change produced by a given amount of CO2 emissions is largely irreversible [8, 9], there is little additional future warming expected from CO2 already in the atmosphere [10]. Consequently, future global warming will be driven primarily by CO2 emissions that have not yet occurred [11].

However, our current dependence on fossil-fuel energy infrastructure represents a societal commitment to continue to burn fossil fuels, and therefore to continue to produce CO2 emissions. The inertia associated with technological change implies that some amount of future CO2 emissions and temperature change will occur, regardless of how quickly we move as to develop non-CO2 energy technologies. These committed future emissions were first quantified by Davis et al [12], and have been shown to comprise a substantial portion of the cumulative emissions budget that is required to maintain global warming below 2 °C [13].

The recent letter by Davis and Socolow [1] has now extended this analysis to show how the emissions commitment associated with fossil-fuel energy technology has grown since the middle of the 20th century. The world is continuing to build new CO2-producing energy infrastructure, and is doing so faster than the rate of retirement of old technology. This means that not only are annual emissions continuing to increase, but the future-emissions commitment associated with this energy infrastructure is also increasing as a result of today's energy policy decisions [1].

Davis and Socolow's work presents a framework for how this emissions commitment could be accounted for at the time that decisions are made to build new CO2-producing energy technology. Their analysis of this "commitment accounting" focuses on power plants fuelled by coal, natural gas and oil. In principle, however, the idea of commitment accounting could be applied to any investment in the use of fossil fuels as an energy source, from the installation of residential heating systems, to the construction of oil and gas pipelines, or even to the exploration for new fossil-fuel reserves.

The financial investment in the construction of a new power plant generates an economic commitment to use this technology. Similarly, the investment in exploration for new fossil-fuel reserves implies an intention to eventually extract and ultimately burn this fuel. The growth of a fossil-fuel reserve base over time can therefore be considered to represent a commitment to future CO2 emissions, whereas the extraction of fuel from that reserve represents the realization of those emissions. As long as the reserve base is growing faster than the rate of fuel extraction, the commitment to future CO2 emissions from the use of this fuel will increase.

Using historical data on fuel reserves and extraction [14], it is possible to quantify how this commitment has changed over time (figure 1(A)). For the case of oil, global oil extraction from 1980 to 2013 totalled 120 billion tonnes (Gt), which using a conversion factor of 3.07 tonnes of CO2 per tonne of oil [14], can be considered to represent realized emissions of approximately 370 Gt CO2. Over this same period of time, however, the oil reserve base also increased by 137 Gt of oil, which means that the oil reserves on average grew over this time at more than twice the rate of oil extraction. Consequently, the commitment to future emissions from oil use increased by 420 Gt CO2 since 1980. Given current (year 2013) proven oil reserves of 230 Gt of oil, the total potential for future emissions from this reserve is about 700 Gt CO2.

The linear relationship between cumulative CO2 emissions and temperature change then enables an estimate of the future warming that would occur if this fuel were used. Using a best estimate of 0.4 °C per 1000 Gt CO2 [6, 7], this means that: (1) oil extraction since 1980 represents 0.15 °C of global temperature change which has already occurred; (2) the growth of oil reserves since 1980 represents an additional potential warming of 0.17 °C that has not yet occurred; and (3) full combustion of the current oil reserve would represent 0.28 °C of future global warming.

What about other fossil-fuel reserves? Using similar data and conversion factors for natural gas and coal [14], there are currently 390 Gt CO2 in proven natural gas reserves1 , and 1700 Gt CO2 in coal2 . Were we to burn this entire reserve base, this would lead to 0.16 °C of warming from natural gas, and 0.68 °C from coal. Combined with oil (0.28 °C), this results in a total of 1.12 °C of potential future warming from current fossil-fuel reserves (figure 1(B)). This by itself, ignoring all other contributions to climate change, would bring the world very close to 2 °C of total global warming.

These calculations are of course approximate, and are subject to large uncertainty in both estimates of fossil-fuel reserve sizes, as well as in the climate response to CO2 emissions. However, it is nevertheless a useful exercise to think about how historical and current exploration for new fossil-fuel reserves represents an economic investment in and therefore commitment to burn this fuel in the future. As in the case of Davis and Socolow's power-plant analysis, we are currently expanding the fossil-fuel reserves faster than the rate of fuel extraction, leading to an increasing commitment to future CO2 emissions. Given the amount of carbon and potential emissions contained in the much larger fossil-fuel resource base, our analysis suggests that continuing to invest in exploration and extraction technologies to expand current proven reserves is inconsistent with a 2 °C climate target. Instead, we will very likely need to decrease the current commitment to future CO2 emissions by leaving some economically viable fossil-fuel reserves in the ground.

For references, see A growing commitment to future CO2 emissions at environmentalresearchweb's sister journal ERL.