Understanding how decreases in CO₂ emissions would affect global temperatures has been hampered in recent years by confusion regarding issues of committed warming and irreversibility. The notion that there will be additional future warming or “warming in the pipeline” if the atmospheric concentrations of carbon dioxide were to remain fixed at current levels has been misinterpreted to mean that the rate of increase in Earth’s global temperature is inevitable, regardless of how much or how quickly emissions decrease. Further misunderstanding may stem from recent studies showing that the warming that has already occurred as a result of past anthropogenic carbon dioxide increases is irreversible on a time scale of at least 1000 years. But irreversibility of past changes does not mean that further warming is unavoidable.

The climate responds to increases in atmospheric CO₂ concentrations by warming, but this warming is slowed by the long time scale of heat storage in the ocean, which represents the physical climate inertia. There would indeed be unrealized warming associated with current CO₂ concentrations, but only if they were held fixed at current levels.

If emissions decrease enough, the CO₂ level in the atmosphere can also decrease.

My comment: The CO₂ level in the atmosphere will go down from now on because we’re at peak oil, coal, and natural gas production — or will be soon.

This potential for atmospheric CO₂ to decrease over time results from inertia in the carbon cycle associated with the slow uptake of anthropogenic CO₂ by the ocean. This carbon cycle inertia affects temperature in the opposite direction from the physical climate inertia and is of approximately the same magnitude.

Because of these equal and opposing effects of physical climate inertia and carbon cycle inertia, there is almost no delayed warming from past CO₂ emissions. If emissions were to cease abruptly, global average temperatures would remain roughly constant for many centuries, but they would not increase very much, if at all. Similarly, if emissions were to decrease, temperatures would increase less than they otherwise would have.

Thus, although the CO₂-induced warming already present on our planet—the cumulative result of past emissions—is irreversible, any further increase in CO₂-induced warming is entirely the result of current CO₂ emissions. Warming at the end of this century and beyond will depend on the cumulative emissions we emit between now and then. But future warming is not unavoidable: CO₂ emissions reductions would lead to an immediate decrease in the rate of global warming.

Why, then, are many different near-term projections of CO₂-induced warming very similar? These modeled estimates are similar because even socioeconomic scenarios that produce very different cumulative emissions by the end of this century are not very different over the next two decades (figs. S1 and S2). The climate system physics implies that further increases in warming could in principle be stopped immediately, but human systems have longer time scales. Carbon-emitting infrastructure is designed to benefit human-kind for many decades; each year’s additional infrastructure implies added stock intended to last and emit CO₂ for many decades. It is this dependence on CO₂-emitting technology that generates a commitment to current and near-future emissions.

The strong dependence of future warming on future cumulative carbon emissions implies that there is a quantifiable cumulative amount of CO₂ emissions that we must not exceed if we wish to keep global temperature below 2°C above preindustrial temperatures. Several recent analyses have suggested that total CO₂ emissions of ∼1000 Pg C (∼3700 Pg CO₂; 1 Pg = 10¹⁵ g) would give us about even odds of meeting the 2°C target (9–12). To meet such a target given historical emissions would mean that the world has roughly half of the allowable emissions budget remaining. This is equivalent to 50 years of emissions at current levels and carries the implication that the longer we delay before beginning to decrease emissions, the faster the rate of decrease must be to stay within this total allowable budget.

Given the irreversibility of CO₂-induced warming, every increment of avoided temperature increase represents less warming that would otherwise persist for many centuries. Although emissions reductions cannot return global temperatures to pre-industrial levels, they do have the power to avert additional warming on the same time scale as the emissions reductions themselves. Climate warming tomorrow, this year, this decade, or this century is not predetermined by past CO₂ emissions; it is yet to be determined by future emissions. The climate benefits of emissions reductions would thus occur on the same time scale as the political decisions that lead to the reductions.