What We Publish

Dependence of climate and carbon cycle response in net zero emission pathways on the magnitude and duration of positive and negative emission pulses

K. U. Jayakrishnan, Govindasamy Bala, and Ken Caldeira · Earth's Future 12, e2024EF004891 · 2024

Key finding. Across nine simulations emitting 1,000, 2,000 or 5,000 gigatonnes of carbon over 150, 250 or 500 years and then removing exactly as much, the climate returns close to preindustrial only on millennial timescales — at the moment net cumulative emissions reach zero, surface air temperature is 0 to 1 degree Celsius higher and atmospheric CO2 is 12 to 29 parts per million lower than preindustrial.

Four time-series panels from year 0 to 2000. The top row is atmospheric carbon dioxide in parts per million and the bottom row is change in surface air temperature in degrees Celsius. The left column fixes the emitted total at 2,000 gigatonnes of carbon and varies the pulse duration over 150, 250 and 500 years; the right column fixes the duration at 250 years and varies the total over 1,000, 2,000 and 5,000 gigatonnes. Every curve rises to a peak and falls back, with vertical lines marking the end of each zero net emission cycle. The carbon dioxide curves settle slightly below their starting dashed line while the temperature curves settle slightly above zero.
Emit a quantity of carbon, then remove exactly as much, and the two curves do not come back together. Carbon dioxide undershoots its starting value while temperature is still above it — the ocean is returning heat it absorbed on a schedule the removals cannot dictate. Figure 3 from Jayakrishnan, Bala and Caldeira (2024), Earth's Future 12, e2024EF004891. Reproduced under CC BY-NC-ND 4.0. Used unmodified, as the licence requires.

What question did this research address?

Net-zero targets assume a symmetry — that carbon put into the atmosphere can later be taken out, and the climate will follow the same path back. The ocean makes that questionable, because it takes up heat and carbon on timescales of centuries.

This paper asked what actually happens when emissions are exactly cancelled by removals, and whether the answer depends on how large the pulses are and how long they take.

What did we find?

Nine stylized runs of a coupled climate and carbon cycle model pair three emission magnitudes with three durations, each followed by identical cumulative negative emissions so that net cumulative emissions end at exactly zero.

On millennial timescales the climate system does return close to its preindustrial state, and which pathway it took to get there stops mattering.

On the centennial timescale that policy operates over, it does not. When net emissions reach zero, temperature is still 0 to 1 degree Celsius above preindustrial while atmospheric CO2 has been pulled 12 to 29 parts per million below it — the air is cleaner than it started and the planet is still warmer.

Both offsets grow with the magnitude and the duration of the pulses, so a bigger, slower excursion leaves a larger residual at the moment the books balance.

Hysteresis between global mean temperature and cumulative emissions behaves differently again: it increases with the magnitude of the pulses but *decreases* with their duration.

The asymmetry is the ocean. Its thermal and biogeochemical inertia means heat and carbon taken up during the emission phase come back on their own schedule, not on the schedule of the removals.

Why does it matter?

It puts a number on the cost of delay in a form net-zero accounting hides. Removing as much carbon as was emitted does not undo the warming on any timescale a government plans over, and the shortfall grows the longer the excursion lasts.

The direction of the residual is counterintuitive and worth stating plainly: you can end up with less CO2 in the air than you started with and still be warmer, because the ocean is still giving back heat it absorbed decades earlier.

The practical conclusion is about timing rather than totals. Earlier emission reduction means smaller and shorter pulses in both directions, and therefore a smaller climate and carbon-cycle residual on centennial timescales — even holding the cumulative budget fixed.

Citation

K. U. Jayakrishnan, Govindasamy Bala, and Ken Caldeira (2024). Dependence of climate and carbon cycle response in net zero emission pathways on the magnitude and duration of positive and negative emission pulses. Earth's Future 12, e2024EF004891.

Related