Reversibility, stability and overshoot of net-zero CO2 emissions across the FaIR ensemble: the eZEC case
The eZEC: the continued temperature response after net-zero CO2 emissions are achieved and sustained7. Here eZEC is evaluated over 40 years (between 2060 and 2100).
To find out the role of stabilization and overshoot for regional reversibility we used simulations that were different from each other and showed how global warming and overshoot affects GSAT. The simulations were done using the AERA61 which adjusts CO2 emissions successively every 5 years to reach stabilization and peak warming. In this setup, the remaining CO2-fe emissions budget is determined every 5 years based on the relationship of past global anthropogenic warming and CO2-fe emissions simulated by the model. The remaining anthropogenic CO2 emissions or removals are then computed assuming non-CO2 and land use change emissions following the RCP 2.6. The future CO2 emissions are redistributed using a function that is constrained to get to any given temperature level. The stabilization case details are given in the AERA model intercomparison simulation protocol.
The agreement between the distribution of ZEC across members of the FaIR ensemble is very good following the extended calibration of FaIR. 2a,b). The agreement of the modelled historical warming across the ensemble was reported. 1d). We can’t exclude high members of the ensemble that drive the tail of the distribution. 2b) . High NNCE outcomes also happen for moderate-highECS and ZEC outcomes.
Estimating the effective Zero Emissions Commitment (eZEC) allows us to separate the stabilization and decline components of NNCE. We use the warm up outcome of the original PROVIDE REN_NZCO2 scenario as the basis for evaluating eZEC.
where F2× is the effective radiative forcing from a doubling of CO2 and λ is the climate feedback parameter. F2× and λ are parameters that are both used directly in FaIR, and therefore ECS can be calculated for each ensemble member.
NNCE stabilization for post-peak cooling and carbon cycle feedback, with an application to the PROVIDE REN_NZCO2 scenario 51
$${{\rm{NNCE}}}{{\rm{stabilization}}}(n)=0\quad {\rm{if}}\;{T}{2060}(n) < 1.5\quad {\rm{else}}\,\frac{{\rm{eZEC}}(n)}{{{\rm{eTCRE}}}_{{\rm{down}}}(n)}$$
Each ensemble member demonstrates a different level of peak warming that depends on eTCREup (Fig. 2c). We calculate the cumulative NNCE (per ensemble member) that is necessary to ensure post-peak cooling to 1.5 °C in 2100 using
The warming of cumulative emissions until net-zero CO2 is captured with the effective Transient Response to Cumulative Emissions or eTCREup.
In our illustrative analysis, we assess the NNCE for the PROVIDE REN_NZCO2 scenario51. The REN_NZCO2 scenario follows the emission trajectory of the Illustrative Mitigation Pathway (IMP) REN from the AR6 of IPCC52,53,54 until the year of net-zero CO2 (2060 for this scenario). After the year of net-zero CO2, emissions (of both GHGs and aerosol precursors) are kept constant.
where ({\Delta E}{{{\rm{CH}}}{4}}) is the change in the emission rate of ({E}{{{\rm{CH}}}{4}}) over the Δt preceding years; H is the CH4 emission rate for the year under consideration; r and s are the weights given to the impact of changing the CH4 emission rate and the impact of the CH4 stock. We used t as follows following ref.68. Because of the dependency on the historical trajectory of the emission and carbon cycle feedback, the values of r and s are scenario-dependent. Here we use r = 0.68 and s = 0.32 (the values used in ref. 68 for RCP2.6), with H = 100 years, GWP100 of 29.8 for permafrost and GWP100 of 27.0 for peatland18.
rmCO_2text- mathrmwe H+s\times {E}{{{\rm{CH}}}{4}}\right)$$
Source: Overconfidence in climate overshoot
A running average for the smoothed GMST time series and its application to climate modelling for the SSP5-34-OS and SSP1-19 scenarios
We apply a running average to the GMST time series. In each simulation run, we identify peak warming as the year in which this smoothed GMST reaches its maximum. Next, we select the years before and after peak warming in which the smoothed GMST is closest to −0.1 K and −0.2 K below peak warming. There is a substantial, model-dependent asymmetry in the average time between the rate of change in GMST before and after peak warming (see ref. 5 for an overview). We average yearly temperatures and precipitation for 31 years around the interest areas. The averaged 31 year periods are used to derive the ensemble median for the 12 ESMs.
In order to analyse climate projections for the SSP5-34-OS and SSP1-19 scenarios, we used 12ESMs.
We compute regional averages for the land and ocean. WNEU corresponds to land grid cells in western central Europe (WCE) and northern Europe (NEU). NAO45 corresponds to ocean grid cells in the North Atlantic region above 45° N (see encircled area in Fig. 3e,f The land regions are AMZ and WAF.
Land cover changes are not included in the two protocols. This points to an implicit assumption that the additional CDR in these simulations is achieved using technical options with little to no land footprint such as Direct Air Capture with CCS (Extended Data Table 2). If the amount of CDR was achieved using land-based CDR methods, we would probably see climate changes because of the land cover changes alone. The regional climate differences resulting from different CDR strategies should be explored in future modelling efforts.
Governments and industry must have a laser-like focus on the risks ahead and how to mitigate them. It means helping communities to become more resistant to shocks by aggressively cutting emissions. To wait and scrub the atmosphere later is to court disaster — for people and the planet.
It’s not that carbon-removal methods don’t work. Some do. It is the simplest way to plant trees. There are more complex measures like taking carbon directly from the atmosphere. The scientists think that 400 gig of carbon would need to be removed from the atmosphere by the end of the century to limit warming to 1.5 C. In emissions terms, that is equivalent to running the US energy industry in reverse for around 80 years.