Offsetting unabated agricultural emissions with CO2 removal to achieve ambitious climate targets

approach would yield insights into the price-driven decrease in emissions, which would affect the amount of CDR required and hence the tax price itself; this, however, introduces substan- tial additional uncertainties associated with the other future determinants of agricultural prices over the long time period of our study, and is beyond our scope. Results Our first results concern the two different approaches to offset additional agricultural methane and nitrous oxide emissions relative to the SSP1-2.6 scenario. Fig 2 portrays the CDR rates and the resulting total CO2 emissions required to offset agricultural methane emissions under the two different approaches described above, “GWP100” and “ERF”. The offsetting policies start in year 2020 and envision additional CDR compared to that already present in the RCP2.6 (dark grey line). S1 Fig shows the additional CDR due to the offsetting scheme only. Both offsetting approaches advance the CDR onset, and have greater annual CDR rates, compared to the SSP1-2.6 scenario. In SSP1-2.6, CDR rates surpass 0.1 GtC/yr in 2030, whereas this occurs already in 2020 for all other scenarios. There are marked differences between the CDR timing and rates under the two offsetting approaches, for each of the emis- sions scenarios. With the GWP100-based offsetting approach, additional CDR rates reach maximal values of 0.6 GtC/yr in the constant emissions scenario and 2.9 GtC/yr in the worst- case scenario, by the end of the century. Under the ERF approach, additional CDR rates peak

higher and sooner, increasing the total CDR use by almost a factor of two compared to the GWP100-based approach. They reach their maximum values at mid-century, at 1.3 GtC/yr for the constant emissions and 3.5 GtC/yr for the worst-case scenario, and then decline slightly in the second half of the century. The ERF offsetting approach, by definition, follows RCP2.6 ERF levels and temperature changes in every year (S2 Fig). It avoids additional warming by up to 0.38˚C under the worst- case and 0.10˚C under the constant agricultural scenarios. By contrast, the GWP100 offsetting approach fails to fully compensate for the additional methane forcing. As shown in Fig 3, around half of the additional temperature rise (0.15˚C under the worst-case and 0.05˚C under the constant scenarios) remains in 2100 after offsetting according to the GWP100-based approach. While this deviation is smaller than the 95% spread of the ensemble simulations, it points to a systematic under-estimation of offsetting rates determined via GWP100 during the Fig 2. Total CO2 emissions and CDR rates under the two offsetting approaches. CO2 emissions under the reference scenario (SSP1-2.6) and under a policy scenario in which alternative agricultural emission pathways (constant in orange and worst-case in blue) are offset via CDR. Light coloured lines correspond to the total CO2 emissions (including CDR) while darker coloured lines correspond to the total CDR rates. We show the results for two different offsetting approaches: (a)

“GWP100” based on the GWP100 metric and (b) “ERF” based on the explicit computation of CDR rates required to achieve the RCP2.6 methane ERF. https://doi.org/10.1371/journal.pone.0247887.g002 PLOS ONE Offsetting future agricultural emissions with CO2 removal PLOS ONE | https://doi.org/10.1371/journal.pone.0247887 March 17, 2021 8 / 19