Neglecting harvest and its uncertainty results in considerably reduced sensitivity to total LULCC uncertainty for simulations started in 1700 and 1850 (not shown). Interestingly, the reduction in cumulative net LULCC flux is largest in HI850NoH if considering the whole simulation (not shown), but from 1850 (Fig. 3), LO850NoH and REG850NoH show the largest reduction by omitting wood harvest. The analysis of the contributions from the four LULCC activities to the total net LULCC flux sensitivity reveals that (1) LULCC uncertainty from harvest causes largest sensitivity in the cumulative net LULCC flux, followed by equal contributions from abandonment and pasture and negligible sensitivity due to crop uncertainty. For harvest, the sensitivity is asymmetric; i.e. the net LULCC flux due to harvest in the HI scenario deviates further from REG than in the LO scenario. (2) Uncertainties in wood harvest cause large sensitivity to starting year of the simulation (StYr), as well as to IC and Trans in the artificial LULCC experiments.
Secondary actions
From 1300 onwards, and for most of the time series, these two LULCC activities affect roughly the same area as wood harvest, though wood harvest exhibits larger temporal variability. Pasture expansion and harvest on primary forested land are only relevant from around 1700 onwards and affect less area than the other LULCC activities. Since the net flux from LULCC cannot be directly measured, we can only rely on values calculated by models, for example dynamic global vegetation models (DGVMs) and bookkeeping models. Bookkeeping models (Houghton, 2003; Houghton and Nassikas, 2017; Hansis et al., 2015) combine observation-based carbon densities with LULCC estimates to determine the net LULCC flux.
Achieving a consistent estimate of the terrestrial carbon budget and its environmental and land-use components
Globally, we find a reduction in SLAND due to LULUCF by 0.7 (0.3, 1.3) GtC yr−1 in 2012–2021 and by 33 (8, 62) GtC cumulatively in 1850–2021 (RSS term in Fig. 2, Table 2). This corresponds to a 23% (8%, 33%) overestimation of the sink if assuming pre-industrial land cover in Airbnb Accounting and Bookkeeping 2012–2021. LASC (RSS plus the environmental contribution to ELUC; see Table 1) amounts to 1.0 (0.5, 1.7) GtC yr−1 averaged over 2009–2018 (Supplementary Table 1) and thus accords with findings from ref. 16 (LASC of 0.8 ± 0.3 GtC yr−1 in 2009–2018).
Data
The dataset captures the challenge of reconstructing the LULCC of the past.LUH2 is the land-use dataset that is – besides many other studies – also applied in CMIP6 (Eyring et al., 2016) for simulations with process-based DGVMs, like in LUMIP (Lawrence et al., 2016). Our findings and discussions regarding DGVM studies are therefore also informative for the interpretation of CMIP6 results. BLUE is a data-driven bookkeeping model (Hansis et al., 2015) used in the GCB for LULCC flux estimates (Friedlingstein et al., 2019).
We use therefore spatially explicit values in the model simulations (as in Fig. A1), but they aresummarised as spatially averaged values in Table A1. Parameters in BLUE and HN2017 are defined on a PFT basis, but HN2017 distinguishes 20 PFTs (3 of them desert PFTs), while BLUE distinguishes 11 PFTs. Most HN2017 PFTs can be aggregated into the often morebroadly defined BLUE PFTs but some of the PFTs in BLUE do not correspond toHN2017 PFTs (e.g. summergreen shrubs) (Table A1). A map of the PFT distribution from HN2017 is not available, as the PFT fractions are definedon a per-country basis. When aggregated globally, the values of BLUE andHN2017 show good agreement in the global extent of croplands (15.3 and 13.8 million km2 for HN2017 and BLUE, respectively, in 2015) and forests (39.9 and 40.9 million km2 for HN2017 and BLUE, respectively, in 2015).
Louise Chini
Such long timescales are needed, however, to capture the slow dynamics of decay and regrowth and thus to capture legacy fluxes accurately. For the last decades, however, more detailed data have become available than those currently used in the models of the global carbon budgets, such as global sets of dynamic carbon-storage factors (Mason Earles et al., 2012) that define a larger number of product pools and time-varying fractions of allocation. A crossing point in the cumulative net LULCC flux between two scenarios can occur if the rate of LULCC varies differently with time in both scenarios.This can, for example, happen when the setups have a similar beginning and end distribution of land cover, as is the case in the LUH2 dataset. The simulation with initially larger number of LULCC activities produces an initially steeper increase of the cumulative net LULCC flux and a weaker increase towards the adjusting entries end, which can potentially imply a crossing point.
This means that it is of little importance for estimates of the net LULCC flux over recent years when a simulation was started, but it is important for cumulative fluxes, with relevant implications for comparisons of the GCB and CMIP6 model simulations. However, not accounting for gross transitions and wood harvest, as is sometimes still the case in DGVMs, can cause even larger differences between model estimates. Finally, it should be noted that the two alternative LULCC scenarios (low and high land-use scenarios) produce relatively smaller or larger estimates of the net LULCC flux than the LUH2 baseline scenario depending on the time period considered.
Regional websites
As the number and definition of PFTs is not standardized in DGVMs, not all PFTs can be directly matched to the BLUE PFTs. In some cases, a spatial mask is applied to PFTs of some DGVMs to fit the spatial extent of the BLUE PFTs (see Supplementary Fig. 13). In other cases, particularly for shrub PFTs, the carbon density of a BLUE PFT is constructed by weighting different PFTs of a DGVM following the cross-walking table by ref. 54 (see their Table 2).
- Note that the experiment groups (LULCC and StYr) are now combined as the presented values show the absolute net LULCC flux.
- As future environmental conditions are expected to diverge further from the present-day state, ELUC estimates are projected to grow further apart in future decades9, yet decisively depending on the future evolution of CO2 concentrations and climate change41.
- Potential natural land refers to the land cover as it would exist without human interventions.
- As a first step, we present the bookkeeping model BLUE used in this study.Then the LUH2 dataset, its high and low LULCC scenarios as well as various future scenarios are introduced.
Despite delivering a conceptually consistent carbon budget, our approach contains several sources of uncertainty, mainly stemming from the limited reliability of the DGVM carbon densities, our processing of carbon densities as global averages, and the LULUCF data. First, DGVMs show a large spread in their sensitivity of carbon fluxes to environmental changes, e.g., due to simplified representations of biogeochemical processes and differences in assumptions on plant productivity, plant allocation, nutrient availability, and carbon turnover times26,30,31,32,33,34. Since we scale the BLUE carbon densities with the DGVM carbon density ratios, these uncertainties of DGVMs propagate to our ELUC and SLAND estimates.