From source to sink: past and projected changes in carbon sequestration in the global forest sector

To reduce atmospheric concentrations of CO2, policymakers are increasingly turning to the world’s forests. Activities related to land use and forestry were responsible for 24% of global greenhouse gas emissions in 2010, primarily due to deforestation and forest degradation (IPCC, 2014). Yet, at the same time, forests have the ability to remove CO2 from the atmosphere by sequestering carbon in biomass, dead organic matter, soils, and long-lived wood products, which can turn the forest sector from a net source of CO2 to a sink (FAO, 2014). Consequently, forests are likely to play a major role in future initiatives to combat climate change.

To investigate this issue further, we modify an economic model of the global forest sector to estimate the carbon mitigating potential of the world’s forests to 2065 for 180 countries assuming future socioeconomic trends that do not change markedly from historical patterns, consistent with the IPCC-SSP2. Forest carbon pools were broken down into four categories; (i) above-ground and below-ground biomass, (ii) forest soil, (iii) dead wood and litter, and (iv) harvested wood products. Changes in forest carbon storage were driven by the dynamic relationship between endogenously determined timber harvest, wood product consumption, evolving forest biomass stock, forest area change and exogenous demographic and income changes. The results suggested that the forest sector was a net carbon source of approximately 3.6 GtCO2e yr-1 in 1992, decreasing to 2.4 GtCO2e yr-1 in 2014 (average rate: -0.05 GtCO2e yr-1), in general agreement with previous historical assessments. In the projections, the global forest sector achieved a net zero carbon balance by the year 2025, but with large variations by region and country. By 2030, the world’s forest sector became a net carbon sink of 1.5 GtCO2e yr-1, and eventually of 6.8 GtCO2e yr-1 by 2065.

Although there are few forecasts of carbon sequestration and emissions for the global forest sector, other studies have made an assessment of past levels and trends. These include the latest UNFCCC country reports (UNFCCC, 2017), the FAOSTAT for forest land (FAO, 2017), the IPCC Fifth Assessment Report (AR5) Working Groups (WG) I and III data (IPCC 2013, 2014), and Grassi et al. (2017) who also provide a complete review of the available data.

While all the data showed similar trends between 1992 and 2010, there were large differences in levels. We estimated that net forest emissions averaged 3.11 (standard error ± 0.33) GtCO2e yr-1 between 1992 and 2010, while the IPCC AR5 reports higher (but not significantly different given the standard errors) average emissions of 4.43 (standard error ± 0.92) GtCO2e yr-1. The IPCC AR5 reports on emissions through a similar book keeping method based on changes in land cover, forest growth, and some forms of management, but has been criticized for the inconsistent treatment of forest area across countries as well as limitations associated with having to reconcile its data with its own global CO2 emission estimates (Grassi et al. 2017). The FAOSTAT forest land emissions dataset estimates net carbon stock changes in above and below ground biomass, and forest land converted to other land uses, concluding that net forest emissions average 1.75 ± 0.39 GtCO2e yr-1 between 1992 and 2010. This is less than the present study as expected due to the narrower coverage of forest carbon pools, and is not entirely consistent with the implied emissions factors reported by the FAO FRA (2015) used in this study.

The data sets reported by the UNFCCC and Grassi et al. (2017) rely on country level reporting. Between 1992 and 2010, the UNFCCC finds that net forest emissions average 0.73 ± 0.51 GtCO2e yr-1, while Grassi et al. (2017) estimate 1.26 ± 0.70 GtCO2e yr-1. Both are less than the present study and significantly so, statistically. However, these low estimates rely on selected countries who self-report land use and forestry related emissions. The latter analyse 68 countries, covering 83% of the global forest area and only 78% of their emissions, while many countries carbon flux was assumed to be zero (Grassi et al., 2017).