Soil carbon sequestration simulated in CMIP6-LUMIP models: implications for climatic mitigation
Akihiko Ito, Tomohiro Hajima, David M. Lawrence, Victor Brovkin, Christine Delire, Bertrand Guenet, Chris Jones, Sergey Malyshev, Stefano Materia, Sonali McDermid, Daniele Peano, Julia Pongratz, Eddy Robertson, Elena Shevliakova, Nicolas Vuichard, David Wårlind, Andy Wiltshire, Tilo Ziehn
Abstract
Abstract Land-use change affects both the quality and quantity of soil organic carbon (SOC) and leads to changes in ecosystem functions such as productivity and environmental regulation. Future changes in SOC are, however, highly uncertain owing to its heterogeneity and complexity. In this study, we analyzed the outputs of simulations of SOC stock by Earth system models (ESMs), most of which are participants in the Land-Use Model Intercomparison Project. Using a common protocol and the same forcing data, the ESMs simulated SOC distribution patterns and their changes during historical (1850–2014) and future (2015–2100) periods. Total SOC stock increased in many simulations over the historical period (30 ± 67 Pg C) and under future climate and land-use conditions (48 ± 32 Pg C for ssp126 and 49 ± 58 Pg C for ssp370 ). Land-use experiments indicated that changes in SOC attributable to land-use scenarios were modest at the global scale, in comparison with climatic and rising CO 2 impacts, but they were notable in several regions. Future net soil carbon sequestration rates estimated by the ESMs were roughly 0.4‰ yr −1 (0.6 Pg C yr −1 ). Although there were considerable inter-model differences, the rates are still remarkable in terms of their potential for mitigation of global warming. The disparate results among ESMs imply that key parameters that control processes such as SOC residence time need to be better constrained and that more comprehensive representation of land management impacts on soils remain critical for understanding the long-term potential of soils to sequester carbon.