Strong dependence of CO₂ emissions from anthropogenic land cover change on initial land cover and soil carbon parametrization

Abstract

The quantification of sources and sinks of carbon from land use and land cover changes (LULCC) is uncertain. We investigated how the parametrization of LULCC and of organic matter decomposition, as well as initial land cover, affects the historical and future carbon fluxes in an Earth System Model (ESM). Using the land component of the Max Planck Institute ESM, we found that the historical (1750-2010) LULCC flux varied up to 25% depending on the fraction of biomass which enters the atmosphere directly due to burning or is used in short-lived products. The uncertainty in the decadal LULCC fluxes of the recent past due to the parametrization of decomposition and direct emissions was 0.6 Pg C yr −1 , which is 3 times larger than the uncertainty previously attributed to model and method in general. Preindustrial natural land cover had a larger effect on decadal LULCC fluxes than the aforementioned parameter sensitivity (1.0 Pg C yr −1 ). Regional differences between reconstructed and dynamically computed land covers, in particular, at low latitudes, led to differences in historical LULCC emissions of 84-114 Pg C, globally. This effect is larger than the effects of forest regrowth, shifting cultivation, or climate feedbacks and comparable to the effect of differences among studies in the terminology of LULCC. In general, we find that the practice of calibrating the net land carbon balance to provide realistic boundary conditions for the climate component of an ESM hampers the applicability of the land component outside its primary field of application.

Interactions between land use change and carbon cycle feedbacks

“…In any case, the results here suggest that carbon feedbacks in coupled models are quite sensitive to the LULCC time series prescribed in a particular scenario. This result compliments previous studies highlighting the importance of the initial carbon amounts, parameterization details, and methodological choices for carbon cycle impacts of LULCC [e.g., Goll et al, 2015;Hansis et al, 2015]. We are also attempting to simulate a process that occurs at a much finer resolution using 1°× 1°grid boxes, which is unlikely to capture important interactions-contributing to additional uncertainties.…”

Effects of bryophyte and lichen cover on permafrost soil temperature at large scale

“…JSBACH uses a process-based approach to simulate both physical and biochemical ecosystem functions, such as the exchange of energy and water at the land surface and carbon fluxes between atmosphere, vegetation and soil which are determined by photosynthesis and respiration. The soil part of JSBACH includes a soil carbon model (Goll et al, 2015) and a five-layer soil hydrology scheme (Hagemann and Stacke, 2015). It has been extended by Ekici et al (2014) to take into account both several dynamic snow layers and the latent heat of fusion associated with freezing and thawing, which is the basis for simulating permafrost extent and active layer thickness.…”

Process-level improvements in CMIP5 models and their impact on tropical variability, the Southern Ocean, and monsoons

“…A recalibration of the cloud processes resulted in a climate sensitivity of about 3 K of the new model system, which is about in the middle of the range of climate sensitivities spanned by the CMIP5 models. JSBACH 3.0 comprises several bug fixes, a new soil carbon model (Goll et al, 2015), and a five-layer soil hydrology scheme (Hagemann and Stacke, 2015) replacing the previous bucket scheme.…”