Numerical model to simulate long-term soil organic carbon and ground ice budget with permafrost and ice sheets (SOC-ICE-v1.0)

Abstract

The degradation of permafrost is a large source of uncertainty in understanding the behaviour and projecting the future impacts of Earth's climate system. The spatial distributions of soil organic carbon (SOC) and ground ice (ICE) provide essential information for the assessment and projection of risks and impacts of permafrost degradation. However, uncertainties regarding the geographical distribution and estimated range of the total amount of stored carbon and ice are still substantial. A numerical soil organic carbon–ground ice budget model, SOC-ICE-v1.0, that considers essential aspects of carbon and hydrological processes in above-ground and subsurface environments and permanently frozen ground (permafrost) and land cover changes (ice sheets and coastlines) was developed to calculate the long-term evolution of local SOC and ICE. The model was integrated to cover the last 125 kyr – from the last interglacial to date for areas north of 50∘ N at 1∘ resolution – to simulate the balance between accumulation and dissipation of SOC and ICE. Model performance was compared with observation-based data and evaluated to assess allogenic (external) impacts on soil carbon dynamics in the circum-Arctic region on a glacial–interglacial timescale. Despite the limitation of forcing climate data being constructed on the basis of a single Greenland ice core dataset, the simulated results successfully reproduced temporal changes in northern SOC and ICE, consist with current knowledge. The simulation also captured regional differences in different geographical and climatic characteristics within the circum-Arctic region. The model quantitatively demonstrated allogenic controls on soil carbon evolution represented by a key parameter that reflects climatological and topo-geographical factors. The resulting circum-Arctic set of simulated time series can be compiled to produce snapshot maps of SOC and ICE distributions for past and present assessments or future projection simulations. Examples of 1∘ resolution maps for the Last Glacial Maximum and mid-Holocene periods were provided. Despite a simple modelling framework, SOC-ICE-v1.0 provided substantial information on the temporal evolution and spatial distribution of circum-Arctic SOC and ICE. Model improvements in terms of forcing climate data, improvement of SOC and ICE dynamics, and choice of initial values are, however, required for future research.