A simulation model of the carbon cycle in land ecosystems (Sim-CYCLE): a description based on dry-matter production theory and plot-scale validation

Abstract

In this paper, we present a new model of the terrestrial carbon cycle (Sim-CYCLE), with the objectives of retrieving the carbon dynamics of various terrestrial ecosystems and estimating their response to global environmental change. The model can be characterized in three ways. (1) It is a compartment model. Ecosystem carbon storage is divided into five compartments; foliage, stem, root, litter, and mineral soil. This approach made the model simple and sound, and allowed us to run the model on a broad scale; indeed, the simulation in this paper was performed using data available at the global scale. (2) It is a process-based model. Sim-CYCLE estimates net primary production (NPP) and net ecosystem production (NEP) by explicitly calculating such carbon fluxes as gross primary production (GPP), plant respiration, and soil decomposition on a monthly time-step; these fluxes are regulated by a multitude of environmental factors at the physiological scale. In relation to global change, responses to increased atmospheric CO 2 and temperature should be modeled in a mechanistic manner. (3) It is a prognostic model. Sim-CYCLE is designed to be applicable not only to the simulation of an equilibrium state under given conditions, but also to the prediction of a transitional state under changing environmental conditions. Importantly, Sim-CYCLE is based on the dry-matter production theory, which enabled us to achieve the scaling-up from single-leaf to canopy and to conceptualize the growth process. Since the model includes both radiation and hydrological conditions, some indirect influences of the initial environmental change can be properly evaluated. We present a comprehensive model description and preliminary results confirmed at the plot scale: (1) intensively in four natural ecosystems and (2) extensively in global 21 sites. At each site, model parameters were calibrated to capture the observed carbon dynamics (e.g. productivity and carbon storage) at the equilibrium state. Successional growth patterns and seasonal variations in CO 2 exchange were also examined in a qualitative manner. Sim-CYCLE successfully expressed the differences between tropical forest and boreal forest and between humid forest and arid grassland in terms of productivity and carbon storage. Next, we simulated transitional ecosystem carbon dynamics, in response to step-wise atmospheric CO 2 doubling and disturbance regime. The simulated temporal patterns of carbon cycle were realistic and ensured that Sim-CYCLE is an effective tool for predicting the impact of global change.