Characterizing uncertainty in modeling primary terrestrial ecosystem processes

Abstract

The simulation results from models participating in the Coupled Climate Carbon Cycle Model Intercomparison Project (C4MIP) highlight the role of positive carbon‐climate feedback in accelerating growth of atmospheric CO2. The large range among models in the strength of this feedback indicates the uncertainty in our understanding of the response of the land and the oceans to continued climate warming and increasing CO2. Most of this uncertainty is associated with the response of terrestrial ecosystems to changes in climate and atmospheric CO2. The reasons for differences between models' responses to transient climate and CO2 forcing are not easily identified because of their complex parameterizations and spatially distributed processes. In this paper, we show that a simple box model can reasonably reproduce the globally averaged primary land‐atmosphere CO2 fluxes and carbon pools of two complex terrestrial ecosystem models (TEMs) over a range of emission scenarios. The parameters of the box model are calculated by fitting the box model to each TEM's response to transient climate and CO2 forcing. This approach is also applied to terrestrial carbon cycle components of carbon‐climate models participating in the C4MIP study. The resulting set of parameter values based on a common box model structure yields a wide range of parameter values, which suggests an absence of clear consensus in modeling primary terrestrial ecosystem processes and provides some insight into the reasons for divergent responses of terrestrial carbon cycle components.