Quantifying Human-Mediated Carbon Cycle Feedbacks
We use the integrated Earth System Model (iESM) to quantify the degree to which human land management in response to changing environmental conditions can alter the global carbon cycle. We further develop a theoretical framework for comparing human-mediated and non-human-mediated carbon cycle feedbacks across models in a consistent fashion, which can help to understand and diagnose sources of uncertainty in this emerging area of research.
The carbon cycle feedback literature often refers to two primary responses of the Earth system to climate change: the concentration-carbon feedback and the climate-carbon feedback. These theoretical constructs provide a useful framework for identifying the broad sources of agreement or disagreement across models. Meanwhile, a growing literature and set of modeling capabilities is concerned with the role that human land management plays in these feedbacks. We expand the existing theoretical framework of the Coupled Carbon-Climate Cycle Model Intercomparison Project (C4MIP) which provides an organizational scheme to understand differences across models going forward.
Changes in land and ocean carbon storage in response to elevated atmospheric carbon dioxide concentrations and associated climate change, known as the concentration-carbon and climate-carbon feedbacks, are principal controls on the response of the climate system to anthropogenic greenhouse gas emissions. Such feedbacks have typically been quantified in the context of natural ecosystems, but land management activities are also responsive to future atmospheric carbon and climate changes. Here we show that inclusion of such human-driven responses within an Earth system model shifts both the terrestrial concentration-carbon and climate-carbon feedbacks toward increased carbon storage. We introduce a conceptual framework for decomposing these changes into separate concentration-landcover, climate-landcover, and landcover-carbon effects, providing a parsimonious means to diagnose sources of variation across numerical models capable of estimating such feedbacks.