Christine Klaas

Biosphere Dynamics: Challenges for Earth System Models

Many lines of evidence have established the role of the biosphere in the Earths climate. Relevant processes act on time scales from minutes (e.g. stomatal control of evapotranspiration) to hundreds of millions of years (coevolution of the atmosphere and life). Successful attempts have been made to model the dynamics of the terrestrial biosphere on time scales up to that of vegetation dynamics, and comparable efforts in global marine biosphere modelling are under way. A long-term goal of biosphere modelling is the interactive representation of biospheric processes in Earth system models built on coupled models of the ocean and atmosphere. The most advanced climate models already contain sufficient biology to represent in some way the coupling between climate change and the global carbon cycle. This work has shown that the representation of terrestrial carbon cycling is a major source of uncertainty for the future of atmospheric CO 2 . However, no climate model as yet includes a representation of land-surface processes and carbon cycling that is consistent with current understanding of plant physiological and ecological processes governing the exchanges of energy, water and carbon between terrestrial ecosystems and the atmosphere. Similarly, ocean biogeochemistry representations in current climate models fall short of representing current understanding of the dynamics of marine ecosystems and their interactions with mixed-layer physics and nutrient cycles. This article illustrates the problem with examples, from both the terrestrial and marine realms, of important processes whose representation in coupled models may be seriously in need of improvement. These examples speak to a need for more effective information exchange and collaboration among the different disciplines involved in Earth system science.