'Static' vegetation and dynamic global climate: Preliminary analysis of the issues time steps and time scales

The implementation of a fully interactive global vegetation model coupled into a Global Climate Model (GCM) first requires evaluation of the question of the 'correct' time step and time scale to employ. The problem of selecting the 'correct' frequency of communication between two such models may be a function of the models themselves: one vegetation model being more sensitive to the variability of climate than another. In turn, one GCM might exhibit a greater 'natural' variability than another GCM or be more responsive to altered vegetation characteristics. We have undertaken a set of sensitivity tests with different vegetation schemes and different climate models. Two questions have been addressed: is it necessary to employ daily instead of monthly climate data, and how many years of climate simulation should be included into the ensemble aggregate of climate. In this taper, two 'static' global vegetation models are assessed using simulated control climates from several GCMs, for ensemble averages of climate over 1 year, 5 years, 7 years, 10 years, 13 years, 15 years, 20 years and 45 years. Different sensitivities are found among the vegetation models and climate models investigated. The sensitivities of the individual biomes to the input climate vary greatly as a function of the GCM and/or the vegetation model employed. This suggests that there is a need for additional sensitivity analysis in order to evaluate how often the vegetation model has to return characteristics to the climate model. The prediction of vegetation using several datasets of simulated control climates seems to be a prerequisite for coupling of biosphere models into global climate models.