Uncertainties linked to land-surface processes in climate change simulations

The impact of climate change on the hydrology of continental surfaces is cr itical for human activities but the response of the surface to this perturb ation may also affect the sensitivity of the climate. This complex feedback is simulated in general circulation models (GCMs) used for climate change predictions by their land-surface schemes. The present study attempts to qu antify the uncertainty associated with these schemes and what impact it has on our confidence in the simulated climate anomalies. Four GCMs, each coup led to two different land-surface schemes, are used to explore the spectrum of uncertainties. It is shown that, in this sample, surface processes have a significant contribution to our ability to predict surface temperature c hanges and perturbations of the hydrological cycle in an environment with d oubled greenhouse gas concentration. The results reveal that the uncertaint y introduced by land-surface processes in the simulated climate is differen t from its impact on the sensitivity of GCMs to climate change, indeed an a lteration of the surface parametrization with little impact on model climat e can affect sensitivity significantly. This result leads us to believe tha t the validation of land-surface schemes should not be limited to the curre nt climate but should also cover their sensitivity to variations in climati c forcing.