The importance of land-surface variability to climate variability

Land surface schemes attempt to simulate the evolution of a set of prognostic and diagnostic variables which are the result of physical and biological processes in the soil and vegetation. These processes are all fundamentally local, as they arise within elements of the land surface (the leaf, the plant, the hillside, the soil column). However, they occur over a variety of time scales. The two most fundamental time scales are the diurnal cycle, and the annual cycle. Upon these mean cycles are perturbations: synoptic variations on times scales of one day plus/minus approximately one order of magnitude, and seasonal-interannual variations on times scales of one year plus/minus approximately one order of magnitude. By decomposing temporal variability in the land surface into these time scales, we can assess their relative importance. We are examining the contributions of variations at the land surface on these time scales using a global climate model consisting of a state-of-the-art coupled atmospheric GCM and land surface scheme. We are also comparing to the effect of ocean temperature variations, which are limited to only the annual cycle and seasonal-interannual variations. Thus, the relative contributions of ocean and land variability to climate variability can also be assessed.