There is an exciting and important new paper that demonstrates the importance of fine enough spatial resolution to adequately resolve terrain and other landscape features. The paper is
Medvigy, D., R. L. Walko, and R. Avissar (2008), Modeling interannual variability of the Amazon hydroclimate, Geophys. Res. Lett., doi:10.1029/2008GL034941, in press. (Need subscription to access paper)
Abstract
“El Niño/Southern Oscillation (ENSO) strongly influences the interannual variability of Amazon hydroclimate. But current global climate models are unable to represent the associated variability, raising the question of their aptitude to simulate current and changing climate in that region. We explore this critical issue with the Ocean-Land-Atmosphere Model (OLAM), a new Earth System Model (ESM) capable of simulating regions like mountain ranges at high resolution while efficiently simulating the rest of the world at coarser resolution. We find that simulations representing the Andes at resolutions coarser than 100 km lead to a ‘reverse’ ENSO effect, with large precipitation rates in the Amazon during ENSO events. In contrast, the model correctly simulates the ENSO dry anomaly provided the Andes are simulated at resolutions finer than 100 km. Furthermore, we find that the grid mesh over the Andes has important implications for ENSO teleconnections, especially over the continents.”
What this paper tells us is that global climate models that do not adequately resolve such landscape features as the Andes cannot accurately simulate climate variability even on seasonal time scales in the Amazon. The obvious conclusion with respect to multi-decadal global climate predictions is that skillful regional projections of climate change due to human climate forcings are also not possible until global models accurately include such terrain features as the Andes.