GISS Lunch Seminar
Speaker: Dervla Meegan-Kumar (UC Irvine)
Title: Influence of resolved topography on model biases in E2.1 and E3

Abstract:
Mountains shape climatology through their interactions with low-level flow. The averaging of Earth's surface height onto coarse-resolution model grids may limit models' ability to fully capture the impact of orographic blocking on circulation, however, as this often leads to unrealistically low mountain heights. I explore the sensitivity of model biases in E2.1 and E3 to resolved topography in suite of idealized experiments forced with modified surface height boundary conditions. In the first set of simulations, I implement envelope topography (calculated as the mean plus 1σ variance of sub-grid-scale topography) over Central and South America in the fully-coupled E2.1 model to evaluate the role of orographic blocking by these ranges on common model biases, including the double-ITCZ and the dipole in precipitation biases over South America (i.e. too-wet Andes and too-dry Amazon). Next, I test the impact of applying grid-cell maximum topography over North and Central America on the North American Monsoon in the atmosphere-only E3. While the topographic modifications improve some aspects of the model circulation in both experiments, the effect of the improved circulation is likely confounded by other model limitations, such as the simulation of stratocumulus clouds in the south Pacific in E2.1 or the representation of the narrow Gulf of California. The results highlight the importance of accurately representing orography in GCMs and that alternate methods of calculating surface height boundary conditions should be considered in model development, however, the optimal interpolation scheme will vary depending on model resolution and the impact of topography on model biases may be limited if the model is overly sensitive to other parameterizations.