Exploring Topography-Based Methods for Downscaling Precipitation for Use in Earth System Modeling

Topography exerts major control on land surface processes. To improve representation of topographic impacts on land surface processes, a new topography-based subgrid structure has been developed for the Exascale Energy Earth System Model (E3SM) Land Model to represent the subgrid heterogeneity of surface elevation. To take advantage of the new subgrid structure for improving land surface modeling, this study explores four variations of topography-based methods for downscaling the grid mean precipitation to the subgrids within the larger grids. In the first three methods, the deviation of the subgrid precipitation from the grid mean precipitation is equal to the grid mean precipitation multiplied by the ratio of the elevation difference between the subgrid and grid mean to a specified elevation equals to the grid mean elevation, the difference between the maximum and minimum subgrid elevation, and the maximum subgrid elevation, respectively. The second method limits the ratio to 0.5 to avoid extreme values in tall mountains and the third method accounts for the slope effect. The fourth method is similar to the third method except that the Froude number is used to limit the numerator of the ratio in a blocking regime of the ambient flow. The downscaled precipitation is evaluated using the PRISM precipitation data over the U.S. based on statistical metrics. Results show that by accounting for topographic slope besides elevation, the third and fourth methods show clear advantages over the first and second methods. Furthermore, introducing the Froude Number in the fourth method improves downscaling skill and shows consistent advantages over the third method in areas with larger subgrid heterogeneity of surface elevation across grids of different sizes.