Geometric Methods in the Study of the Snow Surface Roughness

Abstract

The snow surface is very dynamic, and the roughness of the snowpack surface varies spatially and temporally. The snow surface roughness influences the movement of air across the snow surface as well as the resulting transfers of energy, and is used to estimate the sensible and la- tent heat fluxes to and/or from the snow surface to the atmosphere. In the present work we used dif- ferent metrics, including the random roughness, autocorrelation, and fractal dimension, geometric roughness length, curvature, and power spectrum density to characterize the roughness of a typical snow surface. The data for the surface come from airborne LIDAR measurements taken during from the NASA Cold Land Process Experiment in late March 2003 at the Fraser Alpine intensive study area. The surface elevation data were rotated to be parallel to the dominant wind direction and were interpolated to a 1-m resolution. We provide a comparison of methods and present their possible applicability for other datasets.