Topographic Correction of Broadband Albedo of Snow Measured from a UAV

Abstract:

The albedo of snow is a critical component of Earth’s climate and hydrologic cycles, as it controls energy dynamics that govern the accumulation and melt of snowpack. Satellite-derived albedo products require calibration and validation from ground-based measurements. Recent work has shown that UAVs are capable of supplementing point-based ground measurements as validation sources. When measuring snow albedo from a UAV in complex terrain, the irradiance detected by the platform is a function of many variables. Since the measured surface-reflected irradiance of snow is affected by surface slope and aspect, corrections for topography are critical in areas with high relief. Previous studies have utilized lidar scanners and manual snow surveys to apply topographic corrections to point and satellite albedo measurements. This research utilizes dual-pyranometers (measuring broadband albedo with a 310 - 2700 nm spectral range) and an RGB camera mounted on a UAV to measure snow albedo at decimeter resolutions and apply topographic corrections using photogrammetric terrain models. Despite a COVID-19 limited field season, two flights were conducted in a high alpine meadow during the 2020 Winter. Flights collected overlapping imagery, and Structure-from-Motion photogrammetry was used to generate elevation models of the study area. Corrections were applied to the RGB photography using measurements from the pyranometers to derive high-resolution albedo surfaces, and topographic corrections were applied using the photogrammetric models. The post-processing performed on raw photography improved point matching and orthomosaic generation during photogrammetry. The topographic correction removed a considerable amount of directional lighting artifacts in the albedo model. This platform is advantageous in that it is self-contained, has the ability to produce multiresolution albedo models, and generate topography data used in the correction algorithm without the need for expensive lidar scanners or labor-intensive manual surveys. In addition to proposing a method for improved UAV snow albedo measurements, this research provides valuable insights in performing photogrammetry over bright, homogeneous snow-covered landscapes.