An operational methodology for validating satellite-based snow albedo measurements using a UAV

Abstract

The albedo, or reflectivity, of seasonal snowpack directly controls the timing and magnitude of snowmelt and runoff. Snow albedo is affected by a large number of snow physical and environmental properties that vary considerably at multiple spatiotemporal scales. This variability introduces a high degree of uncertainty into existing modeling techniques. Models for snowmelt that require snow albedo can be improved by incorporating satellite measurements to inform and update estimates of this snow property. However, satellite measurements are susceptible to a multitude of error sources, which requires them to be calibrated and validated by means of ground-based measurements. Ground-based measurements from automated weather stations are often located at sparsely-distributed monitoring sites in homogeneous meadow environments. These spatially restricted in-situ data provide biased validation and calibration data that are not representative of the heterogeneous landscapes that comprise many seasonally snow-covered watersheds. In order to provide comprehensive validation and calibration of satellite albedo products, multiple near-surface measurements should be taken across large areas to capture the high degree of spatial variability that snow albedo can exhibit. UAV albedo measurements can be used to bridge the scaling gap between satellite and point-based measurements. Since these platforms are in a novel stage, the requisite methodologies for topographic correction and comparison to gridded albedo products do not exist. Additionally, there lacks a general understanding of the spatial scaling of albedo measurements in heterogeneous terrain. This research aims to develop these methodologies and provide a comprehensive understanding of how to deploy these platforms and properly interpret their measurements. We first developed and validated a topographic correction using ground-based measurements of snow albedo in a sloping alpine meadow. Sensitivity analyses on both ground validation measurements and UAV-based albedo surveys in our alpine study area highlight the implications of using different user-defined parameters for the proposed topographic correction and satellite comparison methods. Improvements to the methodology can be made in the way it accounts for trees, shading, and cloud cover. This research develops the initial steps requisite to the operationalization of UAV albedo measurements and standardization of the techniques.