Improving GPS-IR Snow Depth Estimation by Considering the Snow Surface Roughness

Abstract

Snow is an important environmental variable influencing weather and climate. The GPS-IR technique is a very effective technique for monitoring snow depth. A GPS-IR snow depth estimation corrected model is proposed to address the impact of the Signal to Noise Ratio (SNR) amplitude attenuation and snow surface roughness variation that are not considered in the standard model of GPS-IR. In this study, the snow depth of the P351 GPS site of the Plate Boundary Observatory (PBO) was obtained using the standard model and the corrected model, and the snow depth observations of the nearby SNOTEL station were compared, and the distribution of residuals was analyzed to assess the performance of the two models. Our results show that the correction model derives an RMSE of 11.8 cm and a bias of 6.0 cm at snow depths compared to the observations, a reduction of 1.6 cm in RMSE and 2.1 cm in bias compared to the standard model. The snow surface roughness coefficients obtained by the corrected model well characterize the snowfall process. The residual statistics show that the stability of the two models is approximate. Our study provides a reference for the research on the enhancement of snow observation networks and the improvement of snow products.