Estimation of snow bulk density and snow water equivalent on the Tibetan Plateau using snow cover duration and snow depth

Abstract

Study region: Tibetan Plateau (TP), China. Study focus: Snow water equivalent (SWE) measurements is few compared with snow depth (SD) but combining with snow bulk density (SBD) can convert SD to SWE. No robust model has been established to estimate SBD and SWE for the TP. In this study, we aims to develop a model capable of obtaining SBD and SWE on the TP based on a significant dataset of simultaneously measured SD and SWE over nearly four decades. New hydrological insights for the region: We found unique seasonal characteristics and elevation differences in the SBD on the TP. The newly introduced factor snow cover duration (SCD) can reflect the beginning and status of each snowfall-snowmelt cycle in this type of discontinuous snow region. The developed artificial neural network (ANN) model based on two-factor of SCD and SD can capture these unique characteristics, with minimal errors and the best performance in simulating SBD and then estimating SWE on the TP. Our method improves the underestimation of the widely used regional average density during the heavy snow period, enables a reasonable conversion of the large amount of historical and growing SDs into critical SWEs, and thus improves our ability to assess snow water resources on the TP.

Keywords

• SWE
• Snow bulk density
• Artificial neural network model
• Tibetan Plateau