Improving snow albedo parameterization scheme based on remote sensing data

Snow albedo is an important parameter for determining the energy budget in high-latitude regions in winter, but it is often overestimated in the model and causes a cold bias of air temperature. To address this issue, we propose a new approach to improve the Canadian Land Surface Scheme (CLASS) snow albedo parameterization in NoahMP by updating the initial value of vegetation snow albedo based on remote sensing data. The modified CLASS scheme resulted in a 0.04 decrease in root mean squared error (RMSE) of the albedo, and the mean bias was reduced by 0.07, representing relative decreases of 37.3% and 89.1%, respectively. In addition, the simulation errors for the upward radiation and sensible heat flux were significantly reduced. Furthermore, the bias and RMSE of air temperature were reduced by 12.9% and 7.9%, respectively, compared to those in the original CLASS scheme. These qualitative and quantitative analyses demonstrated that our modified CLASS snow albedo parameterization scheme has a robust and positive impact on the simulation of snow cover areas. Our approach provides a preliminary investigation for improving snow albedo estimation, and we plan to conduct a widespread simulation and introduce new vegetation data into the model to improve the estimation of albedo in future research.