Empirical Model Development for Ground Snow Sublimation beneath a Temperate Mixed Forest in Changbai Mountain

To develop an empirical model for ground snow sublimation beneath canopy, a weighing measurement experiment was conducted using snow samples with different density in the broadleaved Koreanpine mixed forest in Changbai Mountains, Northeastern China. Eddy covariance measurement for water vapor flux was used to evaluate the model performance. Eddy covariance data showed that the daytime sublimation was much larger than the nighttime sublimation, and 94.3% of daily sublimation occurred within the 8 h from 8:00 to 16:00. Daytime sublimation showed a linear relationship with snow density, and the regression coefficients between them varied with meteorological variables. The regression slope was closely correlated to solar radiation ($R^2=0.92$) and water vapor pressure ($R^2=0.79$), whereas the regression intercept was closely correlated to air temperature ($R^2=0.92$). Based on the regression relationships among sublimation, snow density, and meteorological variables, a nonlinear empirical sublimation model with a combination of snow density, solar radiation, water vapor pressure, and air temperature was developed. Sublimation estimation of the empirical model matched the eddy covariance data giving $R^2=0.83$ and root mean square error $(\mathrm{RMSE})=0.05\mathrm{mm}·{\mathrm{d}}^{-1}$. The surface reflectivity decreased with the increase in snow density. Dense snowpack would absorb more energy and promote snow sublimation by increasing the vertical water vapor pressure deficit between snow surface and atmosphere.