Spatiotemporal variation of snow depth in the Northern Hemisphere from 1992 to 2016

Abstract

Abstract. Snow cover is an effective best indicator of climate change due to its effect on regional and global surface energy, water balance, hydrology, climate, and ecosystem function. We developed a long term Northern Hemisphere daily snow depth and snow water equivalent product (NHSnow) by the application of the support vector regression (SVR) snow depth retrieval algorithm to historical passive microwave sensors from 1992 to 2016. The accuracies of the snow depth product were evaluated against observed snow depth at meteorological stations along with the other two snow cover products (GlobSnow and ERA-Interim/Land) across the Northern Hemisphere. The evaluation results showed that NHSnow performs generally well with relatively high accuracy. Further analysis were performed across the Northern Hemisphere during 1992–2016, which used snow depth, total snow water equivalent (snow mass) and, snow cover days as indexes. Analysis showed the total snow water equivalent has a significant declining trends (~ 5794 km3 yr−1, 12.5 % reduction). Although spatial variation pattern of snow depth and snow cover days exhibited slight regional differences, it generally reveals a decreasing trend over most of the Northern Hemisphere. Our work provides evidence that rapid changes in snow depth and total snow water equivalent are occurring beginning at the turn of the 21st century with dramatic, surface-based warming.

Why Do CMIP6 Models Fail to Simulate Snow Depth in Terms of Temporal Change and High Mountain Snow of China Skillfully?

“…Only errors in precipitation or temperature for the current month were considered in previous studies, however, it has been found that winter snow can also be affected by the precipitation and temperature of the months prior to the winter (Qiao et al., 2021). In addition, the remote sensing (Venäläinen et al., 2021; Xiao et al., 2020) and reanalysis (Bian et al., 2019; Mortimer et al., 2020; Orsolini et al., 2019) snow data may have limitations in certain conditions. For example, mountainous regions are masked out in the widely used Globsnow 3.0 SWE data set, owing to methodological limitations under complex terrains (Luojus et al., 2021).…”

Real-Time Snow Depth Estimation and Historical Data Reconstruction Over China Based on a Random Forest Machine Learning Approach

“…CC BY 4.0 License. application in remote sensing fields is promising (Liang et al, 2015;Bair et al, 2018;Xiao et al, 2018;Xiao et al, 2019). ML techniques can reproduce the nonlinear effects and interactions between variables without assumptions of a functional form.…”