Significant decreasing trends in snow cover and duration in Northeast China during the past 40 years from 1980 to 2020

Snow cover in Northeast China (NC) is not only a vital variable in water availability but also a significant indicator of climate change. Limited by the short time span, incomplete spatial coverage, large uncertainty, and low spatial resolution of the current snow cover products, the long-term snow cover changes in NC remain unclear. To resolve this issue, a high-quality, long-term, and daily 500 m fractional snow cover (NCFSC) data were generated by integrating ERA5-land FSC data and several auxiliary data, with the help of a random forest (RF) regression model, during the period 1980-2020. The validation against in situ snow-depth observations from 2000 to 2018 indicated that the NCFSC effectively distinguished snow cover conditions (the overall accuracy (OA) was 0.88, producer's accuracy (PA) was 0.78, user's accuracy (UA) was 0.81, and Cohen's kappa (CK) value was 0.71). The spatiotemporal comparison between the NCFSC and MODIS FSC data (MDFSC) during 2015-2018 showed that NCFSC was reliable and highly comparable, with bias, RMSE, and R values of -0.02, 0.14, and 0.73, respectively. Additionally, based on NCFSC, the spatiotemporal variations in snow cover in NC over the past 40 years were analyzed for the first time, and the results showed that the FSC had a significant decreasing trend (0.0017/year, p < 0.05). The FSC in farmland had the fastest decline rate, 0.0020/year (p < 0.05), and the snow cover areas decreased by approximately 20 % from 1.55x10(5) km(2) in 1980 to 1.19x10(5) km(2) in 2020. The snow onset date had a delayed trend (0.141 day/year, p > 0.05), the snow end date had a significant advanced trend (0.243 day/year, p < 0.05), and the snow duration days had a significant decreasing trend (0.384 day/year, p < 0.05) in NC. In summary, the NCFSC seamlessly depicts the spatiotemporal changes in snow cover in NC over the past 40 years, providing a scientific basis and decision-making support for hydrological studies and climate change.