@article{XIAO2026, 
author = {Yao XIAO and Guo-Jie HU and Ren LI and Tong-Hua WU and Xiao-Dong WU and Guang-Yue LIU and De-Fu ZOU and Ji-Min YAO and Nan ZHOU and Lin ZHAO},
title = {Autumn snow expansion and spring divergence in Northeast China (2000–2020)},
year = {2026},
journal = {Advances in Climate Change Research},
volume = {17},
number = {1},
pages = {105-116},
keywords = {Northeast China, Snow cover duration, Snow depth, SHAP, Structural equation modeling, Machine learning, Elevation dependence},
url = {https://www.sciopen.com/article/10.1016/j.accre.2025.11.004},
doi = {10.1016/j.accre.2025.11.004},
abstract = {Mid-latitude snow has broadly declined, but Northeast China (NEC) shows a seasonal divergence that remains insufficiently explained. We integrated a cloud-free 500 m MODIS snow-phenology record with passive-microwave snow depth for 2000–2020 to map changes and diagnose drivers. We quantified controls using Random Forest with SHAP and structural equation modeling. Autumn snow cover duration expanded in northern NEC, with rates up to 12 d per decade around 47°―50°N, consistent with localized cooling (−0.1 to −0.2 ℃ per decade) and greater precipitation. In spring, responses diverged: the Greater Khingan Mountains experienced accelerated melt under 0.3–0.9 ℃ per decade warming, whereas parts of the Songliao Plain retained snow longer. Temperature dominated variability, while precipitation, vegetation, and elevation showed seasonally asymmetric and terrain-dependent effects. These findings highlight overlooked regimes in transitional mid-latitudes: autumn expansion driven by early-season cooling and moisture supply versus spring divergence shaped by warming and terrain. The driver diagnostics provide process-oriented evidence for improving parameterizations of shallow, short-lived snow and elevation-dependent melt in regional climate and hydrological models.}
}