Understanding historical changes in land surface albedo (ALB) is critical for projecting future climate changes over the Arctic. Previous studies have reported the important role of snow cover and vegetation growth in regulating Arctic land ALB, but their main influencing paths and separate contributions remain underexamined. In this study, the partial least squares path model (PLS-PM) and random forest model (RFM) are adopted to explore the dominant paths of three climate variables—temperature (Tem), precipitation (Pre) and wet days (WD)—in influencing the interannual variations in ALB (IVA) by regulating vegetation and snow cover. The results show that ALB in October and December exhibit significant (p < 0.05) upward and downward trends, respectively. ALB in the Northwest Territories, Victoria Island and Nunavut of the Arctic significantly (p < 0.05) increase in summer and October. The summer snow cover fraction (SCF) shows a significant positive correlation with ALB (R: 0.60‒0.78), while significant negative (R = -0.60) and positive (R = 0.50) relations are found between the enhanced vegetation index (EVI) and ALB during July and September, respectively. According to PLS−PM, the dominant paths that affect the IVA exhibit a prominent seasonal feature. Pre → EVI → ALB and Tem → SCF → ALB are identified as the main paths for ALB changes during spring, while in summer, the only significant path is Tem → SCF → ALB. In autumn, nonsignificant impacts on the variations in EVI and SCF (p > 0.05) are observed, while a significant path WD → SCF → ALB is found in winter. Based on RFM, the relative contribution (RC) of vegetation change to March IVA can reach 52.1 ± 9.2%, while snow variations play a dominant role in April and May, with contributions of 57.8 ± 8.4% and 68.4 ± 4.1%, respectively. Seasonally, summer IVA is mainly impacted by snow (RC from 50.5 ± 4.5% to 65.9 ± 4.5%), while vegetation exerts the dominant influence on IVA in autumn except for September. In winter, however, the dominant strength of vegetation is weakened except in December, with IVA mainly controlled by snow (RC above 50%).

了解地表反照率 (ALB) 的历史变化对于预测北极未来的气候变化至关重要。先前的研究报告了积雪和植被生长在调节北极土地 ALB 中的重要作用，但它们的主要影响路径和单独的贡献仍未得到充分研究。本研究采用偏最小二乘路径模型（PLS-PM）和随机森林模型（RFM）探索三个气候变量——温度（Tem）、降水（Pre）和湿日（WD）的主导路径——通过调节植被和积雪来影响 ALB (IVA) 的年际变化。结果表明，ALB 在 10 月和 12 月表现出显着（p < 0.05) 分别呈上升和下降趋势。西北地区、维多利亚岛和北极努纳武特地区的 ALB 在 夏季和 10 月显着增加 ( p < 0.05)。夏季积雪分数（SCF）与 ALB 呈显着正相关（R：0.60-0.78），而显着负相关（R  = -0.60）和正相关（R = 0.50) 分别在 7 月和 9 月期间发现增强植被指数 (EVI) 和 ALB 之间的关系。根据 PLS-PM，影响 IVA 的主要路径表现出突出的季节性特征。Pre → EVI → ALB 和 Tem → SCF → ALB 被确定为春季 ALB 变化的主要路径，而在夏季，唯一重要的路径是 Tem → SCF → ALB。秋季，对 EVI 和 SCF 变化的影响不显着（p > 0.05) 被观察到，而在冬季发现了一条重要的路径 WD → SCF → ALB。基于RFM，植被变化对3月IVA的相对贡献（RC）可达52.1±9.2%，而4月和5月则以降雪变化为主，贡献率分别为57.8±8.4%和68.4±4.1%。季节性地，夏季IVA主要受雪影响（RC从50.5±4.5％到65.9±4.5％），而除9月外，植被对秋季IVA的影响占主导地位。但冬季除12月外，植被优势强度减弱，IVA主要受雪控制（RC在50%以上）。