Seasonal snow cover provides a thermally stable and humid refuge for overwintering plants in temperate forests. Canopy composition changes snow processes and shapes the thermal regime. To explore canopy effects on snow depth (SD) and below-the-snow temperature, ultrasonic sensors and time-lapse infrared cameras were used to monitor SD over two winters. Three forest types with different canopy compositions [the ratio of Korean pine (Pinus koraiensis Sieb. et Zucc.) basal area, R_Kp] on adret (sunny) and ubac (shady) slopes, including broadleaved secondary forests, Korean pine-broadleaved mixed forests, and Korean pine plantations, were selected as sample sites. Results showed time-lapse cameras were more accurate for monitoring SD in studied forests. Increasing R_Kp increased the snow interception rate and reduced the peak SD, snow melting rate, and snow cover duration. 21–69% of the snowfall was intercepted by the forests, with the R_Kp explaining 89% of the variance in interception rates. All forest stands had the lowest interception rate for severe snow events. For each forest type, plots on ubac slopes had a lower interception rate and thicker snow cover due to the positive effects of prevailing northerly wind on snow accumulation. Inter-annual variation of snowfall affected SD, snow melting rate, and below-the-snow temperature, with smaller differences among forest stands during the below-normal snow winter (2018/2019). The subnivium established in all plots during the normal snow winter (2019/2020) with the longest maintenance in secondary forest on ubac slopes and the shortest maintenance in Korean pine forests. The temperature increment under snowpack was positively correlated with SD while negatively correlated with air temperature. Overall, the forest with fewer evergreen canopy compositions was covered with thicker snowpack for a longer duration and had a more stable and warmer temperature regime, and fewer freeze-thaw events beneath the snow, which were essential for safe overwintering of regenerated individuals.

季节性积雪为温带森林中的越冬植物提供了一个热稳定和潮湿的避难所。冠层组成改变了雪的过程并塑造了热状态。为了探索冠层对雪深 ( SD ) 和雪下温度的影响，超声波传感器和延时红外摄像机被用来监测两个冬天的SD。三种不同冠层组成的森林类型[红松( Pinus koraiensis Sieb. et Zucc.)基面积比，R _Kp] 在 adret（晴天）和 ubac（阴凉）斜坡上，包括阔叶次生林、红松阔叶混交林和红松人工林，被选为样本点。结果表明，延时摄像机在监测研究森林中的SD方面更准确。增加R _{Kp 会}增加积雪拦截率并降低峰值SD、融雪率和积雪持续时间。21-69% 的降雪被森林拦截，R _Kp解释了拦截率 89% 的差异。所有林分对严重降雪事件的拦截率最低。对于每种森林类型，由于盛行北风对积雪的积极影响，ubac 斜坡上的地块具有较低的截留率和较厚的积雪。降雪的年际变化影响SD、融雪速率和雪下温度，在低于正常雪冬季（2018/2019）期间林分之间的差异较小。正常雪冬（2019/2020）在所有地块中建立的亚层，乌巴克斜坡次生林的维护时间最长，红松林的维护时间最短。积雪下温度增量与SD呈正相关而与气温呈负相关。总体而言，常绿冠层成分较少的森林被更厚的积雪覆盖的时间更长，温度状况更稳定和温暖，雪下的冻融事件更少，这对于再生个体的安全越冬至关重要。