In response to the growing threat of intensified cropland abandonment to food security in the mountainous areas of northern Hebei, this study selected Zhangjiakou City as the study area. Based on the Google Earth Engine (GEE) platform and Sentinel-2 remote sensing imagery, the continuous change detection and classification (CCDC) algorithm was employed to identify the spatiotemporal patterns, occurrence frequency, and trajectory characteristics of abandoned cropland from 2019 to 2023, and to explore the driving factors of cropland abandonment. The results were as follows. 1) During the study period, the total area of abandoned cropland in Zhangjiakou increased from 329.12 km² to 449.41 km², showing a fluctuating upward trend, with an average annual abandonment rate of 4.18%. Spatially, cropland abandonment in the Bashang Plateau region was relatively concentrated, followed by the central and southern river valley basins, while the eastern mountainous and hilly areas had the least, showing clear regional differentiation and a predominantly high–high spatial clustering pattern. 2) In terms of frequency, cropland abandoned twice was dominant, followed by cropland abandoned three times, while cropland abandoned once or more than three times accounted for a relatively small proportion. The latitudinal distribution of abandonment frequency showed a “two-peak and one-trough” pattern, whereas the longitudinal distribution displayed “more in the west and fewer in the east, with multiple peaks and valleys”. 3) In terms of trajectory, newly abandoned cropland was dominant, mostly distributed in the eastern mountainous and hilly region. Reclaimed abandoned cropland ranked second, concentrated in the Yuxian–Yangyuan basin. Fluctuating and regular abandonment ranked third and fourth, respectively, both mainly occurring in the Bashang Plateau. 4) Elevation, slope, and GDP exhibited promoting effects on cropland abandonment in most areas, while population size and distance to roads were predominantly inhibitory. The influencing intensity of each driving factor decreased sequentially in the order of GDP, population size, elevation, slope, and distance to roads.

cropland abandonment, spatiotemporal patterns, abandonment occurrence frequency, evolutionary trajectory, driving factors, Zhangjiakou City

针对冀北山区弃耕加剧威胁粮食安全的问题, 以河北省张家口市为研究区, 基于Google Earth Engine平台, 利用Sentinel-2遥感影像数据, 通过CCDC算法, 识别2019—2023年弃耕的时空格局、发生频次及演变轨迹, 并挖掘弃耕的驱动因素, 得到如下结果。1) 研究时段内张家口市弃耕面积总体上呈波动上升趋势, 从329.12 km²增至449.41 km², 年均弃耕率为4.18%。弃耕主要出现在坝上高原区, 中南部河谷盆地区次之, 东部山地丘陵区最少, 并表现为以“高–高”集聚为主的空间正相关格局。2) 张家口市弃耕频次以二次弃耕为主, 三次弃耕次之, 一次及三次以上的弃耕较少。弃耕频次在纬度方向呈现“两峰一谷”的特征, 在经度方向呈现“西多东少、多峰多谷”的特征。3) 张家口市弃耕轨迹以新增弃耕类为主, 东部山地丘陵区分布居多。弃耕复垦类面积次之, 集中于蔚县–阳原盆地区域。波动性弃耕类与规律性弃耕类面积分别排在第三、四位, 主要分布于坝上高原区。4) 高程、坡度与GDP在多数区域对弃耕表现为促进作用, 人口数量和与道路距离则以抑制作用为主。各驱动因子的影响强度按GDP、人口数量、高程、坡度、与道路距离的顺序依次递减。