标题: Patterns and drivers of terrace abandonment in China: Monitoring based on multi-source remote sensing data

作者: Lu, D (Lu, Dan); Su, KC (Su, Kangchuan); Wang, ZP (Wang, Zhanpeng); Hou, MJ (Hou, Mengjie); Li, XX (Li, Xinxin); Lin, AW (Lin, Aiwen); Yang, QY (Yang, Qingyuan)

来源出版物: LAND USE POLICY 卷: 148 文献号: 107388 DOI: 10.1016/j.landusepol.2024.107388 Early Access Date: OCT 2024 Published Date: 2025 JAN

摘要: As urbanization and industrialization surge in China, the problem of land abandonment intensifies. However, the situation of terrace abandonment in China remains unclear. We conducted the first-ever remote sensing monitoring of terrace abandonment in China with full space coverage by combining high-precision terrace data with land use datasets to reveal the abandonment pattern. By fully considering natural and socio-economic factors, the XGBoost-SHAP framework was used to investigate the driving factors of terrace abandonment. The results show that approximately 2.42 % of terraces were abandoned from 2019 to 2021, mainly distributed in the Southwest and Loess Plateau regions. Agricultural regions with more terraces exhibited higher abandonment rates. The ratio of the population with pension insurance, cropland quality, slope, and land parcel size were prime drivers of terrace abandonment. There were significant spatial differences in the contribution of each factor. It is noteworthy that there was a significant deceleration in terrace abandonment trends in 2021, potentially ascribed to the impact of the COVID-19 pandemic leading to a substantial decrease in non-agricultural employment opportunities, thereby slowing down rural-to-urban emigration and even prompting a return migration of migrant workers. Grasping this critical post-pandemic period is crucial and should support returning migrant workers in engaging in agricultural activities by establishing diverse new agricultural entities and providing agricultural technical guidance.

作者关键词: Land use; Spatiotemporal dynamic; Machine learning; XGBoost; SHAP

地址: [Lu, Dan; Wang, Zhanpeng; Hou, Mengjie; Li, Xinxin; Lin, Aiwen] Wuhan Univ, Sch Resource & Environm Sci, Wuhan 430079, Peoples R China.

[Su, Kangchuan; Yang, Qingyuan] Southwest Univ, Sch Geog Sci, Chongqing 400715, Peoples R China.

通讯作者地址: Yang, QY (通讯作者)，Southwest Univ, Sch Geog Sci, Chongqing 400715, Peoples R China.

电子邮件地址: yizyang@swu.edu.cn

影响因子：6