Unstable changes in ecological quality of the four major sandy lands in northern China based on Google Earth Engine

作者：Haowen Ma, Rui Wang, Rui Wang, Enliang Guo, Shan Yin, Yao Kang, Yanli Wang, Yanli Wang, Jiapei Zhao, Jisiguleng Wu, Quanfei Mu, Delong Zhou · 发表于：Ecological Indicators · 年份：2025 · DOI：10.1016/j.ecolind.2025.113195 · 被引用次数：6 · 研究领域：Land Use and Ecosystem Services、Remote Sensing and Land Use、Remote Sensing in Agriculture

• Monitor ecological quality of China’s 4 major sandy lands from 2000 to 2022. • Decompose trends and abrupt changes using time series decomposition algorithm. • Use Geographical Detector to explore driving mechanisms over different periods. • The ecological quality of the 4 sandy lands improves, while 2 show unstable trends. • Natural factors primarily drive 3 sandy lands, while anthropogenic factors drive 1. Based on the Google Earth Engine (GEE) platform, this study constructed a Remote Sensing-based Ecological Index (RSEI) model for four major sandy lands in northern China from 2000 to 2022. RESI incorporates a Bayesian Estimator of Abrupt change, Seasonal change, and Trend (BEAST) to mine abrupt change years and key information within the time series data. Moreover, it combines Geographical Detector (GD) to conduct a detailed quantitative study on the spatiotemporal variation and driving mechanisms of the Ecological Quality (EQ) in these sandy lands at different times and discusses possible causes for abrupt changes in the EQ. The results indicated that from 2000 to 2022, the EQ of all four major sandy lands improved to varying degrees. However, the introduction of BEAST revealed that the RSEI of all four sandy lands experienced abrupt changes in the latter half of the study period, with the EQ in the Horqin and Hulun Buir showing potential for continued improvement post-change. By contrast, the EQ of the Mu Us and Otindag showed a downward trend within certain intervals...