Evaluation of grassland biomass and driving factors in the Hailar river basin based on random forest model

作者：Chenglong Yao, A Yinglan, Guoqiang Wang, Baolin Xue, Jin Wu, Xianglong Dai · 发表于：Journal of Cleaner Production · 年份：2025 · DOI：10.1016/j.jclepro.2025.146590 · 被引用次数：5 · 研究领域：Remote Sensing in Agriculture、Forest, Soil, and Plant Ecology in China、Remote Sensing and Land Use

Under the dual impacts of climate change and human activities, the response of aboveground biomass (AGB) in grasslands to environmental changes has become increasingly complex, especially in the estimation of biomass and identification of influencing factors in large-scale grasslands, where significant challenges still exist, particularly in the impact of land surface temperature (LST) on AGB, which is often overlooked. Therefore, in this study, an improved random forest algorithm is combined with the normalized difference soil index to estimate the grassland AGB in the Hailar River Basin from 2013 to 2023. We employ Theil–Sen slope estimation and the Mann‒Kendall trend test to analyze spatiotemporal changes and investigate the impacts of natural factors, including altitude, precipitation, and temperature, on AGB. The main research conclusions are as follows: (1) The grassland ecosystem in the Hailar River Basin has shown a gradual improvement trend over the past decade. (2) Climate factors dominate biomass changes: precipitation is positively correlated with AGB, showing a lag effect, whereas high temperatures significantly suppress grassland biomass. The impact of LST on AGB is more significant than that of air temperature. The contribution of grassland growth to LST in midlatitude semiarid regions is more significant than that of precipitation. (3) The dual role of human activities: the policy of returning grazing to grasslands may help restore grasslands, whereas urban ex...