Abstract: Vegetation is a key component of the complex coupled human-nature system, and there is a prominent greening trend in the Yellow River basin (YRB).We used the NOAA Climate Data Record (CDR) of AVHRR Normalized Difference Vegetation Index (NDVI) daily dataset from 1982 to 2020 and applied a land surface phenology method to investigate the differentiation of the human-nature relationship in the YRB from a vegetation perspective.To extract vegetation intra-annual indices, we applied a land surface phenology method that integrates the Savitzky-Golay filter and double logistic fitting, in addition to the traditional annual vegetation index based on annual NDVI series.We analyzed the trend of the inter-annual and intra-annual vegetation indices and summarized the typical types of vegetation changes using k-means clustering.We also conducted an analysis of the driving factors behind the different vegetation changes.Our results demonstrated that the integrated intra-annual indices extraction method is suitable for various vegetation types in the YRB, achieving a high recognition rate of 98%.The spatial and temporal variations in the extracted intra-annual indices revealed differences in ecological engineering and agricultural activities under distinct climatic conditions.We identified five types of ecological engineering-dominated vegetation changes and two types of agricultural-dominated vegetation changes, and found that the significance of agricultural vegetation changes was comparable to that of ecological engineering.Agricultural intensification is a critical factor that cannot be ignored in vegetation research of the YRB.Given the tense relationship between humans and nature, and the severe water-food-ecological contradictions in the YRB, it is crucial to optimize agricultural water use and ensure food security.