The implementation of artificial water allocation scheme (WAS) exerts a significant influence on alleviating water use conflicts between the middle and lower reaches and maintaining ecological health in arid and water-scarce regions. However, current studies mostly focus on analyzing the ecological impacts on the downstream areas after the WAS, without simultaneously investigating the ecohydrological responses in both the middle and lower reaches before and after the WAS. Thus, this study focused on the Heihe River Basin (HRB)-a prototypical arid-semiarid area in northwestern China-as a case study. Leveraging our new Normalized Difference Vegetation Index (NDVI) data (1982-2022) obtained by the fusion of Enhanced spatial and temporal adaptive reflectance fusion model (ESTARFM) algorithm, we analyzed spatiotemporal vegetation dynamics and assessed how natural factors and WAS influence the vegetation changes via Sen + Mann-Kendall trend analysis and Geographically and Temporally Weighted Regression (GTWR) model. As the results manifested: (1) Over the past 40 years, the NDVI in the HRB has shown a significant and slow increasing trend (0.0092/10 yr, R2 = 0.8853). (2) The vegetation in the middle reaches had the most significant growth rate (0.0220/10 yr, R2 = 0.8664). (3) SM2cm and SM100cm were the primary drivers of NDVI changes in the HRB. (4) The vegetation change in lower reaches was significantly correlated with current allocations, but showed higher correlation with allocations lagged by 1 year. The outcomes of this study provide a theoretical basis and feasible approaches for investigating the ecohydrological impacts of WAS on water-competing regions over long-time scales in arid and semi-arid areas.