Ecohydrological processes in a water-limited environment are sensitive to both climate conditions and human activities, but the response mechanisms have rarely been explored for large endorheic river basins via an integrated modeling approach. This study established an integrated surface watergroundwater model for the Heihe River Basin (HRB), China's second largest endorheic river basin, using GSFLOW as the modeling platform. Evapotranspiration (ET) and Leaf Area Index (LAI) data independently derived from remote sensing products were compared and correlated, respectively, with the modeling results. Scale-dependent interrelationships among ecological, hydrological, and human-impact (i.e., diversion and pumping) variables were revealed through multiple regression analyses. Major study findings include: (1) the independent ET and LAI data enabled the modeler to crosscheck the modeling results from a unique angle not possible with conventional groundwater and streamflow observations; (2) controlling factors for the temporal variability of ET and LAI exhibit notable scale-dependence, reflecting distinctive climate, and human impacts on different land covers; and (3) there exists an intricate linkage between the hydrological regimes in the lower HRB and the middle HRB, essentially equivalent to a tradeoff between the ecosystem health of the lower HRB and the sustainable development of the middle HRB. Overall, the integrated modeling assisted by the independent ET and LAI data has provided a coherent understanding on the regional water cycle, and led to new insights on tackling the existing water conflicts in HRB.