China's arid and semi-arid regions are primarily located in northern China characterized by sparse vegetation, low precipitation and high evaporation. The ecological water consumption threshold of vegetation plays a critical role in optimizing afforestation and water resource allocation. However, the limitations of sparse vegetation distribution and water flux pose substantial challenges to quantifying the minimum transpiration water consumption threshold for sparse vegetation. In this study, multi-source remote sensing data were used to monitor spatiotemporal dynamics of sparse vegetation. By integrating in-situ observational data, remote sensing and literature data, the Differential Evolution Markov Chain (DE-MC) method was employed to optimize sensitive parameters (b1, b2, k1, and k2), enhancing the transpiration simulation accuracy of Priestley-Taylor Jet Propulsion Laboratory (PT-JPL) model. On this basis, the sparse vegetation and transpiration changes were analyzed to obtain the minimum transpiration water consumption threshold for sparse vegetation by the latitude of ecosystem resilience (LER) method. The results show that sparse vegetation has increased from 2001 to 2015 with localized degradation observed in eastern Inner Mongolia and northern Xinjiang. Following the model optimization, the transpiration simulation accuracy has been significantly improved. Consistent with vegetation expansion, transpiration exhibits an overall increasing trend, with higher water consumption mainly concentrated in eastern Inner Mongolia and northern Xinjiang, alongside localized decreases. Furthermore, the minimum transpiration water consumption threshold is closely related to vegetation type. In the future, the combination of the individual vegetation water consumption and minimum transpiration water consumption threshold obtained in this study enables the derivation of optimal planting densities for afforestation projects. Additionally, based on the watershed unit, the minimum water demand within each unit can be delineated. The results of this study provide a scientific basis for the optimal allocation of afforestation strategies and water resources in desertification regions.