**5. Conclusions**

Based on the time series MODIS NDVI datasets from 2001 to 2020, we retrieved the vegetation phenological parameters in the QLMs. The spatiotemporal variation in vegetation phenology was analyzed, and divergent correlations between the SOS and EOS and seasonal driving factors were explored. The results demonstrated that vegetation phenology in the QLMs is characteristic of advancing SOS, postponing EOS, and prolonging LOS, but the variation trends of vegetation phenology were not significant (*p* > 0.05) from 2001 to 2020. The seasonal temperature, precipitation, and soil moisture had spatiotemporal heterogeneous effects on the vegetation phenology. Compared with temperature and soil moisture, precipitation had a weaker influence on the vegetation phenology in QLMs. The spring temperature was the key driving factor influencing SOS in the QLMs. The autumn soil moisture and autumn temperature made the largest contributions to the variations in EOS at lower elevations (<3500 m a.s.l.) and higher elevations (>3500 m a.s.l.), respectively. Spring temperature was the key driving factor influencing SOS of most vegetation types. Autumn soil moisture was the main factor influencing EOS in deserts because of the strong soil moisture stress. An increase in summer soil moisture may limit vegetation growth in the QLMs. Under ongoing global change, finding the response of the SOS and EOS to driving factors is beneficial for a better understanding of the interactions between vegetation phenology and future climate change.

**Supplementary Materials:** The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/rs14153645/s1: Figure S1: Comparison of satellite-derived phenology data with ground-based phenology data.

**Author Contributions:** Formal analysis, X.C.; Methodology, X.C.; Software, G.X. and D.L.; Visualization, G.X. and X.H.; Writing—original draft, X.C.; Writing—review & editing, X.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by the Strategic Priority Research Program of Chinese Academy of Sciences (grant XDA20100102) and Gansu Provincial Science and Technology Major Special Plan (20ZD7FA005).

**Conflicts of Interest:** The authors declare no conflict of interest.
