Tracking temporal dynamics trajectories and drivers of net primary productivity in the Yellow River Basin of China

The response of vegetation to global climate change and human activity is diverse, which includes both linear and nonlinear changes. However, the understanding of the characterization and driving mechanism of this differential response remains unclear. Here, we distinguished linear, curvilinear, and abrupt trajectory characteristics of net primary productivity (NPP) from 2001 to 2020 in Yellow River Basin (YRB) and explored the driving factors of three trajectories. We found that 80.4% of the YRB was dominated by positive trend, of which linear, curvilinear and abrupt change accounted for 79.16%, 9.32% and 11.52%, respectively. The three trajectories showed obvious spatial heterogeneity. Positive abrupt trajectory was mainly distributed west of the Heihe-Tengchong Line, and the abrupt increment (39.8 g C m −2 , mainly in 2011) gradually decreased to the west. The positive linear increase rate was highest at the Heihe-Tengchong Line and gradually decreased to the west (average increase rate was 4.34 g C m −2 yr −1 ). Implementing ecological engineering and shifting to ecological land increased the positive abrupt increment and linear increase rate. Hydrothermal condition was the main factor affecting the abrupt increment and linear increase rate, and the increase of both contributed to the increase of abrupt increment and linear increase rate. The positive curvilinear trajectory indicated that although the NPP showed an increasing trend over the 20-year period, the increase rate is slowing down. NPP of more and more regions were decreasing after 2015 or will decrease in the future. NPP on the east side of the Heihe-Tengchong Line was dominated by aridity index (AI), whereas on the west side, it was mainly influenced by palmer drought index (PDSI), human footprint index (HFI), and soil moisture (SM). Significant fluctuations in the HFI caused a shift from a negative linear trajectory to a negative abrupt trajectory. Although NPP in the YRB showed an increasing trend over the 20-year period, the risk of decreasing in the future is increasing. The result showed that the temporal dynamics of NPP exhibited diversity, which was helpful in identifying hotspots of vulnerable vegetation productivity and targeted management measures. • Spatially distinguishing different temporal dynamic trajectory of NPP. • Positive linear trajectory was the dominant dynamic of NPP in the YRB. • Temporal dynamic trajectory exhibited spatial heterogeneity. • Despite showing positive trend, NPP declined in many regions after 2015.