[Spatio-temporal Pattern and Influencing Mechanism of Carbon Sinks in the Yellow River Basin （Qinghai Section）]

The change trends， trend analyses， and Hurst index of net ecosystem productivity （NEP） in the Yellow River Basin （Qinghai section） from 2001 to 2020 were explored via the Carnegie-Ames-Stanford Approach （CASA） and empirical estimation models. The Optimal Parameter Geodetector （OPGD） model was employed to analyze the influences of factors such as human activities， climate， topography， and vegetation on NEP and to discuss the time-lag effect of climate factors on NEP. The results indicated that： ① From 2001 to 2020， the NEP of the Yellow River Basin （Qinghai Section） kept rising. The average NEP （in terms of C） in 2001 was the lowest， ranging from 258.755 to 278.086 g·（m²·a）^-1， with an average annual increase of 1.34 g·（m²·a）^-1. ② The area prone to degradation in the future accounted for 60.83% of the land area， mainly concentrated in the western and central regions （i.e.， from Qumalai County to Tongren County）. In contrast， the areas showing improvement signs were mainly concentrated in Maduo County and Gonghe County. ③ FVC and GPP were dominant in the explanation of NEP. Among the climatic factors， temperature had a stronger explanatory capacity. Furthermore， the interaction effect was more remarkable in explaining NEP than a single factor. ④ Under a lag period ranging from 0 to 3 months， the responses of NEP to variations in temperature and rainfall exhibited pronounced time lags. Specifically， the average lag months of NEP to rainfall and temperature were 2 months and 1.62 months， respectively， demonstrating that the response time of NEP to temperature fluctuations was shorter and more sensitive in contrast to that of rainfall.