Spatial and temporal variations in the end date of the vegetation growing season throughout the Qinghai-Tibetan Plateau from 1982 to 2011

The spatial and temporal variations in the end date of the vegetation growing season (EGS) and their relationships with climate factors across the Qinghai-Tibetan Plateau yet have not been well researched. In this study, we used the rate of the change in the curvature of the S-curve function which integrated a logistic function and an asymmetric Gaussian function and showed a better performance for fitting the LAI (leaf area index) data to extract the EGS from a long-term time series of AVHRR (advanced very high resolution radiometer) LAI data. The spatial distribution pattern of the EGS averaged from 1982 to 2011 presented a gradual decrease from the southeast to northwest plateau. The various vegetation types showed different phenological EGS timing. The EGS occurred earlier with increasing altitude (slope = −3 day km−1, p < 0.001). Throughout the entire Qinghai-Tibetan Plateau, the monthly air temperature and precipitation were positively correlated with the EGS, whereas the monthly sunshine duration showed a negative correlation. At the regional scale, a less pronounced increasing EGS trend (shifting about 1 day over 24 years, p = 0.084) was observed during the entire study period. By analyzing the trend turning points, we found that the EGS occurred later during 1982-1994 (slope = 0.155 day yr−1, p = 0.045) and 1999-2011 (slope = 0.096 day yr−1, p = 0.3), but occurred earlier during 1994-1999 (slope = −0.373 day yr−1, p = 0.049). During 1982-2011, the annual changes of EGS negatively correlated with precipitation (p < 0.1) in June, but positively with precipitation (p < 0.1) in August. As the same time, the annual changes of EGS positively correlated with sunshine duration (p < 0.1) in June, yet negatively with sunshine duration (p < 0.1) in August. During 1982-1994, the annual changes of EGS positively correlated with air temperature (p < 0.01) and negatively with precipitation (p < 0.1) in June. During 1994-1999, the annual changes of EGS only negatively correlated with air temperature (p < 0.05) in August. During 1999-2011, the annual changes of EGS only negatively correlated with sunshine duration (p < 0.1) in August.