Analysis and Simulation of the Start of Growing Season on the Qinghai-Xizang Plateau Based on Remote Sensing Vegetation Index

The Qinghai-Xizang Plateau （QXP） is an important herbage producing area， ecological barrier and water conservation area， the vegetation ecological process on which can directly affect the changes of China and even East Asia.With global warming， the phenological period of vegetation on the QXP is constantly changing， affecting climate and ecosystem through carbon cycle and hydrothermal cycle， etc.The study of phenological change and its influencing factors has become a key issue， and the construction of models that can realize future phenological prediction is of great scientific significance.In this paper， based on the Normalized Difference Vegetation Index acquired by satellites during 2000 -2020 （MODIS NDVI）， the dynamic threshold method was used to extract the start of growing season （SOS） on the QXP， and its spatiotemporal variation was analyzed in combination with vegetation types， so as to construct multiple phenological models of SOS， air temperature and soil moisture， exploring the hydrothermal conditions required for different regions and types of vegetation to start growing.The results showed that： （1） From 2000 to 2020， the overall SOS advance trend of the QXP was most significant in the eastern part of the region， where the SOS advance rate exceeded 10 d·（10a）-1.Coniferous forests， scrub， meadows， and alpine vegetation cover areas had a high percentage of SOS advance， and grasslands had about 50 % of slightly delayed areas.（2） The eastern and northern regions of the QXP showed an obvious warming and humidification trend.The average annual temperature rise rate was about 0.36 ℃·（10a）-1， and the average annual soil moisture increase rate was about 3.8×10-4 m3·m-3 （p<0.01）.（3） The parameters of the four phenological models showed that the vegetation growth in the eastern and southern QXP required higher hydrothermal conditions.The main controlling factor for vegetation SOS in the south was air temperature， while in the north it was soil moisture.The temperature and soil moisture thresholds and main controlling factor of different vegetation types were also closely related to their spatial distribution locations.（4） The cumulative temperature and cumulative soil moisture threshold model established in this paper has the best simulation effect for the main vegetation types （grassland， meadow and alpine vegetation） on the QXP， and the root-mean-square error is only about 8 days， which has reference significance for the future SOS prediction and the interaction mechanism between phenology and climate on the QXP.