Search

Vegetation phenology is closely related to seasonal dynamics of the lower atmosphere and,therefore,is thus considered as an important indicator that influences the feedback mechanisms of global climate variation to the terrestrial ecosystem.Detecting vegetation phenological metrics using time series remote sensing dataset at the regional or global scale has become an active research topic in geography and ecology.This paper presents a dynamic thresholding method to extract the vegetation phenological metrics(greenup date,growing season length,maximum EVI and wilting date) of typical grassland,including alpine shrub meadow in the eastern region,alpine meadow in the central region,alpine steppe and alpine desert steppe in the western region of northern Tibetan Plateau,using a time-series TERRA/MODIS EVI dataset from 2001 to 2010 which was reconstructed using the asymmetric Guassian function fitting method.Furthermore,the relationship between inter-annual variation of phenological metrics and climatic factors was analyzed based on the measured data of weather stations close to the sampling areas,the results showed that:(1) The ordinal relationship of these typical vegetation samples in phenological metrics was in the sequence of alpine shrub meadow,alpine meadow,alpine steppe,and alpine desert steppe(for greenup date and the wilting date,the sequence respectively referred to be earlier and later).(2) The annual,spring,summer and winter average temperature of northern Tibetan Plateau all showed significant ascending trend,the increased rate was between 0.8℃/10a and 3.9℃/10a.While the decreased trend of precipitation was not apparent.Due to the warming and drier climate,the greenup date of four types of typical vegetation was advanced(approximately 7.2—15.5d/10a earlier) and the growing season length became longer(approximately 8.4—19.2d/10a),whereas the wilting date of four types of vegetation cover did not show apparent advanced or lingering trend;(3) The maximum EVI of different typical vegetation cover types was affected by different climatic factors.As follows,alpine shrub meadow was mainly affected by spring average temperature and spring precipitation,and precipitation had more important influence on phenological variation than temperature;Alpine meadow was affected by both spring and summer average temperature and precipitation;Alpine steppe and alpine desert steppe were primarily influenced by summer average temperature and precipitation,while for alpine steppe vegetation the inhibition of summer average temperature was more intense than the promotion effect of summer precipitation,among four types of vegetation cover,alpine desert steppe had the closest relationship with summer precipitation(the correlation coefficient,-0.752);(4) As for greenup date,alpine shrub meadow was mainly affected by autumn precipitation of the previous year,the correlation coefficient was closed to-0.579;however,other three types of vegetation cover were mainly influenced by spring average temperature,and the character was the most obvious for desert steppe(the correlation coefficient,-0.559).That was due to that high temperature in spring expedited snow melting and tundra soil thawing which are indispensable for vegetation greenup,especially for high-altitude regions in northern and western Tibet.In addition,the autumn precipitation and winter average temperature in the previous year also had some impact on the vegetation greenup of alpine steppe and alpine desert steppe.