Abstract:

Based on the product of C5 MOD09A1 from Land Processes Distributed Active Center (NASA LPACC) from 2001 to 2018,the Enhanced Vegetation Index (EVI) over Qinling Mountains,the geographical boundary of south-north China,is reconstructed.The four reconstruction methods,Whittaker,DL,HANTS and AG algorithms were evaluated.Then the phenological parameters of Qinling Mountains was extracted with the maximum change rate and threshold method. We got the followed results.The Whittaker filter method is generally the best among the four methods in reconstructing the EVI of shrubs,fields and forests over Qinling Mountains.And the EVI from the Whittaker filter method shows the lowest difference from the original EVI in the reconstruction of shrub,farmland and forest cover with higher representative than that from the DL,AG and HANTS algorithms. Along the main ridge of the Qinling Mountains from the high elevations to the farming area,the start of growth season is gradually moved earlier,the end of growth season is gradually delayed and the length of growth seasons become much longer.The growing season in the northern and southern shallow-mountain areas and Funiu Mountains starts in early March,while it begins in early and mid-May (120-135 d of the year) in the coniferous forest belt.While the end of growth season in the coniferous forest belt ends earlier than that of shallow-mountain areas which mainly appear in early of November,and both of them is mainly in the range of 270-310 d of the year.The length of the growth (LOG) season over Qinling Mountains is mainly about 150-200 d,and the LOG at lower elevation (about 180 d) is longer than that of forest regions at higher elevation which is about 150-170 d.The distribution of the initial period of SOG over 95% areas of the Qinling Mountains generally shows an advance of 1-2 d,especially for the high-altitude area,which shows an obvious advance trend with about 2 d·a^-1,while there is a delay trend over the areas with the elevation lower than 500 m and the Funiu mountains.There is a delayed trend of the growth seasons (EOG) over Qinling Mountains,especially over the north of Qinling Mountains and low-altitude areas (about 2-3 d·a^-1).Furthermore,the length of growth seasons over Qinling Mountains become longer,and the change range of 72% of the forest is about 1-23 d·a^-1. There is a slightly increased trend of average temperature with a rate of 0.02 ℃·a^-1,while the total annual sunshine duration is decreased with a rate of 14.6 h·a^-1 over the Qinling Mountains.The change of hydrothermal conditions further affected the phenological parameters of Qinling Mountains.The correlation analysis between accumulated temperature,total precipitation and total sunshine duration and the phenological parameters at the temperatures of 0 ℃,5 ℃ and 10 ℃ respectively,which shows that the response sequence of the phenological parameters to the three climatic factors is: accumulated temperature> precipitation>sunshine.The increasing of temperature slightly leads to an earlier SOG,a delayed EOG and a longer LOG,especially in the areas with the elevation between 1 000 and 2 000 m over the southern and northern part of Qinling Mountains.While for higher mountains the correlations becomes more complex.The response of phenological parameters to accumulated temperature over Funiu Mountains in the east of Qinling Mountains is not obvious and there may be some other restriction factors.

Key words: Qinling Mountains, phenological parameters, EOG, SOG, LOG