Research On The Temporal And Spatial Changes Of Grassland Coverage In The Source Area Of ??the Three Rivers In The Region From 2000 To 2016 Based On Remote Sensing Data

The Three-River Headwaters Region(TRHR)is located in the interior of the Qinghai-Tibetan Plateau,which is a typical research area in East Asia of fragile environment and high ecological value.The main vegetation cover type in TRHR is grassland which conserves water resource,preserves soil,and grazes livestock.The grassland cover is traditionally studied by in-situ measurement,but it is inefficient in grassland cover change monitoring due to large area of grassland.Thus,the remote sensing images,product indicators,and methods were much more effective in researches of grassland cover change.This paper studied the characteristics of grassland cover changes in the TRHR between 2000 and 2016 using method of area division(AD)based on natural conditions of elevation,air temperature,and precipitation,and the tabulate area(TA)method based on terrain factors of elevation,aspect,and slope was also used dependent on Moderate-resolution Imaging Spectroradiometer(MODIS)44B product.Further investigations were conducted on some of the typical areas to determine the characteristics of the changes and discuss the driving factors behind these changes.The AD and TA methods were also utilized to illustrate the distribution rules of the corresponding coefficient between grassland cover of MOD44B and hydrothermal factors of air temperature,land surface temperature,and precipitation.Classification And Regression Trees(CART),Random Forest(RF),Bayesian(BAYE),and Support Vector Machine(SVM)Machine Learning(ML)methods were employed to evaluate the correlation between grassland cover changes and corresponding variables.The CART,RF,BA YE,and SVM machine learning algorithms were also used to simulate and evaluate the grassland cover of year 2015 based on the training set of year 2016,then made prediction of grassland cover of year 2017 through the same training set.The overall trend for grassland cover in the TRHR towards recovery that rose 0.91%during the 17-year study period.The results showed that:(1)The change in grassland cover was more divisive in similar elevation and temperature conditions when the precipitation was stronger.The higher the temperature was,the more significant the rise of grassland cover was in comparable elevation and precipitation conditions.(2)There was a distinct decline and high change standard deviation of grassland cover in some divided areas,and strong correlations were found between grassland cover change and aspect,slope or elevation in these areas.(3)The main influence factors of grassland cover across years were elevation and precipitation,while the main influence factors of grassland cover within a year were air temperature and elevation,the time difference when grassland cover reached the maximum value among different areas in TRHR within a year could be up to 32 days.(4)In the study of the corresponding coefficient between grassland cover of MOD44B and hydrothermal factors of air temperature,land surface temperature,and precipitation,observable linear relationships were found between the corresponding coefficient and the terrain factors of elevation and slope.(5)The study methods of AD and TA achieved enhancing performance in interpretation of grassland cover changes in the broad and high elevation variation areas.(6)RF and CART methods showed higher stability and accuracy in application of grassland cover change study in TRHR among the four ML methods utilized in this study,while the RF algorithm had better performance of accuracy in the simulation,evaluation,and prediction of grassland cover in TRHR than the CART,BA YE,and SVM algorithms.