Dynamic Monitoring Of Vegetation Participants In Typical Oasis Based On Sentinel-2 Data

Oasis is the basis for the survival of humans and various living creatures in the arid regions of northwest China.Vegetation is the key to maintaining the existence and healthy development of oasis.Especially in the national policy-Lucid waters and lush mountains are invaluable assets,it is of great significance to achieve accurate,efficient,economical and wide-ranging dynamic monitoring of vegetation in the oasis,and to timely grasp the development of oasis and protect the oasis ecological environment.Remote sensing can achieve this by effectively monitoring several major vegetation parameters.At this stage,the use of remote sensing technology for vegetation parameter monitoring mainly relies on vegetation parameter products and a few high-resolution remote sensing data.However,the spatial resolution of common vegetation parameters based on MODIS and AVHRR data is low(>250m),which makes it difficult to achieve high-precision vegetation monitoring in the oasis region;and WorldView-2,ZY-3,and GF-1 data is costly and generally lacks spectral bands that are sensitive to vegetation.In view of the above problems,this study takes the typical oasis in the middle reaches of the Tarim River-the Werigan-Kuqia River Delta Oasis(referred to as " WKRDO ")as the target area,using economical,high precision(10m),sensitive to vegetation.The Sentinel-2 data combined with the random forest model,PROSPECT+SAIL radiation transmission model,selected four typical vegetation parameters(vegetation coverage,canopy chlorophyll content,photosynthetically active radiance,canopy water content)to achieve vegetation in the oasis Dynamic monitoring on a monthly basis.The study found that:(1)Sentinel-2 data showed high estimation accuracy under both models(R2>0.6).For the canopy chlorophyll content,the random forest model reached R2= 0.826;the radiation transfer model achieved a high precision of R2= 0.925.(2)Under different vegetation cover,the relationship between the parameters of 4 plantings and soil properties is different.The 4 plots under cotton,jujube and corn cover were higher in autocorrelation,above 0.8,and closely related to soil properties of various soil layers;the autocorrelation of vegetation parameters under reed and Haloxylon was low,some were About 0.6,and only 1 to 2 planting parameters are closely related to soil properties.(3)Vegetation parameters vary significantly with the seasons,from April to November each year is the growth period of local vegetation;from May to June is the fastest growing period of vegetation;the variation of vegetation parameters during the period from June to September is small;7 August is the peak of vegetation parameters in the region.(4)Vegetation parameters and salinization degree are reversed in the interior and periphery of the oasis.The degree of salinization is heavier in October each year,and the degree of salinization is the lightest in July and August.This study provides a reference method and data source for vegetation monitoring in the oasis.In the future,we will conduct continuous time parameter monitoring for one or two economic crops in the oasis,in order to better serve local precision agriculture while achieving higher inversion accuracy.