Research On Remote Sensing Monitoring Method Of Land Degradation In Hulunbuir Grassland Region

Land is an essential element for human survival and reproduction.At present,about one-quarter of the land in the world is in degradation,and the situation of land degradation is extremely severe.Therefore,carrying out land degradation monitoring based on remote sensing is very important.Among the existing remote sensing monitoring methods for land degradation,the vegetation index method using net primary productivity(NPP)as an indicator is a widely used method.It monitors degraded land based on the spatiotemporal characteristics of NPP,which requires NPP accuracy higher.Existing MODIS NPP product is affected by data scale and MODIS GPP product quality,its spatial resolution and product accuracy are significantly lower.And there is a significant lack of applicability in the research on the small and medium scale area.So it is necessary to reinvert NPP in combination with regional feature.Now,the research on NPP inversion is mostly based on MODIS data,but the limitation of the data spatial scale determines the inversion result cannot characterize local feature better,resulting in large error.Therefore,the study comprehensively used the multi-source remote sensing data to improve the insufficiency of NPP inversion based on MODIS data,so as to better show the local characteristic of the region and improve the accuracy.In fact,in arid and semi-arid areas,NPP is particularly affected by precipitation.If land degradation is monitored based on NPP change,it is easy to misjudge land degradation area in years when precipitation significantly changes.Therefore,the study introduced the vegetation precipitation use efficiency(PUE)indicator for land degradation monitoring.It characterizes the effective use of water by vegetation and is defined as the ratio of NPP to precipitation.Based on the constructed PUE and land degradation correlation model,the research completed land degradation monitoring in Hulunbuir grassland region of Inner Mongolia in 2010-2018,and analyzed the driving factors of land degradation.The main results are as follows:(1)Using Landsat-8,Sentinel-2 and MOD13Q1 images as remote sensing data source,combined with meteorological data,the CASA(Carnegie-Ames-Stanford Approach)model was driven to obtain NPP data with the spatial resolution of 30 m in the study area.Compared with MODIS NPP product,the inversion result can identify the high value distribution area of NPP accurately,and the overall accuracy is higher.In addition,the comparison result of three scale NPP obtained by the above remote sensing data showed that the NPP inversion with high spatial resolution data can characterize the local feature of region clearly,which reduced misjudgment of the vegetation area and the inversion error.(2)The change trend of NPP was consistent with precipitation in the study area,indicating that NPP is indeed significantly affected by precipitation.Therefore,the PUE index was introduced to eliminate the impact of precipitation,and the land degradation monitoring was completed based on the constructed PUE and land degradation correlation model.The example verification result through Google Earth images showed that the PUE index can indicate the land degradation area well,and the monitoring result is effective.(3)The analysis result of driving factors about land degradation showed that the area of land degradation caused by single meteorological factor change in study region accounted for less than 0.1%,which had little effect on land degradation;human activity was the main influencing factor.Therefore,while developing and constructing,human should pay attention to protect the ecological environment.