Research On Identification Of Abandoned Cultivated Land In Youyang County Based On LT-GEE

Cultivated land is a limited land resource available for human survival and a basic condition to ensure human survival.Abandonment is a form of insignificant utilization of cultivated land,which threatens the national food security.In recent years,the abandonment of cultivated land has evolved into a global trend,especially in mountainous areas.At present,remote sensing technology is widely used for cultivated land abandonment monitoring at home and abroad,and some research progress has been made.However,due to the limitations of hardware conditions such as insufficient computer memory and long processing time of remote sensing data,it is still difficult to grasp the details of cultivated land abandonment in the existing research,especially the research on the long-time series of cultivated land abandonment is immature,and it is easy to overestimate or underestimate the area of abandoned cultivated land,The resulting data error is easy to mislead policy-making.Therefore,taking Youyang County of Chongqing as an example,based on Google Earth Engine(GEE)cloud platform,using SNIC segmentation method,random forest model and Land Trendr(LT)method,this paper draws the cultivated land abandonment map of Youyang County from 2001 to2020,identifies the start time point of cultivated land abandonment,and studies the temporal and spatial distribution characteristics of cultivated land abandonment in Youyang County by using global autocorrelation analysis,local autocorrelation analysis and emerging temporal and spatial hot spot analysis,Calculate the abandonment rate of cultivated land,and rely on the tool of geographical detector to explore the driving factors affecting the abandonment of cultivated land from two aspects of natural environment and social economy.The research contents and results are as follows:(1)Due to the lack of previous remote sensing data,it is sometimes difficult to obtain the annual land use classification data.This study selects sample points based on Google Earth Engine platform and uses random forest algorithm to draw the land type possibility classification map from 2001 to 2020.(2)In the past,pixel based land use classification methods failed to eliminate the noise in remote sensing images,resulting in low accuracy of classification results.Based on the object-oriented idea,this study uses SNIC segmentation technology to classify land use.The classification results show that the overall accuracy of land use classification from 2001 to 2020 is more than 80.0%,and the classification effect is good.(3)Using Land Trendr algorithm to identify abandoned farmland has a significant effect,and the accuracy of abandoned results can reach 81.0%.Through calculation,the abandonment rate of cultivated land in Youyang County from 2001 to 2020 was 9.983%.The abandoned cultivated land was concentrated in the north and central regions,and the western region was the cold spot of abandoned cultivated land.(4)During the three years of 2002-2003,2008-2009 and 2015-2016,the abandoned area of cultivated land increased significantly;In the five years of 2005-2008,2009-2010 and 2016-2017,the abandoned area of cultivated land increased positively,but on the whole,it showed a downward trend,with an annual growth rate of only 0.79% in 2008.(5)Precipitation,agricultural population and GDP are the significant driving factors affecting the abandonment of cultivated land in Youyang County.