Land Cover Change Detection In Shanxi Province Based On LT-GEE

Remote sensing technology has developed rapidly and is widely used in land use/cover change(LUCC)research as an important means of obtaining spatial and temporal changes.However,when remote sensing technology is used in land use/cover change research,it is often limited by problems such as limited computer storage space and long processing time of remote sensing data.Therefore,this paper retrieves Landsat images from 1986 to 2017 in Shanxi Province from the Google Earth Engine database to produce a long-term sequence of Landsat remote sensing image datasets.On the Google Earth Engine platform,LandTrendr time series algorithm and supervised classification algorithm are combined to generate long-term sequence change results of various types in Shanxi Province,and the land use/cover change of various regions in Shanxi Province is analyzed from the time point of view.The duration of change and the impact of changes in land use/cover change with economic changes.From the perspective of space,the changes of various years in various cities of Shanxi Province,the changes of various cities and the changes of slope grades on land use/cover change are analyzed.The results show that:(1)LandTrendr time series segmentation algorithm has obvious effects in detecting land use/cover change.And the LandTrendr land use/cover change result is extracted according to the supervised classification,and then the superposition analysis of the regional change results between different time periods can effectively analyze the land use/cover transfer situation.According to tests,the cultivated land,forest land and grassland in Shanxi Province have changed drastically,and the construction land and water area have changed less.From the late 1980 s to the 1990 s,the area of cultivated land,woodland,grassland and water changes changed more than the area changed after 2000.The change in the area of construction land is opposite to that of other land types.The five types of land transfer have a large area of mutual transfer,and the area transferred to cultivated land,construction land,forest land,grassland and water body is the most during the period of 2010-2015.(2)The LandTrendr time series segmentation algorithm can effectively detect the duration of land use/cover change.According to tests,the changes in the cultivated land,construction land,forest land,grassland and water bodies in Shanxi Province have a short duration and last for one year.(3)Slope and economy have a severe impact on land use/cover change in Shanxi Province.The changes of five land types such as cultivated land,construction land,forest land,grassland and water body in Shanxi Province are mainly distributed between 0° and 15°.When the slope is greater than 15°,the ratio of the area of change of each type to the total area of each category is significantly reduced.The economic development of Shanxi Province has a severe impact on the changes of these five land types.The secondary industry,which is the leading industry before 2014,has a greater impact on these five land types.The tertiary industry,which is the leading industry after 2014,has a greater impact on these five types of land types.The results of various studies show that it is necessary to use LandTrendr in combination with the supervised classification algorithm to automatically monitor the land use/cover change work in various places,and to detect land use/cover change qualitatively,quantitatively and quantitatively.On the one hand,it provides an important source of data for China's “finding the bottom of the family”,and on the other hand,it adds assistance to China's land use supervision system and saves costs.