Study On Land Use Change And Spectral Recognition Of Lycium Barbarum L. Area Based On "3S" Technique In Dulan County

Land use is a long-term,dynamic use and transformation from human.Land use change not only directly affects the sustainable development of social economy,but also has an important impact on the current ecological environment.The study of land use change has a great effect on the comprehensive utilization of land,especially agricultural land,and the coordination between economic production and land ecology.The comprehensive study of land use by remote sensing(RS)and geographic information system(GIS)is still one of the hot spots of land science research.Therefore,this study analyzes the land use change in Dulan County from 2000 to2015 by the dynamic degree and transfer matrix.On this basis,we try to use the spectral informationwhichbased on model method to identify Lycium barbarum L.,to identify the difference between Lycium barbarum L.areaand other similar landand the planting area of Lycium barbarum L.in Dulan County.The results show that the unused land is the largest typein Dulan County,accounting for about half of the total area of the county.The sum of unused land and grassland is about 80% of the total area of the county.From 2000 to 2015,the area of unused land and farmland decreased year by year.And the area of grassland,forest land and construction land increased obviously.The area of water area was stable.Among them,the larger area of change was unusedland(506.97km2),grassland(316.85km2)and woodland(up 208.22km2).In all types of transfer change matrix,the forest land transfer rate is the largest.The transfer area of unused land,grassland is relatively larger thanother types,and the water have no transfer relationship with other types.In addition,based on the compared between vegetation coverage and land use status,verifying the vegetation coverage area and distribution of grassland,woodland and farmland.They are roughly the same.The Lycium barbarum L.spectrum was collected by Landsat remote sensing image and wild field.According to the method of minimum noise Fraction-MNF and Pixel Purity Index-PPI,the target pixel were mostly non-pure pixel,which was actually the same with the Lycium barbarum L.cultivation that cannot be completely cover.Then we establish two models which the first based onthe relationship of spectral information and the second based on the method of Spectral Information Divergence(SID)and Mixture Tuned Matched Filtering(MTMF).The twovalues,the measured data,in the relationship is significantly related,but the inversion view possiblyaffected by the different things with the same spectrum and non-pure pixels.In addition,models based on the SID and MTMF methods can clearly reflect the main planting areasof Lycium barbarum L.in the inversion view and distinguish them from land use types such as unused land,woodland and farmland.