Inverse Study Of Sand Particle Size Based On Remote Sensing Images

Land desertification seriously affects regional ecological balance and seriously damages the ecological environment.Desertification assessment is one of the key issues in the study of desertification,which make us grasp the situation of regional desertification timely and accurately.In Golmud,the climate is arid and windy all year round.Desertification is one of the most serious environmental problems.In this paper,the sand particle size index was imported to explore the method and process of using single point remote sensing inversion of surface sand size,which is of great significant to improve the evaluation index system and improve the remote sensing image of desertification efficiency research.In this study we use Landsat8 OLI as data source.The Hapke model was introduced and made a simplification and improvement,which proposed a single scatter albedo inversion model for single angle spectral data.The simplified model is used to retrieve the single scattering albedo of sand in each wave band of pretreatment images.The average particle size of the samples obtained by field measurement.Optimal input variables and their coefficients,which can influence the change of sand particle size were obtained through GA-PLS,which established the statistical regression model between measured sand particle size and single scattering albedo.So as to explore the method and process of estimating the surface sand particle size based on single angle remote sensing image data.The conclusions of this paper are as follows: As an important index of desertification assessment,Particle size of sand,which provide effective basis in the evaluation of regional land desertification especially the desertification with wind erosion.Research shows that the size of sand particles in the area becomes larger with deepen of wind erosion desertification degree.Simplified Hapke model can well adept to the single scattering albedo inversion in single angle spectral data of sandy desertification of desertification areas,and achieved good results.The single scattering albedo of each band were successful inversed base on improved Hapke model,and the trends of single scattering albedo can reflect the change regulation of sand particle size,to a certain extent.The GA-PLS model,which widely used in high spectral quantitative analysis also be demonstrated to be well fit in multi spectral data and particle size inversion.Research shows that by calculating the band ratio of single scattering albedo,the input variable of GA-PLS were added,therefore the model was successfully applied to the inversion of multi spectral sand particle size and achieved good results.Estimate the sand particle size of study area from 2013 to 2015 using inverse model.Analysis results show that apart from 2014 when the area were badly influenced by human factors,2013-2014 and 2015-2016,the average particle size in study area becomes larger due to the wind erosion and other factors.