Research Of Spectral Characteristics Of Water Content In Different Textured Soils Of Xinjiang

This research takes the farmland soil from blunt flood product fan middle and lower reaches soil in manas county of xinjiang province as subject, using HR-768field-portable spectroradiometer made by the Spectra Vista Corporation(SVC) of America obtained by laboratory experiment and field spectral date, applying a series of methods, to analyze the spectral character and determine the bands sensitive at laboratory experiment worked and field soil, to analyze the correlation coefficients of spectral character and the different textured soil content; To establish the estimate model of spectral prediction soil content; Explore the influence of different soil conditions of laboratory experiment and field geometry spectral character and determine the optimal indoor spectral acquisition geometry condition; Through the ENVI create the xinjiang different textured soil content of the spectral library to create.The results show:(1). Spectral curves of different texture soils had the same trend, but the order of the reflectivity was: Clay> Silty clay loam> Sandy loam.(2). The critical points of field capacity in sandy loam, silty clay loam and clay were20.01%,24.10%and30.43%, respectively. Within the range of water content which were below the critical points, water content and spectral reflectance had inverse proportion, but above the points, then reverse.(3). Within the1390-1623nm band, the correlation coefficients of soil water content and normalized spectra reflectance were all above-0.9and significant at the0.01level, R2of the model established for soil water content prediction exceeded mostly0.8, and the average relative error of the model was10%, model accuracy is better.(4). Through the different mathematical transformation, soil prediction model of soil water content spectral established, inspection found, forecasting the soil moisture content is the best model by continuum removal of linear equations,R2of the model was0.887and the average relative error of the model was12.3%.(5). When different conditions of laboratory experiment and field geometry spectral, spectral curves had the same trend, but the order of the reflectivity was different. Through the variance analysis of geometric condition, such as the angle of the beam, the distance between sensor and soil surface, the distance between lamp and soil surface,15°,30cm and40cm are recommended as optimization values of angle, sensor distance and lamp distance. It was better reflect laboratory geometric condition of soil sample spectral curve of the consistency of the influence.