Spectral Characteristics Of Saline-alkali Patch Salt And PH In Oasis Farmland

The formation of soil salinization is usually caused by the change of natural conditions and the use of land by human beings. Due to the long dry climate, scarce rainfall, strong salt accumulation, resulting in a large number of Saline-alkali patch in farmland, seriously affect the ecological environment of the oasis and the healthy development of agriculture. This paper selects Manas County farmland topsoil?0-10cm? Saline-alkali patch soil as the research object, analysis of saline alkali spot soil spectral curve features, mining, soil salinity and pH sensitive spectral bands, soil salt content and pH value were the best estimation model is established. Based on the Landsat8 remote sensing image, the regional salinization remote sensing monitoring model?SDI? was established by using the NDVI-SI feature space relationship, and the distribution map of the different degree of Salinization Soil in the study area was made by this model. Through the above research, Main conclusions are listed below:?1? Saline-alkali patch soil spectral reflectance curve trend is consistent, the spectral features in the form of basic approximation; spectral reflectance curve in 1450 nm and 1950 nm and 2200 nm near has three obvious absorption peaks, and with the degree of soil salinization aggravated gradually and reflectance showed increasing trend, soil spectral reflectance and surface salinity show certain correlation.?2? Soil spectral reflectance by seven kinds of differential transform, enlarge the spectral curve features, correlation coefficient; the spectral range of the high correlation coefficients generally appear between the 400-1259 nm. One of the reflectivity first-differential, logarithmic derivative, the reciprocal of count first-derivative and salt content correlation coefficient compared to the original reflectance were improved.?3? By several transformation of soil spectral reflectance and soil containing salt modeling analysis that establish the spectral reflectance of the logarithm of the reciprocal of the first derivative differential effect of salinity inversion model best response to salinity of the best sensitive wavelength at 428 nm, the correlation coefficient reached 0.81, construct the salinity of the prediction equation?y=30097.5x428+38.9, R²=0.742, RMSE=0.836?. This model can provide a scientific basis for quantitative inversion of soil salinization in oasis farmland.?4? Using five kinds of differential transform spectral data and Saline-alkali patch soil pH data correlation analysis found that: spectral reflectance logarithm of reciprocal first derivative differential constructed model is better, soil pH sensitive bands for 681 nm, 909 nm, 954 nm, 997 nm predicted equation(Y=20669.81X₉₀₉-2784.23X₆₈₁-27948.27X₉₉₇+5089.03X₉₅₄+7.53, R²=0.714, RMSE=0.35). The accuracy of the model is higher, and the model can provide reference for estimating the pH value of the soil in the study area.?5? The relationship between vegetation index?NDVI? and salinity index?SI? was analyzed by using Landsat8 image and field measured salinity data, and a remote sensing monitoring model?SDI? based on NDVI-SI feature space was established. The results show that the SDI and surface soil containing salt high correlation, the correlation coefficient was 0.83. The model construction is conducive to soil salinity remote sensing image interpretation, and can very good area of different levels of saline soil spatial distribution of future regional scale.