干旱区盐渍化土壤高光谱遥感信息分析与提取

Analysis and extraction of saline soil hyperspectral remote sensing information for arid area

以干旱区典型区域新疆渭干河-库车河三角洲绿洲为研究区,以环境小卫星高光谱影像及野外实测土壤含盐量为主要数据源,进行光谱反射率及其变换形式与土壤含盐量的相关性分析,筛选盐渍化土壤响应最敏感波段,利用多元线性回归分析方法,建立基于HSI影像的研究区土壤含盐量定量反演模型。结果表明:研究区土壤含盐量与HSI波段的敏感性随着波长的增加而增强,位于近红外波段范围(797.826-923.913nm)的相关系数R普遍较高,基本在0.7左右。土壤光谱反射率对数的倒数一阶微分变换在628.261nm和923.913nm的波段组合为最佳敏感波段,所构建的土壤含盐量反演模型为最优模型,模型方程为Y=-11.731-114.996X628.261-186.637X923.913,模型及检验的决定系数R2都在0.85以上,均方根误差RMSE约为2.7。该模型的建立为地区土壤含盐量信息的提取及监测提供了参考。

Choosing typical arid area of Weigan and Kuqa Rivers delta oasis as a study area,this study proposed a multiple linear regression model to assess the degree of soil salinization,based on the correlation between soil salinity and HSI reflectance variation and screening sensitive wavebands on soil salinity of HSI images.The results show that the sensitivity of soil salinity and HSI band increased with increasing of spectrum wavelength,and the sensitive wavebands was located in the near infrared band range（797.826-923.913nm）,the correlation coefficient R was larger,generally about 0.7.The model using the reciprocal of HSI reflectance logarithm in first order differential could get the best result（Regression equation：Y=-11.731-114.996X628.261-186.627X923.913）,the coefficient determination R2 was higher than 0.85 and the RMSE was about 2.7.The establishment of the model provided reference for regional soil salinity information extraction and monitoring.