土壤含盐量是土地质量的一个重要因素。光谱反射率以快速获取和高质量等特点,在土壤研究方面广泛受到重视。选择塔里木河中游典型绿洲作为研究区,探讨了利用土壤光谱反射率来估算土壤含盐量,为该区域农业的可持续发展提供参考。首先,对土壤光谱反射率做了17种形式的变换,寻找能突现土壤含盐量的最佳的光谱波段。其次,对土壤光谱反射率与土壤含盐量进行耦合分析,建立了光谱反射率与含盐量的相关性方程。最后,使用统计软件SPSS16.0利用验证土壤验证方程的精度。结果显示,非盐渍土的最好估算方程为土壤光谱反射率的一阶微分形式,总均方根差RMSE为0.457。盐渍土的最好估算方程为对数倒数的二阶微分形式,总均方根差RMSE为0.523。结果表明,可以用土壤的光谱反射率间接测量土壤的含盐量,正确有效地指导农业生产。
Soil salt content is an important factor of land quality.Spectrum reflectivity attracts extensive attention in soil research because of its fast acquirement and high quality.Typical oasis of Tarim River's middle reaches was chosen as the study area to explore the problem of how to use soil spectral reflectivity to predict the soil salt content,and to provide the basis of sustainable development of agriculture for the region.First,seventeen transforms of the soil spectral reflectivity were conducted to look for the optimal spectral band to present soil salt content.Then,coupling analysis were carried out between the soil spectral reflectivity and soil salt content to establish the correlative equation between the two.At last,the accuracy of the model was tested through the verification of soil samples with the use of the statistical software SPSS 16.0.The results show that,the best estimation equation of non-saline soil is first differential of soil spectral reflectivity,and the total RMS difference is 0.457;while the best estimation equation of saline soil is second differential of the reciprocal of logarithm of soil spectral reflectivity,and the total RMS difference is 0.523.In all,it is feasible to use soil spectral reflectivity to predict soil salt content and guide agricultural production correctly and effectively.