中文摘要：

跟踪初生盐渍土壤的微生物修复实验,采用同步实测得土壤盐含量和光谱数据,详细分析了基于34种光谱变换,修复过程中盐渍土的光谱特征。对于选取的6种光谱变换,采用全波段（400-1650 nm）和分析获得的最佳敏感波段分别建立了土壤盐含量的光谱反演PLSR（partial least squares regression）模型。研究表明,光谱变换处理使土壤盐含量与平滑后的光谱反射数据的相关性明显增强,且最佳敏感波段范围进一步聚焦。本研究得到最佳光谱变换为导数变换,基于全波段的土壤盐含量预测模型以SGSD变换效果最好,与原始光谱相比,模型的r、RMSEP分别从0.537和1.928改善到0.823和1.256。而SGSD（Log R）是基于最佳波段所建立的盐含量预测模型的有效光谱变换方法,该研究为进一步实现盐渍土中盐含量快速定量分析提供了方法和数据参考。

英文摘要：

In this paper, soil salinity content （SSC） and its spectral reflectance were measured during the microbial remediation process of saline soil. The aim of the paper was to analyze and compare the effects of 34 pre-processing methods on spectral characteristics of saline soil during the remediation process. Partial least squared regression （PLSR） analysis was then used to predict SSC based on reflectance spectra by using full bands （400 - 1650 nm） and the optimal sensitive bands for 6 selected pre-processing methods. The results showed that spectral pre-processing methods had the advantage of enhancing the correlation of SSC and smoothed reflectance spectra, and the range of optimal sensitive bands was further focused. The derivative turned out to be the best pre-processing methods in this study, and the prediction accuracy of SGSD was the best in full bands. Compared to the raw reflectance spectra （R）, the corresponding r and RMSEP of the predicted model were improved, respectively, from 0.537 and 1.928 to 0.823 and 1.256. Based on optimal sensitive bands of PLSR predicting models of SSC, SGSD（LogR） obtained more robust calibration and prediction accuracies than other pre-processing inversion models. The results obtained in this study provided a method and data reference for further quantitative analysis of SSC in saline soil quickly.