摘要:以宁夏银北地区典型龟裂碱土为研究对象，表层土壤光谱反射率选择平滑、倒数等7种数据处理方式，采用全回归、逐步回归和偏最小二乘三种回归方法，分析龟裂碱土光谱特征，筛选对土壤pH值和ESP的敏感波段，建立龟裂碱土碱化信息的预测模型。结果表明：龟裂碱土的光谱反射曲线属于缓斜型；土壤表层反射率与土壤pH值和ESP在研究波段内均呈极显著正相关关系；反射率倒数对数的一阶微分和反射率的一阶微分在特征波段范围表现较好；反射率与土壤pH值的相关性优于与土壤ESP的相关性。从拟合度和选用敏感波段的多少整体考虑，采用偏最小二乘回归来拟合土壤pH值和ESP的方程最佳，拟合度分别达到0.93和0.8367。

关键词:龟裂碱土  光谱特征  碱化信息

AbstractBy using the typical Takyr Solonetzs in northern Yinchuan as research object, seven methods were selected to process the reflectance data of surface soil, and total regression, stepwise regression and partial least squares regression were adopted to analyze the spectral characteristics of Takyr Solonetzs, with the purpose of determining the sensitive wavelengths to pH and ESP of surface soil and establishing the prediction model of alkalization information. The results show that: the reflectance curve of Takyr Solonetzs belonged to slow-oblique type; there were significant positive correlations between the reflectance and pH as well as ESP of surface soil; the methods of reciproca l, first-order differential of reciprocal logarithmic and first-order differential of reflectance were relatively good in the characteristic wavelength range; and the relativity between reflectance to pH was higher than that to ESP. Considering both the fitting degree and the quantity of sensitive wavelengths, partial least squares regression was the best method to estimate pH and ESP of surface soil, whose fitting degree were 0.93 and 0.8367 respectively.

Key wordsTakyr Solonetzs  characteristic reflectance  alkalization information

文章编号： 0258_7106 (2016) 01_0018_15 中图分类号： P618.41 文献标志码：A

改回日期：2015_07_11

基金项目

**通讯作者耿新霞， 女， 1979年生， 助理研究员， 成矿规律研究方向。 Email： gen gxinxia@cags.ac.cn