摘要
以宁夏平罗县龟裂碱土为研究对象,以实测植被光谱和土壤pH值为基础数据源,通过对原始光谱数据进行小波阈值去噪,和对数、一阶微分、多元散射校正、归一化等8种变换,筛选土壤碱化程度最佳光谱变换方式和敏感波段,用一阶傅里叶和三次多项式进行回归分析、比较,来构建更加精确的龟裂碱土信息预测模型。研究表明:植被光谱反射率一阶微分变换在波段861nm处为最佳敏感波段,相关系数为0.86;多项式拟合比傅里叶拟合效果好;以最佳光谱指标和土壤pH值为变量,构建的pH含量三次多项式预测模型精度最高,在0.01显著性水平上通过检验,该模型可为干旱区半干旱地区土壤碱化程度遥感定量反演提供依据。
In the study, the paper takestypical soil salinization area in Pingluo County of Ningxia Hui Autonomous Region as the research object.Based on the measured spectral reflectance of vegetation and the value of pH in the la-horatory as the basic data source, the original spectral data through threshold denoising and classification were used to analyze the spectral characteristics of different levels of soil salinization. The reflectance data were transformed to 8 kinds of spectral indices, such as logarithm, first-order derivative, multiphcative scatter correction and normaliza-tion, etc.Then,the correlation analysis was carried out between the obtained vegetation spectral and the value of pH to extract sensitive wavelengths of pH parameters. Fourier and polynomial regression analysis were employed to establish takir information prediction model. The correlation coefficient is used to verify the prediction effect of the two models. The results showed The wavelength of 861nm at the first order differential transformation of the spectral reflectance of vegetation is the best sensitive band; the correlation coefficient is 0.86. Polynomial fitting is better than Fourier fitting effect to the best spectral indices and soil pH as a variable, the content of the pH value of the building three times polynomial prediction model is the highest precision.The model can provide the basis for the quantitative remote sensing inversion of soil alkalization degree of semi-arid area.
作者
刘欢
贾科利
张俊华
Liu Huan Jia Keli Zhang Junhua(College of Resource and Environment, Ningxia University ,Yinchuan , Ningxia , 750021, China Institute of Environmental Engineering, Ningxia University, Yinchuan , Ningxia , 750021 ,China)
出处
《绿色科技》
2017年第20期1-5,共5页
Journal of Green Science and Technology
基金
国家自然科学基金(编号:41561078)
关键词
盐渍化
龟裂碱土
光谱信息
salinization
alkaline soil
spectral information
