不同干扰程度的盐渍土与其光谱反射特征定量分析

发布时间：2018-05-14 12:18

本文选题：干扰程度 + 盐渍土　；参考：《光谱学与光谱分析》2017年02期

【摘要】：通过对新疆阜康500水库下游的盐渍化土壤实地定点取样和光谱测量,利用光谱变换、相关分析等方法,定量探讨了不同人为干扰程度的土壤盐分、水分与光谱反射率之间的关系,并建立了土壤反射光谱与盐分含量之间的多元线性回归预测模型。结果表明:(1)人为干扰程度与土壤盐分呈极显著正相关,而与土壤水分呈极显著负相关,相关系数分别为0.961和-0.929。(2)在不同干扰程度与土壤光谱反射率的关系中,重度干扰的土壤反射率比轻度干扰土壤的反射率高10%,比未干扰高17%。这是由于人为干扰破坏了土壤表面的少量植被及生物、物理结皮,土壤表层因缺乏保护,水分会迅速蒸发,并将土壤下部的盐分带到上部,加之降水稀少,盐分在表层聚集。干扰程度越高,结皮破坏越严重,土壤积盐越多,反射率越高。(3)随干扰程度的不断增加,土壤原始光谱反射率与盐分相关系数的两个最大值逐渐向近红外波段偏移(999,876~979,1 182~1 370和1 900nm),这预示着,在近红外区土壤光谱反射率对盐分含量更为敏感。(4)利用反射率R、反射率一阶导数R′、反射率R+水分分别建立了不同干扰程度的三类土壤盐分含量预测模型。综合R~2和RMSE判断模型精度,在不同干扰程度下,同类型的土壤含盐量预测模型中,干扰程度越小,模型精度越高;而在相同干扰程度下,不同类型的土壤含盐量预测模型中,均以一阶导数R′建立的模型预测效果最优,R~2均超过0.983。总体上,模型精度提高了5%~10%,表明原始光谱经过一阶导数变换处理,可以去除部分线性背景值的干扰,提高预测土壤含盐量的精度。
[Abstract]:Through sampling and spectrum measurement of salinized soil in the lower reaches of Fukang 500 Reservoir in Xinjiang, the soil salinity with different degree of human interference was quantitatively discussed by means of spectral transformation and correlation analysis. The relationship between water content and spectral reflectance, and a multivariate linear regression prediction model between soil reflectance spectrum and salt content was established. The results showed that the degree of human disturbance was significantly positively correlated with soil salinity, but negatively correlated with soil moisture. The correlation coefficient was 0.961 and -0.929. 2 respectively. The reflectivity of heavily disturbed soil was 10% higher than that of mild disturbance and 17% higher than that of non-interference. This is due to human disturbance to destroy a small amount of vegetation and organisms on the soil surface, physical crust, soil surface for lack of protection, water will evaporate rapidly, and the lower part of the soil salt to the upper, coupled with the lack of precipitation, salt accumulation in the surface layer. The higher the interference degree, the more serious the crust damage, the more salt accumulated in the soil, the higher the reflectivity. The two maximum values of the original spectral reflectance and the salt correlation coefficient of the soil gradually shifted to the near infrared band. In the near infrared region, soil spectral reflectance is more sensitive to salt content. 4) three kinds of soil salt content prediction models with different degree of interference were established by using reflectivity R, first derivative R and R water. By synthesizing RX2 and RMSE, the precision of the model is higher in the same type of soil salinity prediction model with different degree of interference, but in the same degree of interference, in different types of soil salt content prediction model, the lower the degree of interference is, the higher the precision of the model is. The prediction results of the models based on the first derivative R 'are all above 0.983. On the whole, the accuracy of the model has been improved by 5% and 10%, which indicates that the original spectrum can remove the interference of partial linear background value and improve the precision of predicting soil salt content after the first order derivative transformation.
【作者单位】：新疆大学资源与环境科学学院教育部绿洲生态重点实验室;北京联合大学应用文理学院城市系;
【基金】：国家自然科学基金项目(41171165) 北京市属高等学校高层次人才引进与培养计划项目(IDHT20130322)资助
【分类号】：S156.4;O657.3

【相似文献】