博斯腾湖西岸湖滨绿洲芦苇地土壤盐分特征高光谱分析

发布时间：2018-06-13 03:52

本文选题：湖滨绿洲 + 高光谱　；参考：《扬州大学学报(农业与生命科学版)》2016年01期

【摘要】：以博斯腾湖西岸湖滨绿洲为研究区,统计分析芦苇地土壤盐分特征,通过多元线性回归分析建立芦苇地土壤盐分含量、Na~++K~+和Ca~(2+)的高光谱反射率估算模型。结果表明:1芦苇地土壤盐分含量平均值为4.21g·kg~(-1),为轻度盐渍化水平,不同株高下土壤盐分含量差异较大,标准误为9.37,方差为87.75;2不同芦苇株高下土壤高光谱反射率曲线特征表现为形态相似,基本平行,光谱的形状特征大致可以由910、1 200、1 450、1 910nm 4处的吸收带控制;3基于土壤高光谱反射率对土壤盐分含量、Na~++K~+和Ca~(2+)建立高光谱反射率估算模型进行定量反演具有良好精度:盐分含量高光谱反射率估算模型均方根误差(RMSE)为0.32,回归分析决定系数(R2)为0.88;Na~++K~+高光谱反射率估算模型RMSE为0.34,R~2为0.90;Ca~(2+)高光谱反射率估算模型RMSE为0.25,R2为0.93(F95%)。
[Abstract]:Taking the oasis on the west coast of Bosten Lake as the study area, the characteristics of soil salinity in Reed land were statistically analyzed. The hyperspectral reflectance estimation model of soil salt content (Na ~ K ~ and Cajiao _ 2) of Reed soil was established by multiple linear regression analysis. The results showed that the average value of soil salt content was 4.21 g 路kg ~ (-1) ~ (-1), which was a mild salinization level, and there was a great difference in soil salt content under different plant height. The standard error was 9.37, and the variance was 87.75 / 2. The characteristics of soil hyperspectral reflectance curves under different Reed plants were similar in shape and basically parallel to each other. The shape characteristics of the spectra can be roughly controlled by the absorption band at 910 ~ 1 200 ~ 1 450 ~ 1 910nm _ 4. Based on the soil hyperspectral reflectance, the quantitative inversion of the hyperspectral reflectance estimation model based on the soil salt content (Na ~ K ~ and Cajiao _ 2) is good. Accuracy: the root mean square error (RMSE) of the model for estimating the hyperspectral reflectance of salt content is 0.32, and the regression determination coefficient (R2) is 0.88 ~ K ~ (-1). The RMSE of the model is 0.34 ~ K ~ (-1). The RMSE of the model is 0.25 R2, and the estimation model of hyperspectral reflectance (RMSE) is 0.93 (F95A).
【作者单位】：新疆师范大学地理科学与旅游学院/新疆干旱区湖泊环境与资源实验室;
【基金】：国家自然科学基金资助项目(41061025、31360039) 新疆维吾尔自治区重点实验室专项资金项目(2014KL016) 新疆师范大学开放基金课题(XJNU-DL-201316)
【分类号】：S156.4

【相似文献】