原状水稻土土壤湿度高光谱特征及反演模型

发布时间：2018-05-11 01:26

本文选题：原状土 + 吸收峰特征参数　；参考：《土壤通报》2017年01期

【摘要】：室内高光谱遥感定量反演土壤湿度的相关研究,可以为利用星载或机载遥感数据源进行大尺度土壤水分监测提供理论基础;原状土土样相较于当前广泛使用的研磨土土样,能更好地保持土壤的原始光谱特性和持水性能。为了得到更可靠的土壤湿度光谱特征与更具有适用性的反演模型,利用ASD Field Spec 3光谱仪采集了江苏省金坛市竹林镇的62个原状水稻土土样12套不同湿度梯度的可见光-近红外反射光谱,通过反射率、相对反射率以及水分吸收峰特征参数探究原状湿土光谱特征;在此基础上,选择对土壤湿度变化响应明显的变量,采用偏最小二乘回归反演土壤湿度,并对比选出最佳反演模型。结果表明:(1)随着风干过程中土壤湿度的降低:土壤反射率整体先下降后上升,可见光波段在土壤湿度较低时存在误差;1450 nm和1950 nm吸收峰的位置向短波方向偏移,吸收峰深度、宽度和面积不断减小,2 200 nm处吸收峰右肩高度不断降低。(2)土壤水分吸收峰特征参数对土壤湿度响应明显;反射率和相对反射率与土壤湿度相关性较高的波段均出现在1300 nm以后。(3)反射率(R)反演原状土湿度精度较低,R2val=0.752,RPD=1.81,经过数学形式和相对反射率(Rr)变换后,[lg(1/Rr)]'建模中R2val=0.820,RPD=1.99;基于水分吸收峰特征参数建立的模型R2val=0.800,RMSEP=62.70 g kg-1,RPD=2.29,最佳主成分数为3,是反演原状土土壤湿度的最佳建模变量。
[Abstract]:The related research on quantitative retrieval of soil moisture by hyperspectral remote sensing can provide a theoretical basis for large-scale soil moisture monitoring using space-borne or airborne remote sensing data sources. It can better maintain the original spectral properties and water retention of the soil. In order to obtain more reliable soil moisture spectral characteristics and more applicable inversion model, ASD Field Spec _ 3 spectrometer was used to collect 12 sets of visible and near infrared reflectance spectra of 62 undisturbed paddy soil samples from Zhulin Town, Jintan City, Jiangsu Province. The spectral characteristics of undisturbed wet soil were investigated by the relative reflectance and the characteristic parameters of water absorption peak, on the basis of which, the soil moisture was retrieved by partial least square regression, and the variables that were obviously responsive to the change of soil moisture were selected. The optimal inversion model is compared and selected. The results showed that with the decrease of soil moisture during air drying, the soil reflectivity decreased first and then increased. The position of absorption peaks at 1450nm and 1950 nm was shifted to the shortwave direction and the depth of the absorption peaks in the visible light wave region was lower than that in the low soil moisture. The width and area of the absorption peak at 2200nm decreased continuously (the height of the right shoulder decreased) the characteristic parameters of the soil moisture absorption peak showed obvious response to soil moisture. The bands with high correlation between reflectivity and relative reflectivity appear after 1300 nm.) reflectivity and R) inversion of undisturbed soil moisture is less accurate. After mathematical form and relative reflectance transform, [lg1 / rr] 'model, R2val1 0.820 RPD-1.99; The model R2valn 0.800 ~ (-1) RMS EPN 62.70 g / kg ~ (-1) RPD-2.29, with an optimum principal fraction of 3, is the best modeling variable for retrieving the soil moisture of undisturbed soil.
【作者单位】：南京农业大学资源与环境科学学院;
【基金】：中央高校基本科研业务费专项资金(KYZ201522) 江苏高校优势学科建设工程项目国家自然科学基金项目(41171071)资助
【分类号】：S152.7

【相似文献】