宣城市岗坡地土壤有机质含量光谱预测分析

利用荷兰AvanSpect2048×14型号地物光谱仪,在实验室条件下,对采自于安徽省宣城市岗坡地29个土壤样品进行反射率光谱数据的采集.对所获取的土壤样品高光谱反射率数据进行标准化比值变换,计算有机质诊断指数,同时将该指数与土壤有机质含量进行相关性分析,寻找最大正相关中心波段.结果表明:最大正相关中心波段为489.87nm,两个次正相关中心波段为481.16nm、490.45nm.随机选取20个样品,对三个中心波段的有机质诊断指数与有机质含量分别进行回归分析,其中以在489.87nm处线性回归模型拟合精度最佳,用该线性模型对其他9个样品土壤有机质含量进行预测,精度达84.660%,表明该线性模型对土壤有机质含量具有较好的预测效果.

Spectral Prediction on Soil Organic Matter Content of the Slope in Xuancheng City

The article uses AvanSpect 2048×14 models spectral analyzer made in Netherlands,under laboratory conditions,to collect reflectance data from 29 soil samples obtained in slopes of Xuancheng City,Anhui Province.Through normalized ratio transformation of hyperspectral reflectance data of soil samples,the author calculates the diagnostic index of organic matter and then makes the correlative analysis with soil organic matter content to find the best positive correlation center band.The result shows that the maximum positive correlation center band is 489.87nm,and next two onespositive correlation center band are 481.16nm,490.45nm.To select 20 samples randomly and make regression analysis on the diagnostic index and organic matter content of three central bands respectively,the result shows that the fitting accuracy of linear regression model at the 489.87nm is best.And then the author uses the linear model to predict soil organic matter content with other 9 samples,and the accuracy reaches up 84.660%.It shows that the linear model is applicable to predict soil organic matter content.