根据多光谱传感器的光谱响应函数，采用实测ISI921VF反射光谱数据模拟Landsat卫星 ETM＋传感器多光谱数据，在模拟光谱的基础上，通过光谱特征提取、构建土壤指数对土壤重金属Cu ，Pb ，As进行预测分析。研究显示，Cu ，Pb与模拟ETM＋光谱的B2，B3波段显著相关，As与DSI ，RSI ，NDSI相关系数在0．6以上，基于模拟多光谱建立的Cu ，As模型精度较高，平均相对误差分别为7．9％，2．7％，表明模拟的Landsat卫星ETM＋传感器多光谱具有预测耕地土壤重金属的潜力，为实现大范围监测土壤重金属污染提供新思路。%In this paper , based on the spectral response function of multi‐spectral sensors , measured ISI921VF reflection spectrum data are used to simulate the multispectral data of ETM+ sensor of Landsat satellite ,and based on the simulated spectrum ,the extracted spectral features and constructed soil index are used to analyze and predict the soil heavy metal of Cu ,Pb and As .The study shows the Cu and the Pb are significantly associated with band B3 ,B4 in ETM + ;the correlation coefficients between the As and index DSI ,RSI ,NDSI are above 0.6 ;based on the simulated multispectral ,the Cu prediction model and the As prediction model have high precision and the average relative errors are 7.9% and 2.7% , respectively .It is indicated that simulated multispectral data of ETM + sensor of Landsat satellitecan predicts the heavy metal in farmland soil ,which is a new way provided to achieve a wide range monitoring of soil heavy metal .
展开▼
