论文部分内容阅读
为研究土壤养分含量分布信息,以从北京郊区一块试验田采集的72个土壤样品为试验材料,应用傅里叶变换近红外光谱技术分析了土样的全氮、全钾、有机质养分含量和pH值。采用偏最小二乘法(PLS)对光谱数据与土壤养分实测值进行回归分析,建立预测模型,以模型决定系数(R2)、校正标准差(RMSECV)、预测标准差(RMSEP)和相对分析误差(RPD)作为模型精度的评价指标。结果表明,利用该模型与光谱数据对土壤全氮、全钾、有机质养分含量和pH值进行预测,结果与实测数据具有较好的一致性,最高决定系数R2达到0.9544。偏最小二乘回归方法建立的养分预测模型能准确地对北京地区褐土土质全氮、有机质、全钾和pH值4种养分进行预测。
In order to study the distribution of soil nutrient content, 72 soil samples collected from a test field in the suburbs of Beijing were used as experimental materials. The contents of total nitrogen, total potassium, organic matter and pH of soil samples were analyzed by Fourier transform near infrared spectroscopy . The PLS was used to conduct regression analysis between the spectral data and the measured values of soil nutrients, and the prediction model was established based on the model determination coefficient (R2), RMSECV, RMSEP and relative analysis error RPD) as the evaluation index of model accuracy. The results showed that using the model and spectral data to predict soil total nitrogen, total potassium and organic matter nutrient contents and pH value, the results were in good agreement with the measured data, and the maximum determination coefficient R2 was 0.9544. The nutrient prediction model established by partial least squares regression method can accurately predict the four nutrients in the cinnamon soil of Beijing area, including total nitrogen, organic matter, total potassium and pH value.