基于高光谱反射特性的土壤全氮含量预测分析

发布时间：2018-06-17 15:31

本文选题：土壤全氮 + 高光谱　；参考：《遥感技术与应用》2017年01期

【摘要】：随着高光谱遥感技术的快速发展,光谱技术已经在土壤理化性质、土壤养分等预测研究中得到了广泛应用。通过土壤高光谱反射率及其变形全氮含量的相关性,提取土壤光谱特征波段;采用多元回归和偏最小二乘回归法对全氮含量进行预测分析。结果表明:土壤光谱一阶微分显著提高了全氮与高光谱之间的敏感度;在多元逐步线性回归模型和偏最小二乘回归分析法建立的模型中,二者均能较好地进行预测,但在偏最小二乘模型中,反射率二阶微分的预测模型最高达到0.956,总均方根误差最低为0.045。其模型的稳定性和预测精度优于多元逐步线性回归所建立模型,可以更好地快速预测土壤全氮,为土壤质量的评价提供数据基础,也为研究土壤退化地区的预测与防治提供信息,对未来农业的发展具有重要意义。
[Abstract]:With the rapid development of hyperspectral remote sensing technology, spectral technology has been widely used in the prediction of soil physical and chemical properties and soil nutrients. Based on the correlation of soil hyperspectral reflectance and deformed total nitrogen content, the characteristic bands of soil spectrum were extracted, and the content of total nitrogen was predicted by multivariate regression and partial least square regression. The results showed that the sensitivity between total nitrogen and hyperspectral was significantly improved by the first order differential of soil spectrum, and both of them could be well predicted in the multivariate stepwise linear regression model and the partial least square regression model. However, in the partial least squares model, the prediction model of second-order differential reflectivity is the highest 0.956, and the total root mean square error is the lowest 0.045. The stability and prediction accuracy of the model is better than that of the model established by multivariate stepwise linear regression. The model can predict soil total nitrogen more quickly and provide the data basis for the evaluation of soil quality. It also provides information for the prediction and prevention of soil degradation areas, which is of great significance to the development of agriculture in the future.
【作者单位】：南京信息工程大学环境科学与工程学院;南京信息工程大学大气环境与装备技术协同创新中心;南京信息工程大学应用气象学院;
【基金】：国家科技支撑计划项目(2012BAD16B0305、2012BAC23B01) 中国清洁发展机制基金(2013013)资助
【分类号】：S151.93

【相似文献】