论文部分内容阅读
作物关键生育时期冠层氮素含量的实时监测对于优化氮肥用量和减少环境风险具有重要的意义。为了寻求预测不同作物氮素含量的最佳光谱参数,实现作物氮素无损营养诊断。本研究通过2008年—2011年在德国慕尼黑弗莱辛和河北曲周的不同氮量的小麦玉米田间试验,采用高光谱仪获取小麦玉米冠层的反射光谱,利用光谱理论模型进行光谱指数波段的优化,从而抽取不同冠层结构条件下的小麦玉米氮素营养敏感波段。结果表明与传统的基于红光的光谱指数相比,优化光谱指数显著提高了小麦玉米冠层氮素含量的预测能力,克服了传统的基于红光光谱指数的饱和问题。优化光谱指数的波段结合随着作物品种及其冠层结构的变化而变化,其优化波段范围主要集中在红边(730~760nm)和红边向近红外的过渡区域(760~880nm)。优化结果显示玉米最佳光谱指数为R_(λ766)/R_(λ738)-1,小麦最佳光谱指数为R_(λ796)/R_(λ760)-1,玉米小麦相结合优化后的最佳光谱指数为R_(λ876)/R_(λ730)-1。结果进一步验证了优化光谱指数估测的不同作物含氮量的预测值与实测值相关性最高,且验证偏差最小,证实了优化后的光谱特征参数可对不同作物氮素丰缺状况进行快速、准确、无损估测。试验结果也为设计作物冠层氮素传感器和更好的利用现有基于卫星的传感器实施区域上的作物氮素营养监测提供了理论基础。
Real-time monitoring of canopy nitrogen content during key crop growth stages is of great importance for optimizing nitrogen use and reducing environmental risks. In order to find the best spectral parameters for predicting nitrogen content in different crops, crop nitrogen nutrition diagnosis is realized. In this study, field experiments were conducted on wheat and corn with different nitrogen contents in Freising, Hebei province and Quzhou, Hebei during 2008-2011. The reflectance spectra of wheat canopy were obtained by using hyperspectral spectroscopy, and the spectral index band was optimized by spectral theory , So as to extract the sensitive bands of wheat nitrogen under different canopy structure. The results showed that compared with the traditional spectral index based on red light, the optimized spectral index significantly improved the prediction of wheat nitrogen content in canopy, which overcomes the traditional saturation problem based on red spectral index. The band of optimized spectral index changes with crop variety and its canopy structure. The optimized band mainly focuses on the transition from red edge (730 ~ 760nm) and red edge to near infrared (760 ~ 880nm). The optimized results showed that the optimum spectral index of maize was R_ (λ766) / R_ (λ738) -1, and the optimum spectral index of wheat was R_ (λ796) / R_ (λ760) Is R_ (λ876) / R_ (λ730) -1. The results further verify that the predicted value of nitrogen content of different crops with the optimized spectral index has the highest correlation with the measured value and the minimum of the verification deviation. It is confirmed that the optimized spectral characteristic parameters can quickly and easily determine the nitrogen abundance and deficiency of different crops, Accurate, non-destructive estimation. The results also provide a theoretical basis for designing crop canopy nitrogen sensors and better monitoring crop nitrogen nutrition in areas where existing satellite-based sensors are used.