全生育期夏玉米氮素利用率垂直分布的高光谱定量反演与动态监测

Nitrogen utilization (assimilation) efficiency (NUtE) measures the ability to use N and the reasonable allocation of N, indicating the impact of N on plant growth and photosynthetic productivity. As one of the most mobile indicators, NUtE exhibits a pronounced heterogeneity in its vertical distribution in crop canopies. Although the vertical heterogeneity of crop NUtE has been recognized, the vertical NUtE gradient in canopies has not been considered in canopy hyperspectral remote sensing (CHRS) so far. Additionally, summer maize is a crop with short growth period and has a dilution effect on the accuracy of NUtE estimation by models derived from CHRS. Therefore, it is not very effective to apply CHRS technique to assess NUtE during different growth periods under various N rates of summer maize, because the NUtE estimation considering vertical gradients is influenced by natural dynamics of N nutrition and growth periods. Although N nutrition and growth periods are inherently correlated, current studies usually treat them as independent factors when analyzing the impacts of N nutrition and growth periods on the vertical NUtE distribution assessment. The study will investigate the effects of the interactive changes in N nutrition and growth periods on the vertical distribution of NUtE as well as on the evaluation of the capabilities of CHRS to determine the vertical distribution of NUtE within summer maize canopies. Moreover, considering the discontinuity of the effects of the two abovementioned factors on NUtE and canopy spectra, the stability and utility of the monitoring model based on the vertical distribution patterns of NUtE was tested using continuous wavelet transformation, partial least square and support vector machine technique. The study will provide theoretical basis and effective methods for applying CHRS technology to real-time detection of NUtE and high yield and efficiency of crop precision management for summer maize.

氮素利用率（NUtE）是表征作物氮吸收利用性能的核心指标，在植株内分布具有明显的垂直特征。然而目前利用冠层高光谱技术在定量反演NUtE时往往忽视其固有的垂直分布特性。夏玉米是一种生育期短、氮“稀释效应”显著且响应敏感的作物，使得基于高光谱的NUtE垂直分布监测同时受氮营养和生育期双因子联动变化影响，但目前研究均将氮营养和生育期视为相对独立因子分析其与高光谱间内在关系，所构建反演模型很难应用于不同生长阶段并做到对氮素的稳定响应。本研究以夏玉米为分析对象，探究氮营养和生育期联动变化对其NUtE垂直分布及高光谱动态监测精度影响。此后，考虑到氮水平和生育期变化对夏玉米NUtE以及冠层光谱影响的不连续性，研究高光谱经连续小波变换后，采用偏最小二乘回归和支持向量机的全波段分析方法动态监测NUtE垂直分布的准确性和适用性。研究结果将为夏玉米NUtE实时监测和丰产增效的氮素管理技术提供理论基础与有效方法。

夏玉米是一种生育期短、氮“稀释效应”显著且响应敏感的作物，使得基于高光谱的氮营养垂直分布监测同时受氮营养和生育期双因子联动变化影响，所构建反演模型很难应用于不同生长阶段并做到对氮素的稳定响应。基于此，本项目首先在叶片尺度上系统探明了夏玉米全时域氮素营养生化指标与叶片SPAD值垂直分布特征及两者间定量关系，确立基于叶绿素仪的夏玉米氮营养无损诊断敏感叶位和叶片部位；进一步阐明了不同叶层氮素营养高光谱分布特征及光谱监测有效叶层。在群体尺度上明确了氮营养和生育期联动变化对其氮素利用率（NUtE）垂直分布及高光谱动态监测精度影响。研究了高光谱经连续小波变（CWT）换后，采用偏最小二乘回归（PLS）的全波段分析方法动态监测叶片氮浓度（LNC）和NUtE垂直分布的准确性和适用性。项目主要结论如下：1）叶片尺度上，不同叶位间夏玉米叶片SPAD值和氮营养指标于植株间分布均呈典型的“钟型”特征，至TL5或TL6时达至峰值。同一叶位不同部位间SPAD值由20%至100%位点时则逐步升高，且80%~100%位点间无显著差异。基于PLS模型中各叶位和叶片部位无量纲评价指标变量重要性投影值（VIP），确定TL4完展叶60%~80%区间为夏玉米氮营养诊断的敏感区域，平均VIP值分别为1.691和1.648，预测效果较为理想。2）群体尺度上，夏玉米LNC和NUtE冠层空间分布具有典型的垂向异质性变化规律，即上层最低，下层最高，且中下层（第二层和第三层）LNC与NUtE为冠层高光谱技术实时和定量监测的有效和关键叶层；CWT变换光谱可显著提高夏玉米全生育期不同叶层LNC和NUtE冠层高光谱预测精度；夏玉米LNC与NUtE有效波段虽在可见光-红边-近红外区均有明显和均匀分布，但在不同生育期和叶层间却表现出明显的差异性，即生育后期（吐丝期-乳熟期）相比于前期（拔节期-大喇叭口期），植株下层相比于上层其LNC和NUtE有效波段均呈规律性的长波“红移”现象。研究结果将为夏玉米氮素营养实时监测和丰产增效的氮素管理技术提供理论基础与有效方法。