As an essential feature of plant autotrophy, Nitrogen (N) is the major nutrient affecting plant growth in terrestrial ecosystems, thus is of not only fundamental scientific interest, but also a crucial factor in crop productivity. Timely non-destructive monitoring of canopy nitrogen concentration (N%) demands fast and highly accurate estimation, which is often quantified using spectroscopic analyses in the 400—2500 nm spectral region. However, extracting a set of useful spectral absorption features from canopy spectra to determine N% remains challenging due to confounding canopy architecture. Deep Learning as a statistical learning technique is useful to extract biochemical information from canopy spectra. We evaluated the performance of a one-dimensional convolutional neural network (1D-CNN) and compared it with two state-of-the-art methods: partial least squares regression (PLSR) and gaussian process regression (GPR). We utilized a large and diverse in-field multi-season (autumn, winter, spring and summer) spectral database (n = 7014) over 8 years (2009–2016) of dairy and hill country farms across New Zealand to develop season specific and spectral-region specific (VNIR and/or SWIR) 1D-CNN models. Results on the independent validation dataset (not used to train the model) showed that the 1D-CNN model provided higher accuracy (R² = 0.72; nRMSE% = 14) than PLSR (R² = 0.54; nRMSE% = 19) and GPR (with R² = 0.62; nRMSE% = 16). Season specific models based on 1D-CNN indicated apparent differences (14 ≤ nRMSE ≤19 for the test dataset), while the performance of all seasons combined model was remained higher for the test dataset (nRMSE% = 14). The full spectral range model showed higher accuracy than the spectral region-specific models (VNIR and SWIR alone) (15.8 ≤ nRMSE ≤18.5). Additionally, predictions derived using 1D-CNN were more precise (less uncertain) with <0.12 mean standard deviation (uncertainty intervals) than PLSR (0.31) and GPR (0.16). This study demonstrated the potential of 1D-CNN as an alternative to conventional techniques to determine the N% from canopy hyperspectral spectra.

氮（N）是植物自养的重要特征，是影响陆地生态系统植物生长的主要营养素，因此不仅具有根本的科学意义，而且还是作物生产力的关键因素。对冠层氮浓度（N％）进行及时的非破坏性监测需要快速且高度准确的估算，通常使用400-2500 nm光谱区域中的光谱分析法对其进行量化。但是，由于冠层结构混杂，从冠层光谱中提取一组有用的光谱吸收特征以确定N％仍然具有挑战性。深度学习是一种统计学习技术，可用于从冠层光谱中提取生化信息。我们评估了一维卷积神经网络（1D-CNN）的性能，并将其与两种最新方法进行了比较：偏最小二乘回归（PLSR）和高斯过程回归（GPR）。我们利用了一个大型且多样化的现场多季节（秋季，冬季，春季和夏季）光谱数据库（n  = 7014）在8年（2009-2016年）的整个新西兰范围内的奶牛场和山地农场，以开发特定季节和特定光谱区域（VNIR和/或SWIR）的1D-CNN模型。独立验证数据集（未用于训练模型）的结果表明，一维CNN模型提供的准确性（R ²  = 0.72； nRMSE％= 14）比PLSR（R ²  = 0.54； nRMSE％= 19）和GPR更高。 （带有R ² = 0.62; nRMSE％= 16）。基于1D-CNN的特定季节模型显示出明显的差异（测试数据集为14≤nRMSE≤19），而测试数据集的所有季节组合模型的性能仍然更高（nRMSE％= 14）。全光谱范围模型显示出比特定光谱区域模型（仅VNIR和SWIR）更高的准确性（15.8≤nRMSE≤18.5）。此外，与PLSR（0.31）和GPR（0.16）相比，使用1D-CNN得出的预测更精确（不确定性更低），平均标准偏差（不确定区间）<0.12。这项研究证明了1D-CNN替代传统技术从冠层高光谱光谱中确定N％的潜力。