| 引用本文: | 韩 康,于 静,李 瑞,等.基于光谱特征值反演马铃薯叶片氮累积量[J].灌溉排水学报,2021,(6):53-58. |
| HAN Kang,YU Jing,LI Rui,et al.基于光谱特征值反演马铃薯叶片氮累积量[J].灌溉排水学报,2021,(6):53-58. |
|
| 摘要: |
| 【目的】明确多种光谱特征值与滴灌马铃薯叶片氮累积量的定量关系。【方法】在大田滴灌模式下，选用不同马铃薯品种为试验材料，设置5个氮肥水平，分别为0、150、300、450、600 kg/hm2，测定了马铃薯各生长阶段光谱特征值与叶片氮累积量，并对二者进行相关分析和回归分析。【结果】马铃薯各生长阶段光谱特征值红蓝边面积比值[RI(SDr，SDb)]与马铃薯叶片氮累积量（LNA）关系较密切，为敏感光谱特征值。在出苗后20、35、50、65 d，RI(SDr，SDb)与LNA的数量关系均为二次项模型，而全生育期基于RI(SDr，SDb)与LNA拟合的最佳监测模型为指数模型。出苗后20、35 d以及全生育期基于RI(SDr，SDb)构建的叶片氮累积量最佳监测模型可定量描述不同氮水平下的马铃薯LNA，且不受生长环境的影响，为马铃薯关键生育阶段氮素营养状况的监测模型，R2高于0.7，RMSE范围分别为4.348~8.844、6.665~17.725、8.862~17.725 kg/hm2。【结论】RI(SDr，SDb)可以估测马铃薯关键生育阶段氮素营养状况。 |
| 关键词: 马铃薯；叶片氮累积量；光谱特征值；监测模型 |
| DOI：10.13522/j.cnki.ggps.2020639 |
| 分类号: |
| 基金项目: |
|
| Optimal Spectral Eigenvalues to Estimate Nitrogen Content in Potato Leaves |
|
HAN Kang, YU Jing, LI Rui, FAN Mingshou
|
|
1.Inner Mongolia Agricultural University, Hohhot 010019, China;2. Ordos Soil and Fertilizer and Water-saving Agriculture Work Station, Erdos 017010, China
|
| Abstract: |
| 【Background】Hyperspectral reflectance information from crop canopy can be used to estimate crop healthy, and the spectral eigenvalues extracted from it have been used to successfully diagnose plant nitrogen content. This can help fertilization thereby improving nitrogen use efficiency and reducing its detrimental impact on the environment. The sensitivity of the spectral eigenvalues to plant nitrogen content varied between crops and their cultivars. Currently, it remains elusive if the sensitive spectral eigenvalue obtained from one crop applies to other crops.【Objective】The aim of this paper is to study the quantitative relationship between different spectral eigenvalues extracted from hyperspectral imageries and the nitrogen content in leaves of potatoes watered by drip irrigation. 【Method】The experiments were conducted in a field grown with various potato varieties, all fertilized with a nitrogen gradient ranging from 0 to 600 kg/hm2. The relationship between nutrient content in the leaves of all varieties at different growth stages and different spectral eigenvalues was analyzed using regression models.【Result】The ratio of the red edge area to the blue edge area, RI, and the nitrogen accumulation in the leaves (LNA) were most closely correlated at all growth stages. The relationship between RI and LNA was quadratic in less than 65 days after the seedling emergence, while for the whole growing season, an exponential model was more accurate. The regression model was accurate and reliable to predict leaf nitrogen content in less than 35 days after the seedling emergence and for the whole growing season, regardless of the nitrogen fertilization and changes in the environment and crop varieties, with R2>0.7 and RMSE ranging in 4.348~8.844, 6.665~17.725 and 8.862~17.725 kg/hm2, depending on growth stages and crop varieties.【Conclusion】RI can be used as the spectral eigenvalue to estimate nitrogen content in potato leaves at different growth stages. |
| Key words: potato; leaf nitrogen content; spectral eigenvalue; predicting model |
