Stu dy on leaf information collection of spring maize under different water nitrogen treatment conditions based on ground-based multispectrum

doi:10.6048/j.issn.1001-4330.2024.10.005

Abstract

Abstract:

【Objective】 To investigate the effects of different water and nitrogen treatments on multispectral collection of leaf information from spring maize. 【Methods】 Three levels of irrigation quotas (75%, 100%, 125% ETc) and four levels of N application (0, 200, 400, 600 kg/hm²) were set, Ground-based multispectral photography was used to obtain spectral information of spring maize leaves, and five vegetation indices were selected to study the effects of different water and nitrogen treatments on multispectral information acquisition. Combined with the measured data, the BP neural network with correlation analysis and particle swarm optimization was used to analyze the trend of the measured values and vegetation index. 【Results】 The results showed that the inversion of vegetation index NDVI was better for SPAD values at the middle of vegetation development, and both irrigation and nitrogen application affected the inversion of vegetation index for SPAD values. The inversions of vegetation indices OSAVI and SAVI for surface soil moisture under medium irrigation treatment (W₂) were superior, and the PSO-BP neural network modeling results of OSAVI with soil moisture data from 0 to 20 cm were better than those of SAVI for soil moisture from 10 to 30 cm. 【Conclusion】 In summary, it is more accurate to use NDVI with SOAVI for inversion of SPAD values and soil moisture at the surface 0-20 cm at 100% ETc irrigation level and above 400 kg/hm² N application level.

Key words: spring maize; multispectral; water nitrogen; neural network

摘要：

【目的】 研究不同水氮处理对多光谱采集春玉米叶片信息的影响。【方法】 设置3个水平灌水定额(75%、100%、125% 作物需水量ETc)和4个水平的施氮量(0、200、400、600 kg/hm²)处理,采用地基多光谱拍摄的方法获取春玉米叶片的光谱信息,选取5个植被指数分析不同水氮处理对多光谱信息采集的影响。结合实测数据处理,分析相关性、粒子群优化的BP神经网络变化,研究实测值与植被指数的变化趋势。【结果】 植被指数NDVI在植株发育中期对SPAD值的反演效果较好,灌水量和施氮量均会影响植被指数对于SPAD值的反演。在中灌水处理(W₂)条件下植被指数OSAVI与SAVI对表层土壤水分的反演较优,且OSAVI与0~20 cm土壤水分数据的PSO-BP神经网络建模优于SAVI对于10~30 cm土壤水分的PSO-BP神经网络建模。【结论】 在100%ETc灌水水平、施氮400 kg/hm²条件下,使用NDVI与SOAVI进行SPAD值和地表0~20 cm的土壤水分的反演较为准确。

关键词: 春玉米, 多光谱, 水氮, 神经网络

CLC Number:

S513

LI Chi, CHEN Gang, YANG Jige, YANG Tingrui, ZHAO Jinghua, MA Mingjie. Stu dy on leaf information collection of spring maize under different water nitrogen treatment conditions based on ground-based multispectrum[J]. Xinjiang Agricultural Sciences, 2024, 61(10): 2374-2387.

李池, 陈刚, 杨继革, 杨庭瑞, 赵经华, 马明杰. 基于地基多光谱的不同水氮处理条件下春玉米叶片信息采集[J]. 新疆农业科学, 2024, 61(10): 2374-2387.

Figures/Tables 18

Tab.1 Soil pH and trace element content

土壤深度 Soil depth (cm)	pH值 pH value	有机质 Organic matter (g/kg)	全氮 Total nitrogen (g/kg)	全磷 Total phosphorus (g/kg)	全钾 Total potassium (g/kg)	碱解氮 Nitrogen alkali digestion (mg/kg)	速效磷 Fast-acting phosphorus (mg/kg)	速效钾 Fast-acting potassium (mg/kg)
0~20	8.30	17.880	0.830	0.890	18.022	62.916	11.810	140.200
20~40	7.88	17.283	0.787	0.879	18.000	60.123	7.540	135.200

Fig.1 Average temperature and rainfall distribution in 2022

Fig.2 Plan view of the main factors

Fig.3 Planting pattern diagram

Tab.2 Test programme

处理名称 Treatments name	灌水定额 Irrigation quota	氮肥施用量 Nitrogen fertilizer application rate (kg/hm²)
W₁N₁	75% ETc	0
W₁N₂	75% ETc	200
W₁N₃	75% ETc	400
W₁N₄	75% ETc	600
W₂N₁	100% ETc	0
W₂N₂	100% ETc	200
W₂N₃	100% ETc	400
W₂N₄	100% ETc	600
W₃N₁	125% ETc	0
W₃N₂	125% ETc	200
W₃N₃	125% ETc	400
W₃N₄	125% ETc	600

Tab.3 Vegetation index calculation method and provenance

植被指数 Vegetation index	全称 Full name	计算公式 Calculation formula
NDVI	归一化植被指数	$(N I R - R) / (N I R + R)$ ^[10]
CCCI	冠层叶绿素含量指数	$(N I R - R_{r e d e d g e}) / (N I R + R_{r e d e d g e})$ ^[11]
GRVI	比值植被指数	$N I R / G$ ^[12]
SAVI	土壤调整植被指数	$1.5 (N I R - R) / (N I R + R + 0.5)$ ^[13]
OSAVI	优化调节土壤植被指数	$(1 + 0.16) (N I R - R) / (N I R + R + 0.6)$ ^[14]

Tab.3 Vegetation index calculation method and provenance

植被指数 Vegetation index	全称 Full name	计算公式 Calculation formula
NDVI	归一化植被指数	$(N I R - R) / (N I R + R)$ ^[10]
CCCI	冠层叶绿素含量指数	$(N I R - R_{r e d e d g e}) / (N I R + R_{r e d e d g e})$ ^[11]
GRVI	比值植被指数	$N I R / G$ ^[12]
SAVI	土壤调整植被指数	$1.5 (N I R - R) / (N I R + R + 0.5)$ ^[13]
OSAVI	优化调节土壤植被指数	$(1 + 0.16) (N I R - R) / (N I R + R + 0.6)$ ^[14]

Tab.4 Correlation between vegetation index and SPAD values

处理名称 Treatments name	归一化植被指数 NDVI	比值植被指数 GRVI	优化调节土壤植被指数 OSAVI	土壤调整植被指数 SAVI	冠层叶绿素含量指数 CCCI
W₁N₁	0.630^*	0.307	0.4	0.4	0.303
W₂N₁	0.646^*	0.194	0.309	0.321	0.331
W₃N₁	0.834^*	0.577^*	0.436	0.453	0.634^*
W₁N₂	0.910^*	0.433	0.173	0.181	0.898^*
W₂N₂	0.655^*	0.195	0.395	0.401	0.267
W₃N₂	0.566	0.605^*	0.534	0.548	0.681^*
W₁N₃	0.636^*	0.077	0.353	0.368	0.584^*
W₂N₃	0.516	0.22	0.416	0.439	0.257
W₃N₃	0.748^*	0.651^*	0.165	0.178	0.706^*
W₁N₄	0.591^*	0.357	0.192	0.2	0.342
W₂N₄	0.816^**	0.609^*	0.232	0.236	0.459
W₃N₄	0.169	0.154	0.343	0.34	-0.069

Tab.5 Correlation between vegetation index and soil moisture from 0 to 20 cm

处理名称 Treatments name	归一化植被指数 NDVI	比值植被指数 GRVI	优化调节土壤植被指数 OSAVI	土壤调整植被指数 SAVI	冠层叶绿素含量指数 CCCI
W₁N₁	0.696	0.614	-0.761^*	-0.740^*	0.578
W₂N₁	0.029	0.214	-0.790^*	-0.795^*	0.076
W₃N₁	-0.214	-0.003	-0.500	-0.501	-0.209
W₁N₂	0.016	-0.391	-0.868^**	-0.880^**	0.227
W₂N₂	-0.106	-0.241	-0.791	-0.695	-0.099
W₃N₂	-0.201	-0.017	-0.752^*	-0.764^*	0.007
W₁N₃	0.046	0.439	-0.529	-0.533	0.067
W₂N₃	-0.059	0.134	-0.746^*	-0.763^*	0.136
W₃N₃	-0.758^*	-0.751	-0.597	-0.605	-0.574
W₁N₄	-0.399	-0.403	-0.677	-0.667	-0.479
W₂N₄	0.283	-0.031	-0.766^*	-0.767^*	0.235
W₃N₄	0.108	-0.704	-0.806^*	-0.815^*	0.085

Tab.6 Correlation between vegetation index and soil moisture from 10 to 30 cm

处理名称 Treatments name	归一化植被指数 NDVI	比值植被指数 GRVI	优化调节土壤植被指数 OSAVI	土壤调整植被指数 SAVI	冠层叶绿素含量指数 CCCI
W₁N₁	-0.256	-0.566	0.259	0.261	-0.268
W₂N₁	-0.129	0.141	-0.689	-0.695	0.058
W₃N₁	-0.307	-0.166	-0.253	-0.256	-0.363
W₁N₂	0.137	-0.105	-0.789^*	-0.782^*	0.196
W₂N₂	0.039	0.18	-0.728	-0.716^*	-0.022
W₃N₂	0.164	0.177	-0.857^**	-0.861^**	0.122
W₁N₃	-0.238	-0.418	-0.829^*	-0.834^*	-0.276
W₂N₃	-0.145	-0.13	-0.901^**	-0.910^**	-0.158
W₃N₃	0.015	-0.486	-0.782	-0.667	-0.403
W₁N₄	-0.026	-0.179	-0.834^*	-0.829^*	-0.221
W₂N₄	-0.483	-0.550	-0.446	-0.448	-0.721
W₃N₄	-0.354	-0.950^**	-0.44	-0.462	-0.472

Fig.4 Trend of NDVI SPAD changes by treatment

Fig.5 Trend of soil moisture changes in OSAVI 0 to 20 cm by treatment

Fig.6 Trend of soil moisture changes in SAVI 10 to 30 cm by treatment

Fig.7 Comparison of the predictions of the BP neural network with the PSO-BP neural network for NDVI-SPAD values

Fig.8 Regression results of PSO-BP neural network

Fig.9 Comparison of the predictions of OSAVI-0-20 cm soil moisture BP neural network and PSO-BP neural network

Fig.10 Regression results of PSO-BP neural network

Fig.11 Comparison of the predictions of SAVI-10~30 cm soil moisture BP neural network and PSO-BP neural network

Fig.12 Regression results of PSO-BP neural network

References 31

[1]	张园梦. 施氮量对膜下滴灌高产(15 000 kg/hm²)春玉米生长发育及产量效应研究[D]. 石河子: 石河子大学, 2020.
	ZHANG Yuanmeng. Effect of Nitrogen Application on Growth and Yield of Spring Maize with High Yield(15,000 Kg/hm²)under Drip Irrigation[D]. Shihezi: Shihezi University, 2020.
[2]	仇焕广, 李新海, 余嘉玲. 中国玉米产业:发展趋势与政策建议[J]. 农业经济问题, 2021, 42(7): 4-16.
	QIU Huanguang, LI Xinhai, YU Jialing. China maize industry: development trends and policy suggestions[J]. Issues in Agricultural Economy, 2021, 42(7): 4-16.
[3]	李少昆, 赵久然, 董树亭, 等. 中国玉米栽培研究进展与展望[J]. 中国农业科学, 2017, 50(11): 1941-1959. DOI
	LI Shaokun, ZHAO Jiuran, DONG Shuting, et al. Advances and prospects of maize cultivation in China[J]. Scientia Agricultura Sinica, 2017, 50(11): 1941-1959. DOI
[4]	许海涛, 冯晓曦, 许波, 等. 氮水协同对玉米干物质和氮素累积与氮素运移及利用效率的影响[J]. 新疆农业科学, 2022, 59(12): 2957-2968. DOI
	XU Haitao, FENG Xiaoxi, XU Bo, et al. Effects of nitrogen and water collaborative supply on accumulation and distribution of dry matter and nitrogen, nitrogen transport and use efficiency of corn[J]. Xinjiang Agricultural Sciences, 2022, 59(12): 2957-2968. DOI
[5]	孟庆岩, 顾行发, 余涛, 等. 我国民用卫星遥感应用现状、问题与趋势[C]// 中国地震学会空间对地观测专业委员会.中国地震学会空间对地观测专业委员会成立大会暨学术研讨会论文集. 地震杂志社(Earthquake Magazine Press), 2008:8.
	MENG Qingyan, GU Xingfa, YU Tao, et al. Current Situation, Problems and Trends of Remote Sensing Applications of Civilian Satellites in China[C]// The Committee on Space-to-Earth Observation of the Chinese Seismological Society. Proceedings of the Inaugural Meeting and Symposium of the Space-to-Earth Observation Committee of the Chinese Seismological Society. Earthquake Magazine Press, 2008:8.
[6]	马仪, 黄组桂, 贾江栋, 等. 基于无人机-卫星遥感升尺度的土壤水分监测模型研究[J/OL]. 农业机械学报:1-24 [2023-04-22].
	MA Yi, HUANG Zugui, JIA Jiangdong, et al. Soil moisture monitoring model based on UAV-Satellite remote sensing scale-up[J/OL]. Transactions of the Chinese Society for Agricultural Machinery:1-24 [2023-04-22].
[7]	郑超磊, 贾立, 胡光成. 高分一号卫星遥感数据驱动ETMonitor模型估算16 m分辨率蒸散发及验证[J]. 遥感学报, 2023, 27(3): 758-768.
	ZHENG Chaolei, JIA Li, HU Guangcheng. Evapotranspiration estimation and validation at 16 m resolution based on ETMonitor model driven by GF-1 satellite remote sensing datasets[J]. National Remote Sensing Bulletin, 2023, 27(3): 758-768.
[8]	闫成川, 曲延英, 陈全家, 等. 基于无人机多光谱影像的棉花SPAD值及叶片含水量估测[J]. 农业工程学报, 2023, 39(2): 61-67.
	YAN Chengchuan, QU Yanying, CHEN Quanjia, et al. Estimation of cotton SPAD value and leaf water content based on UAV multispectral images[J]. Transactions of the Chinese Society of Agricultural Engineering, 2023, 39(2): 61-67.
[9]	李诗瑶, 丛士翔, 王融融, 等. 基于无人机多光谱遥感的干旱胁迫下玉米冠层SPAD值监测[J/OL]. 干旱区地理:1-14 [2023-04-22].
	LI Shiyao, CONG Shixiang, WANG Rongrong, et al. Monitoring of maize canopy SPAD value under drought stress based on UAV multi-spectral remote sensing.[J/OL]. Arid Land Geography:1-14 [2023-04-22].
[10]	Chen L, Gao J, Chang M, et al. Effects of Spatial Density of Farmland Shelterbelts on NDVI on the Northern Slope of Tianshan Mountains[J]. Asian Agricultural Research, 2022, 14(10):4.
[11]	冯文斌. 基于无人机多光谱遥感的夏玉米长势监测及产量估测[D]. 泰安: 山东农业大学, 2022.
	FENG Wenbin. Summer corn growth monitoring and yield estimation by UAV-based Multispectral remote sensing[D]. Tai'an: Shandong Agricultural University, 2022.
[12]	Gitelson A A, Kaufman Y J, Merzlyak M N. Use of a green channel in remote sensing of global vegetation from EOS-MODIS[J]. Remote Sensing of Environment, 1996, 58(3): 289-298.
[13]	许民, 宜树华, 叶柏生, 等. 植被盖度及太阳/观测角度对疏勒河上游NDVI和SAVI值的影响[J]. 干旱区资源与环境, 2012, 26(5): 101-107.
	XU Min, YI Shuhua, YE Baisheng, et al. Influence of PVC and Sun/view geometry on NDVI and SAVI in the upstream regions of Shule River Basin[J]. Journal of Arid Land Resources and Environment, 2012, 26(5): 101-107.
[14]	杨祯婷. 基于无人机多光谱遥感的夏玉米长势监测研究[D]. 杨凌: 西北农林科技大学, 2022.
	YANG Zhenting. Study on Monitoring of Summer Maize Growth Based on Multi-spectral Remote Sensing of UAV[D]. Yangling: Northwest A & F University, 2022.
[15]	林位衡, 黄文骞, 李广会, 等. 基于粒子群优化神经网络的水深反演模型[J]. 海洋测绘, 2020, 40(5): 26-29.
	LIN Weiheng, HUANG Wenqian, LI Guanghui, et al. A model for water depth retrieval based on neural network optimized by particle swarm optimization[J]. Hydrographic Surveying and Charting, 2020, 40(5): 26-29.
[16]	崔乃丹. 基于粒子群优化算法与BP神经网络的高铁客运量预测算法[J]. 自动化技术与应用, 2022, 41(4): 148-150.
	CUI Naidan. High speed railway passenger volume prediction algorithm based on particle swarm optimization algorithm and BP neural network[J]. Techniques of Automation and Applications, 2022, 41(4): 148-150.
[17]	祝子涵, 谭峰, 田芳明, 等. 基于BP神经网络的粳稻种子拉曼光谱鉴别方法研究[J/OL]. 中国粮油学报:1-12 [2023-03-10].
	ZHU Zihan, TAN Feng, TIAN Fangming, et al. Research on the raman spectroscopic identification of japonica rice seeds based on BP Neural Network[J/OL]. Journal of the Chinese Cereals and Oils Association:1-12 [2023-03-10].
[18]	杨晓慧, 王碧胜, 孙筱璐, 等. 冬小麦对水分胁迫响应的模型模拟与节水滴灌制度优化[J/OL]. 作物学报:1-15 [2023-04-22].
	YANG Xiaohui, WANG Bisheng, SUN Xiaolu, et al. Modeling the response of winter wheat to deficit drip irrigation for optimizing irrigation schedule[J/OL]. Acta Agronomica Sinica:1-15 [2023-04-22].
[19]	Zhang J, Liu F L, Wang Q H, et al. Effect of light wavelength on biomass, growth, photosynthesis and pigment content of Emiliania huxleyi (Isochrysidales, Cocco-lithophyceae)[J]. Journal of Marine Science and Engineering, 2023, 11(2): 456.
[20]	张智韬, 劳聪聪, 王海峰, 等. 基于FOD和SVMDA-RF的土壤有机质含量高光谱预测[J]. 农业机械学报, 2020, 51(1): 156-167.
	ZHANG Zhitao, LAO Congcong, WANG Haifeng, et al. Estimation of desert soil organic matter through hyperspectra based on fractional-order derivatives and SVMDA-RF[J]. Transactions of the Chinese Society for Agricultural Machinery, 2020, 51(1): 156-167.
[21]	Roig-Oliver M, Fullana-Pericàs M, Bota J, et al. Adjustments in photosynthesis and leaf water relations are related to changes in cell wall composition in Hordeum vulgare and Triticum aestivum subjected to water deficit stress[J]. Plant Science: an International Journal of Experimental Plant Biology, 2021, 311: 111015.
[22]	Pang Y, Huang Y, Zhou Y, et al. Identifying spectral features of characteristics of Sphagnum to assess the remote sensing potential of peatlands: A case study in China[J]. MIRES and PEAT, 2020,26.
[23]	汪沛, 陈海波, 李就好. 控制灌溉下甘蔗叶水势、叶片光谱反射率和土壤含水量的关系[C]// 中国农业工程学会农业水土工程专业委员会,云南农业大学水利水电与建筑学院.现代节水高效农业与生态灌区建设(上). 云南大学出版社(Yunnan University Press), 2010:9.
	WANG Pei, CHEN Haibo, LI Jiuhao. The relationship Among sugarcane leaf water potential, leaf spectral reflectance and soil moisture under controlled irrigation[C]// China Agricultural Engineering Society, College of Water Conservancy, Hydropower and Construction, Yunnan Agricultural University. Modern water-saving and efficient agriculture and ecological irrigation district construction (above). Yunnan University Press, 2010:9.
[24]	Zhou G F, Moayedi H, Bahiraei M, et al. Employing artificial bee colony and particle swarm techniques for optimizing a neural network in prediction of heating and cooling loads of residential buildings[J]. Journal of Cleaner Production, 2020, 254: 120082.
[25]	贺露, 万莉, 高会议. 基于高光谱成像技术识别番茄干旱胁迫[J]. 光谱学与光谱分析, 2023, 43(3): 724-730.
	HE Lu, WAN Li, GAO Huiyi. Recognition of drought stress in tomato based on hyperspectral imaging[J]. Spectroscopy and Spectral Analysis, 2023, 43(3): 724-730.
[26]	祝榛, 李天胜, 崔静, 等. 基于高光谱成像估测冬小麦不同生育时期水分状况[J]. 新疆农业科学, 2022, 59(3): 521-532. DOI
	ZHU Zhen, LI Tiansheng, CUI Jing, et al. Study on estimation of water status of winter wheat in different growth stages based on hyperspectral imaging[J]. Xinjiang Agricultural Sciences, 2022, 59(3): 521-532. DOI
[27]	李庆禄. 渍水胁迫下小麦高光谱特征分析与主要生理参数估测研究[D]. 扬州: 扬州大学, 2021.
	LI Qinglu. Hyperspectral characteristics analysis and estimation of main physiological parameters of wheat under waterlogging stress[D]. Yangzhou: Yangzhou University, 2021.
[28]	韩康, 于静, 石晓华, 等. 不同光谱指数反演马铃薯叶片氮累积量的研究[J]. 作物学报, 2020, 46(12): 1979-1990. DOI
	HAN Kang, YU Jing, SHI Xiaohua, et al. Inversion of nitrogen accumulation in potato leaf with different spectral indices[J]. Acta Agronomica Sinica, 2020, 46(12): 1979-1990. DOI
[29]	努热曼古丽·托乎提, 聂臣巍, 余汛, 等. 基于高光谱指数的叶片尺度叶绿素荧光参数反演[J]. 玉米科学, 2021, 29(5): 73-80.
	Nuerman Tuohuti, NIE Chenwei, YU Xun, et al. Retrieval of leaf-scale chlorophyll fluorescence parameters based on hyperspectral index[J]. Journal of Maize Sciences, 2021, 29(5): 73-80.
[30]	王俞茜. 华北地区夏玉米水分和氮素光谱诊断模型研究[D]. 天津: 天津农学院, 2021.
	WANG Yuqian. Study on Spectral Diagnostic Model of Water and Nitrogen of Summer Maize in North China[D]. Tianjin: Tianjin Agricultural University, 2021.
[31]	李雪, 顾沈明, 年浩. 改进粒子群算法优化BP神经网络的粮食产量预测[J]. 漳州师范学院学报(自然科学版), 2014, 27(1): 56-61.
	LI Xue, GU Shenming, NIAN Hao. Prediction for grain yield based on improved PSO optimized BP neural network[J]. Journal of Minnan Normal University (Natural Science), 2014, 27(1): 56-61.