铜污染植被指数的玉米叶片污染程度探测模型

引用本文

程凤, 杨可明, 崔颖, 陆天宇, 陈立帆, 荣坤鹏. 铜污染植被指数的玉米叶片污染程度探测模型[J]. 光谱学与光谱分析, 2020,40(1): 209-214.
CHENG Feng, YANG Ke-ming, CUI Ying, LU Tian-yu, CHEN Li-fan, RONG Kun-peng. A Model on Detecting the Polluted Degree of Maize Leaves by Cu Pollution Vegetation Index[J]. Spectroscopy and Spectral Analysis, 2020,40(1): 209-214.
Doi:10.3964/j.issn.1000-0593(2020)01-0209-06 复制到剪切板

Permissions

《光谱学与光谱分析》期刊社所有

铜污染植被指数的玉米叶片污染程度探测模型

程凤, 杨可明^*, 崔颖, 陆天宇, 陈立帆, 荣坤鹏

中国矿业大学(北京)煤炭资源与安全开采国家重点实验室, 北京 100083

*通讯联系人 e-mail: ykm69@163.com

作者简介: 程凤, 女, 1992年生, 中国矿业大学(北京)煤炭资源与安全开采国家重点实验室硕士研究生 e-mail: 369179776@qq.com

收稿日期: 2018-11-17

基金: 国家自然科学基金项目(41971401, 41271436), 煤炭资源与安全开采国家重点实验室2017年开放基金课题基金项目(SKLCRSM17KFA09)资助

摘要

高光谱遥感监测农作物重金属污染已成为遥感研究的重要内容之一。受污染的作物叶片中重金属含量映射到光谱上的信息量差异较微弱, 如何灵敏地挖掘其所包含的价值信息具备一定挑战性。以农作物叶片光谱为研究对象, 通过多个光谱特征波段组合的方式, 提出了一种铜污染植被指数(CPVI)的污染程度探测模型, 来表征重金属Cu对农作物的污染程度。首先设置盆栽实验, 将不同浓度梯度的CuSO₄·5H₂O粉末添加到土壤中, 模拟Cu污染土壤环境, 胁迫玉米生长。采集玉米穗期的老、中、新叶片光谱, 测定叶片中Cu²⁺含量及相对叶绿素浓度。而后利用随机选取的58组玉米叶片光谱作为实验数据, 在380~900 nm波长范围内选取波长 λ₁和 λ₂的两组叶片光谱反射率并计算相应的CPVI[ λ₁, λ₂]模型指数及其与对应叶片中Cu²⁺含量的皮尔逊相关系数, 得到相关性特征绝对值矩阵。其次, 根据得到的相关性特征绝对值矩阵, 提取皮尔逊相关系数较高的光谱特征波段690和465 nm, 并结合波段850 nm建立针对玉米叶片的铜污染植被指数(CPVI_m)。之后, 利用另外26组数据对CPVI_m指数进行检验, 同时将该指数与归一化植被指数(NDVI)、陆地叶绿素指数(MTCI)等常规植被指数进行比较以验证CPVI_m的有效性与优越性。结果表明, NDVI, MTCI, REP和DVI与叶片中Cu²⁺含量相关系数最高仅为0.68, 残差平方和RSS最低为70.99, 而CPVI_m与叶片中Cu²⁺含量显著负相关, 相关系数达-0.80, 残差平方和为48.52, 均优于NDVI和MTCI等常规植被指数, 证明CPVI_m对重金属胁迫更敏感。同时利用两期不同年份不同品种的玉米光谱数据进行CPVI_m指数的鲁棒性验证, CPVI_m与叶片Cu²⁺含量的相关系数 r分别为-0.90和-0.96, 均显著相关, 说明该指数对于不同品种的玉米污染程度探测仍具有良好的适用性。另外, 利用玉米叶片中Cu²⁺含量、 CPVI_m和叶片中叶绿素相对浓度构建三维分析模型, 从空间角度直观地反映了三者之间具有一定的相关关系。通过光谱特征波段组合方式构建的CPVI探测模型可作为评价农作物重金属污染程度的参考方法, 基于该方法构建的CPVI_m指数可有效甄别玉米受重金属Cu²⁺污染的程度。

关键词: 光谱; 玉米叶片; 皮尔逊相关系数矩阵; 铜污染植被指数; 叶绿素

中图分类号:TP7 文献标志码:A

A Model on Detecting the Polluted Degree of Maize Leaves by Cu Pollution Vegetation Index

CHENG Feng, YANG Ke-ming^*, CUI Ying, LU Tian-yu, CHEN Li-fan, RONG Kun-peng

State Key Laboratory of Coal Resources and Safe Mining, China University of Mining & Technology (Beijing), Beijing 100083, China

*Corresponding author

Abstract

The use of hyperspectral remote sensing to monitor heavy metal pollution in crops has become an important part of remote sensing research. The difference in the amount of heavy metal content in the contaminated crop leaves mapped to the spectrum is weak, so it is challenging to dig sensitively the value information contained in it. In this paper, based on the spectrum of crop leaves, a pollution detection model of copper pollution vegetation index (CPVI) was proposed by combining multiple spectral feature bands to characterize the pollution degree of heavy metal copper on crops. Firstly, a pot experiment was conducted to add CuSO₄·5H₂O powder with different concentration gradients to the soil to simulate copper-contaminated soil environment and stress corn growth. The spectra of old, middle and new leaves at the ear of corn were collected, and the Cu²⁺ content and relative chlorophyll concentration in the leaves were determined. Then, using 58 randomly selected corn leaf spectra as experimental data, the spectral reflectances of the two groups of wavelengths λ₁ and λ₂ were selected in the wavelength range of 380~900 nm. The Pearson correlation coefficient between CPVI [ λ₁, λ₂] and Cu²⁺ content in the corresponding leaves was calculated, and the absolute value matrix of correlation characteristics was obtained. Secondly, according to the obtained correlation feature matrix, the characteristic band of 690 and 465 nm with high correlation coefficient was extracted and combined with the band 850 nm to establish the Copper pollution index of maize (CPVI_m). After that, CPVI_m index was verified based on 26 other groups of data, and Normalized difference vegetation index (NDVI), MERIS terrestrial chlorophyll index (MTCI) and other conventional vegetation indexes were compared to verify the effectiveness and superiority of CPVI_m. The results showed that the highest correlation coefficient between NDVI, MTCI, REP, DVI and Cu²⁺ content in leaves was 0.68, and the lowest residual sum of squares was 70.99. However, CPVI_m was significantly negatively correlated with Cu²⁺ content in leaves. The correlation coefficient was -0.80, and the residual sum of squares was 48.52, which were better than conventional indexes such as NDVI and MTCI. It proved that CPVI_m is more sensitive to heavy metal stress. At the same time, the robustness of CPVI_m index was verified by using the spectral data of different varieties of maize in different years. The correlation coefficient of CPVI_m and Cu²⁺ content were -0.90 and -0.96, respectively, which were significantly correlated. It shows that CPVI_m is still suitable for detecting the pollution degree of different maize varieties. In addition, using Cu²⁺ content, CPVI_m and chlorophyll relative concentration in maize leaves, a three-dimensional analysis model was constructed, which reflected the correlation between them from a spatial point of view. The CPVI detection model based on the combination of spectral characteristic bands can be used as a reference method to evaluate the pollution degree of heavy metals in crops. The CPVI_m index based on this method can effectively identify the degree of heavy metal Cu²⁺ pollution in maize.

Keyword: Spectrum; Maize leaves; Pearson correlation coefficient matrix; Copper pollution vegetation index; Chlorophyll

文章图片

引言

近年来, 随着矿山开采、工业废水排放及污水灌溉、农药化肥不合理使用等问题的出现, 大量重金属进入到土壤环境中。由于重金属污染具有波及范围广、隐蔽性强和难降解等特点, 已成为国际环境污染治理的研究热点^{[1, 2, 3]}。据2014年发布的“ 全国土壤污染调查报告” , 全国土壤污染物总超标率达16%。铜、镉、汞等重金属污染点位超标率分别达2.1%, 7.0%和1.6%。重金属不仅会引起土壤生态环境的改变, 还会在农作物中累积, 并且会影响植物叶片叶绿素含量变化、阻碍光合产物的传输、抑制植物根系生长, 进而影响到农作物产品质量及农田生态系统的稳定, 最终严重威胁到人类生命健康^{[4, 5]}, 因此对于农田生态系统土壤及农作物重金属污染监测尤为重要。传统的重金属污染监测方法存在成本高、过程繁琐, 不适于大范围应用等缺陷, 而高光谱遥感具有信息量大、可进行快速无损、自动且大范围监测地表等优势, 已在农业生态环境监测等方面得到广泛应用。

在利用高光谱遥感监测植被生化特性参量等方面, 常以具备光学特性的植被指数或经模型算法处理的光谱作为媒介来反映植物的生化参量及生长状态。国内外已有众多专家学者通过常规植被指数以及算法模型等方式来反映受重金属污染植物的生理状态。 Hede^[6]等通过绿波短波红外植被指数(vegetation index considering greenness and shortwave infrared, VIGS)探究了植被受重金属Cu, Pb, Zn和Cd污染的影响程度, 并与传统的归一化植被指数(normalized difference vegetation index, NDVI)进行了比较, 证明植被指数VIGS对重金属胁迫的敏感程度更强。 Rathod^[7]等指出水分胁迫指数等与叶片砷含量有显著相关性, 可鉴别重金属胁迫下植物生长状态, 并利用回归模型显示了反射光谱数据在监测植物健康方面的能力。刘美玲^[8]等运用小波变换分析了受金属锌胁迫水稻光谱的奇异性特征。李蜜^[9]等建立了用于预测水稻农田重金属污染水平的模糊神经网络模型, 表征重金属胁迫下水稻叶片叶绿素含量存在微小变化, 进而确定农田重金属污染状态。原有的植被指数建立的目的是反演植被的生化参量但对监测重金属污染方面缺乏针对性, 已有的模型算法虽具有针对性但较为复杂, 实际应用存在一定局限性。且除金属含量外, 植物的供水量、有机质及养分等相互重叠的参数也会影响植被的光谱反射率^[10]。基于此, 在保持各参数一致的情况下, 针对玉米叶片光谱的变化情况, 利用多个光谱特征波段组合的方式, 提出一种铜污染植被指数的污染程度探测模型CPVI, 在此基础上建立探测模型与叶片Cu²⁺含量的相关系数绝对值矩阵, 构建针对玉米铜污染的植被指数(copper pollution vegetation index of maize, CPVI_m), 最终获得CPVI_m与叶片Cu²⁺含量的变化规律, 并与NDVI, MTCI, DVI和REP等光谱特征参量进行对比分析, 验证其在监测重金属污染方面的优越性, 并进一步探究了CPVI_m与叶片叶绿素含量之间的关系。

1 实验部分

1.1 胁迫实验

植株培育及污染梯度设置。选取“ 密糯8号” 玉米作为研究对象, 种植在有底漏花盆中。根据《GB15618— 2018: 土壤环境质量农用地土壤污染风险管控标准(试行)》中土壤污染标准设置铜胁迫梯度为0(空白对照实验), 50, 100, 150, 200, 300, 400, 600, 800和1 000 μ g· g^-1, 同一梯度设置三组平行实验, 分别标注为Cu(ck)-1, Cu(ck)-2, Cu(ck)-3, Cu(50)-1, Cu(50)-2, Cu(50)-3, …, Cu(1 000)-1, Cu(1 000)-2, Cu(1 000)-3, 共计30组。对栽培土壤进行过筛处理, 以逐层喷洒、翻土均匀混合的方式将不同浓度的CuSO₄· 5H₂O粉末添加入相应胁迫梯度的土壤中, 并向花盆喷撒适量水, 模拟Cu污染土壤环境。对玉米种子进行催芽处理, 待发芽后将其栽种于含有不同浓度梯度的CuSO₄· 5H₂O的土壤中。在玉米生长过程中添加适量NH₄NO₃, KNO₃和KH₂PO₄营养液, 并定期给玉米浇水、保持通风。

1.2 数据采集

(1)玉米叶片光谱数据采集。使用波长范围为350~2 500 nm SVC HR-1024I高性能地物光谱仪进行光谱数据采集, 50 W的卤素灯光源和25° 视场角探头, 所采集的光谱反射系数经专用平面白板进行标准化处理。采集玉米穗期的叶片光谱, 选取位于植株上、中、下部位具备代表性的新(N)、中(M)、老(O)三类叶片的叶中位置进行光谱测量。在培育植株过程中, 胁迫梯度为Cu(1 000)-2的玉米在幼苗期叶片枯黄, 故无法用于后续实验分析。对玉米N, M和O叶片的反射光谱分别平行测取3组, 取均值作为最终结果, 最终获取87组光谱数据。选取第一组平行样的新叶叶片光谱展示, 如图1所示。

	Figure Option View Download New Window
	图1 不同Cu²⁺胁迫梯度下玉米叶片新叶光谱Fig.1 Spectra of new maize leaf stressed by different Cu²⁺ concentrations

(2)叶片叶绿素相对浓度测定。使用SPAD-502叶绿素测定仪对各类叶片进行叶绿素测定, SPAD-502所获值代表叶片中的叶绿素浓度相对值。对每组叶片测取3次, 取均值为不同叶片叶绿素浓度相对值。本工作直接以SPAD-502叶绿素测定仪读数代表叶绿素含量。

(3)叶片Cu²⁺含量测定。各项实验完成后, 对玉米叶片进行冲洗、烘干、粉碎、封装保存、标注、编号处理。在相同规范条件下, 经微波处理后, 采用电感耦合等离子发射光谱仪(ICP-OES)测定玉米叶片样品中的Cu²⁺含量, 各胁迫梯度N, M, O叶片Cu²⁺含量结果如图2所示。其中胁迫梯度为Cu(150)-1O, Cu(600)-3O, Cu(800)-2O叶片(图2中用圆圈标注)Cu²⁺含量较其他两组平行样差异较大, 属异常数据, 故将其剔除。

	Figure Option View Download New Window
	图2 土壤Cu²⁺胁迫梯度及叶片中Cu²⁺含量关系Fig.2 Relationship between Cu²⁺ stress concentration in soil and Cu²⁺content in leaves

随机选取两组平行样中的N, M, O叶片0, 50, 100, 150, 200, 300, 400, 600, 800和1 000 μ g· g^-1胁迫梯度共58组数据作为实验数据用于CPVI探测模型构建, 以剩余平行样N, M, O叶片相同胁迫梯度共26组数据作为检验数据进行验证。

2 理论与方法

2.1 CPVI污染探测模型建立

重金属吸收等胁迫因素会影响叶片的反射光谱, 因此可利用植被叶片光谱反射率作为媒介来反映植物的生长状况。许多研究侧重利用两个或多个特征波段经线性或非线性组合的方式, 构成对某个生理及生化参量比较敏感的指数, 以达到监测植被生长规律特征的目的。本文将两个宽波段反射率修改为波段间隔为1 nm分辨率的窄波段反射率, 并在近红外波段中选择R₈₅₀作为基础波段光谱反射率, $R_{λ_{1}}$ 与 $R_{λ_{2}}$ 作为变量, 建立针对重金属铜污染的植被指数探测模型

$CPVI [λ_{1}, λ_{2}] = \frac{R_{850} - R_{λ_{1}}}{R_{850} - R_{λ_{2}}}$ (1)

式中, λ 为波长, 且λ ₁≠ λ ₂≠ 850 nm, R_λ为波长对应的反射率。

利用随机选取的两组平行样共计58组光谱数据及对应叶片Cu²⁺含量作为实验数据, λ ₁, λ ₂光谱波长范围均取380~900 nm。通过λ ₁和λ ₂分别组合, 计算对应探测模型CPVI[λ ₁, λ ₂], 进行遍历计算后将所获探测模型结果与对应梯度的叶片Cu²⁺含量进行相关性分析, 求取二者之间的相关系数, 进而得到皮尔逊相关系数绝对值矩阵, 如图3所示。在此, 横轴代表波长λ ₁, 纵轴代表波长λ ₂, 色带|ρ |代表相关系数的绝对值, 即由探测模型CPVI[λ ₁, λ ₂]与叶片Cu²⁺含量之间以λ ₁和λ ₂作为变量的相关性函数所得, 点的颜色代表相关性程度的强弱。

	Figure Option View Download New Window
	图3 CPVI[λ ₁, λ ₂]与叶片Cu²⁺含量相关性图Fig.3 Correlation map between CPVI [λ ₁, λ ₂] and leaf Cu²⁺ content

2.2 特征波段选择及植被指数构建

根据皮尔逊相关系数绝对值矩阵, 通过波长λ ₁和λ ₂相互组合、分析比较, 挑选与玉米叶片Cu²⁺含量相关性相对较高的波段λ ₁=690 nm, λ ₂=465 nm(图3中红色区域), 构建针对农作物玉米的铜污染植被指数CPVI[690, 465](简称CPVI_m)。在此并未选择相关性最高的λ ₁=690 nm, λ ₂=530 nm波段作为探测模型的特征波段, 根据CPVI[λ ₁, λ ₂]与叶片Cu²⁺含量相关性图, 当λ ₂取值为530 nm时, 相对应λ ₁取值范围较小, 顾及各胁迫梯度原始光谱具有差异性, 会产生误差。因此选择λ ₂=465nm, 波长λ ₁相对取值范围较大的波段作为铜污染植被指数的特征波段, 以保证模型的有效性。

3 结果与讨论

3.1 CPVI_m优越性验证

对于构建的植被指数CPVI_m, 利用剩余26组光谱、叶片Cu²⁺含量数据进行验证。计算各胁迫梯度叶片光谱对应的CPVI_m及常规植被指数NDVI, MTCI, DVI, REP^{[12, 13, 14]}, 并分别与叶片中Cu²⁺含量进行相关性分析, 得到对应的相关系数r和残差平方和(the residual sum of squares, RSS)。 RSS是衡量模型拟合程度的指标, 其值越小, 证明拟合程度越好。分析结果如图4和图5所示, CPVI_m与重金属含量呈显著负相关, 样本个数n=26, 相关系数r=-0.80, RSS=48.52, NDVI, MTCI, DVI和REP指数与重金属含量的相关系数r的绝对值均小于0.8, 且RSS均大于48.52。由此说明CPVI_m较其他指标(即NDVI, MTCI, DVI和REP)对重金属胁迫更敏感, 并且可更准确地反映植物受重金属的胁迫程度。常规植被指数定义及计算方法如表1所示。

	Figure Option View Download New Window
	图4 植被指数CPVI_m与叶片Cu²⁺含量关系Fig.4 Relationship between CPVI_m and Cu²⁺content in leaves

	Figure Option View Download New Window
	图5 NDVI (a), MTCI (b), DVI (c), REP (d)与叶片Cu²⁺含量关系Fig.5 Relationship between NDVI (a), MTCI (b), DVI (c), REP (d) and Cu²⁺ content in leaves

表1 植被指数 Table 1 Vegetation index

Index	Name	Formula
NDVI	Normalized difference vegetation index	$\frac{R_{800} - R_{670}}{R_{800} + R_{670}}$
MTCI	MERIS terrestrial chlorophyll index	$\frac{R_{754} - R_{709}}{R_{709} - R_{681}}$
DVI	Difference vegetation index	R₈₀₀-R₆₇₀
REP	Red edge position	700+ $\frac{40 (R_{780} + 1.5 R_{673} - R_{700})}{R_{740} - R_{700}}$

表1 植被指数 Table 1 Vegetation index

3.2 CPVI_m鲁棒性验证

为了检验CPVI_m的鲁棒性, 在2014年和2016年采集的“ 中糯1号” 玉米叶片光谱数据中随机选择梯度分别为Cu(ck), Cu(250), Cu(500)及Cu(ck), Cu(200), Cu(300), Cu(500), Cu(600), Cu(800)和Cu(1 000)对指数CPVI_m进行验证。以2014年数据为基础的结果如图6所示, 不同污染水平下CPVI_m与叶片Cu²⁺含量显著负相关, 相关系数r=-0.90, 以2016年数据为基础的结果如图7所示, CPVI_m与Cu²⁺含量的相关系数r= -0.96。由此说明CPVI_m植被指数对于不同品种的玉米污染程度探测均具备一定可靠性。

	Figure Option View Download New Window
	图6 植被指数CPVI_m与叶片Cu²⁺含量关系Fig.6 Relationship between CPVI_m and Cu²⁺ content in leaves

	Figure Option View Download New Window
	图7 植被指数CPVI_m与叶片Cu²⁺含量关系Fig.7 Relationship between CPVI_m and Cu²⁺content in leaves

3.3 叶片Cu含量、叶绿素含量及CPVI_m相关性分析

叶绿素在光合作用中起着至关重要的作用, 且作物重金属含量与叶片中的叶绿素相对浓度密切相关^[11]。因此, 可通过掌握叶绿素与叶片Cu²⁺含量的关系规律, 间接利用叶绿素含量反映植被受重金属污染的程度。因实验所测玉米N, M和O叶片叶绿素含量变化范围不同, 所以将实测数据(包括实验数据与验证数据)共计84组, 分新叶、中叶、老叶分别进行讨论。为避免偶然因素影响, 体现数据的普遍性, 对三组平行样中N, M和O叶片叶绿素含量、 Cu²⁺含量及CPVI_m值取平均值作为该梯度的代表值。叶片叶绿素含量、 Cu²⁺含量及CPVI_m两两结合, 分析任意两者之间相关性, 为了综合得到更好的评估效果, 将这3个指标作为三维参量, 构建玉米叶片污染程度三维分析模型, 如图8所示。三维分析模型中, 可直观观察到土壤中铜胁迫程度范围在0~100, 150~400和600~1 000 μ g· g^-1分别集中在模型的顶部, 中部和底端, 即随着胁迫程度的增强散点随着体对角线呈现逐步下移的趋势, 更直观清晰地呈现了不同土壤铜胁迫的程度, 可作为评估分析模型。其次, 根据分析模型可直观体现叶片叶绿素含量与指数CPVI_m呈正相关, 叶绿素含量与叶片Cu²⁺含量呈负相关, CPVI_m与叶片Cu²⁺含量呈负相关。由此说明植被指数CPVI_m不仅可作为玉米叶片铜污染程度的指标, 也可作为探究叶片叶绿素含量的指标。

	Figure Option View Download New Window
	图8 玉米叶片Cu²⁺、叶绿素含量、 CPVI_m三维分析模型 (a): 新叶; (b): 中叶; (c): 老叶Fig.8 Three-dimensional analysis model of Cu²⁺ and chlorophyll content and CPVI_m in maize leaves (a): New leaves; (b): Middle leaves; (c): Old leaves

4 结论

提出一种以光谱波段组合遍历计算为基础的CPVI污染程度探测方法, 分析其与叶片Cu²⁺含量的相关性, 并构建了针对玉米叶片铜污染的CPVI_m指数, 得到以下结论: (1)CPVI_m可表征玉米受重金属铜污染的程度, 即随着叶片中Cu²⁺含量的增加, CPVI_m呈下降趋势, 相关系数为r=-0.80, 残差平方和RSS=48.52。通过与常规植被指数NDVI, MTCI, DVI和REP对比, 验证了CPVI_m的优越性。同时通过另外两组实验数据验证了指数CPVI_m的鲁棒性, CPVI_m与叶片Cu²⁺含量相关系数分别达到了-0.90和-0.96。 (2)基于新、中、老叶分别探究了叶片叶绿素含量与叶片重金属含量及CPVI_m的相关性, 并构建了三维分析模型, 证明了叶片Cu²⁺含量和CPVI_m与SPAD-502读数中的叶片叶绿素含量相关, 即叶绿素含量与CPVI_m呈正相关, 叶绿素含量与叶片中Cu²⁺含量呈负相关。

由此表明通过光谱特征波段组合方式构建的CPVI探测模型可作为评价农作物重金属污染程度的参考方法; 基于该方法构建的CPVI_m指数可有效甄别玉米受重金属Cu²⁺污染程度, 也可作为反映玉米叶片尺度叶绿素含量的指标。

参考文献

文献列表

[1]	ZHU Ye-qing, QU Yong-hua, LIU Su-hong, et al(朱叶青, 屈永华, 刘素红, 等). Journal of Remote Sensing(遥感学报), 2014, 18(2): 335. [本文引用:1]
[2]	YU Min-da, ZHANG Hui, HE Xiao-song, et al(虞敏达, 张慧, 何小松, 等). Chinese Journal of Environmental Engineering(环境工程学报) , 2016, 10(3): 1500. [本文引用:1]
[3]	LIU Sha-sha, FU Jian-ping, CAI Xin-de, et al(刘沙沙, 付建平, 蔡信德, 等). Ecology and Environmental Sciences(生态环境学报), 2018, 27(6): 1173. [本文引用:1]
[4]	SONG Wei, CHEN Bai-ming, LIU Lin(宋伟, 陈百明, 刘琳). Research of Soil & Water Conservation(水土保持研究), 2013, 20(2): 293. [本文引用:1]
[5]	CHEN Wei-ping, YANG Yang, XIE Tian, et al(陈卫平, 杨阳, 谢天, 等). Acta Pedologica Sinica(土壤学报), 2018, 55(2): 261. [本文引用:1]
[6]	Hede A N H, Kashiwaya K, Koike K, et al. Remote Sensing of Environment, 2015, 171: 83. [本文引用:1]
[7]	Rathod P H, Brackhage C, Meer F D V D, et al. European Journal of Remote Sensing, 2015, 48(3): 283. [本文引用:1]
[8]	LIU Mei-ling, LIU Xiang-nan, LI Ting, et al(刘美玲, 刘湘南, 李婷, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2010, 26(3): 191. [本文引用:1]
[9]	LI Mi, LIU Xiang-nan, LIU Mei-ling(李蜜, 刘湘南, 刘美玲). Acta Scientiae Circumstantiae(环境科学学报), 2010, 30(10): 2108. [本文引用:1]
[10]	Gotze C, Glasser C, Jung A. International Journal of River Basin Management, 2016, 14(4): 499. [本文引用:1]
[11]	Schlemmer M, Gitelson A, Schepers J, et al. International Journal of Applied Earth Observations and Geoinformation. 2013, 25(1): 47. [本文引用:1]
[12]	Jay S, Maupas F, Bendoula R, et al. Field Crops Research, 2017, 210: 33. [本文引用:1]
[13]	LIU Yan, NIE Lei, YANG Yun(刘艳, 聂磊, 杨耘). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2018, 34(9): 182. [本文引用:1]
[14]	Danson F M, Plummer S E. Remote Sensing, 1995, 16(1): 183. [本文引用:1]

2014

0.0

ZHU

Ye-qing

, QU

Yong-hua

, LIU

Su-hong

, et al(朱叶青, 屈永华, 刘素红, 等). Journal of Remote Sensing(遥感学报), 2014, 18(2): 335.

The leaf reflectance spectra of plants polluted with heavy metal copper exhibit abnormal behavior. This paper investigates the spectral changes in copper-polluted leaves using seven spectral signatures and the spectral angle method. Wheat and lettuce were selected as the experimental subjects and were cultured in copper-polluted soil. Leaf spectra reflectance and scanning electron micrograph of ten copper treatment groups were measured in four regular periods. The reflectance spectra of the leaves were obtained using an ASD Field Spectrometer. Results showed that the irregular spectral changes of copper-polluted leaves depend on crop growth stages and crop types. The spectral angle, a simple method of comparing the vector angle of the leaf spectra with the given threshold values, has been justified as an efficient method of describing the spectral differences between copper-polluted leaves and healthy leaves. The method is very sensitive to light and the severity of copper pollution. Results also show that the leaf structural parameter ( N ) of copper-polluted leaves is larger than that of healthy leaves. The linear relationship between N and the reflectance value of the red shoulder was fitted with a correlation coefficient of 0.978. As a result, the deduced N from the leaf reflectance spectra may be used as an indicator to reflect the leaf structure distortion under copper pollution conditions. The significance of this study lies in its provision of data asbasis and theoretical support for the construction of a copper-polluted leaf reflectance spectral model.

重金属铜污染植被的反射光谱特性会发生明显改变。在本研究中，采用不同程度的铜污染土壤作为培养基质，选择春小麦、上海青两种农作物进行铜胁迫实验，获取了4个不同生育期、10个不同铜污染强度下的植被叶片的反射光谱，并采用铜污染叶片7个特征波段和光谱角的方法研究了铜污染叶片的光谱特征。结果表明，铜污染叶片光谱差异与作物时期和作物类型有关，可以采用叶片光谱角描述铜污染叶片与健康叶片的光谱差异。该方法只需与阈值做简单的比较，方法简便易行，而且对轻度及重度铜污染十分敏感。叶片光谱辐射传输模型反演结果表明铜污染叶片内部结构参数N明显变大，这也证明了铜污染使叶片内部结构更加散乱无序。在此基础上进一步建立了N与红肩处反射率值的线性关系，相关系数为0.978。本文为铜污染叶片光谱反射模型的建立提供了初步的数据基础与理论支持。

... 由于重金属污染具有波及范围广、隐蔽性强和难降解等特点, 已成为国际环境污染治理的研究热点^[1,2,3] ...

2016

0.0

... 由于重金属污染具有波及范围广、隐蔽性强和难降解等特点, 已成为国际环境污染治理的研究热点^[1,2,3] ...

2018

0.0

... 由于重金属污染具有波及范围广、隐蔽性强和难降解等特点, 已成为国际环境污染治理的研究热点^[1,2,3] ...

2013

0.0

... 重金属不仅会引起土壤生态环境的改变, 还会在农作物中累积, 并且会影响植物叶片叶绿素含量变化、阻碍光合产物的传输、抑制植物根系生长, 进而影响到农作物产品质量及农田生态系统的稳定, 最终严重威胁到人类生命健康^[4,5], 因此对于农田生态系统土壤及农作物重金属污染监测尤为重要 ...

2018

0.0

2015

0.0

... Hede^[6]等通过绿波短波红外植被指数(vegetation index considering greenness and shortwave infrared, VIGS)探究了植被受重金属Cu, Pb, Zn和Cd污染的影响程度, 并与传统的归一化植被指数(normalized difference vegetation index, NDVI)进行了比较, 证明植被指数VIGS对重金属胁迫的敏感程度更强 ...

2015

0.0

... Rathod^[7]等指出水分胁迫指数等与叶片砷含量有显著相关性, 可鉴别重金属胁迫下植物生长状态, 并利用回归模型显示了反射光谱数据在监测植物健康方面的能力 ...

2010

0.0

LIU

Mei-ling

, LIU

Xiang-nan

, LI

Ting

, et al(刘美玲, 刘湘南, 李婷, 等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2010, 26(3): 191.

The aim of this study is to detect the stress levels of rice under Zn pollution by analyzing the characteristics of hyperspectral singularity. Hyperspectral reflectance and heavy metal concentration from different growth stages of rice in two consecutive years were collected in the field experiment. Wavelet transform was performed to hyperspectral reflectance (350-1 300 nm) by using “Daubechies 5” wavelet function, and three categories of singularity parameters as indicators of hyperspectral singularity of rice under Zn stress were explored, including singularity range (SR), singularity amplitude (SA) and singularity exponent (SE). The results showed that the wavelet coefficient of the fifth decomposition level of “Daubechies 5” wavelet function proved successful for accurately identifying the hyperspectral singularity of rice under Zn stress. SR of rice was concentrated on the region around 480-850 nm of spectral signal under Zn stress. Maximum value of SA occurred at the tillering stage, and the SA increased sharply from seeding stage to tillering stage and then decreased from jointing stage to anthesis stage and to mature stage, while the SE increased constantly from seeding stage to mature stage when rice was growing. Zn concentration in rice leaves had better linearity with the SA and SE with correlation coefficient (R2) of 0.8445 and root mean square error (RMSE) of 5.60. The hyperspectral singularity applying wavelet transform technique has been shown to indentify and quantitatively analyze the stress levels of rice under Zn pollution effectively and provide important reference for detecting other metal-induced stress on crop.

通过研究锌污染胁迫下水稻的光谱奇异性特征来诊断水稻锌污染水平。根据试验区水稻各个生长期的ASD实测高光谱数据和同步获取的作物与农田土壤重金属含量2 a田间试验数据，利用Daubechies小波系中的“Db5”小波函数对锌污染胁迫水稻的350～1 300 nm光谱波段进行5层小波分解，并利用奇异范围、奇异幅度和奇异指数等参数对水稻光谱奇异性进行定量计算与分析。结果表明，“Db5”小波函数分解的第5层小波系数能精准探测水稻光谱的奇异性，其奇异范围集中在480～850 nm波段范围；奇异幅度从出苗期到分蘖期迅速增加，并达到最大值，在随后的分蘖期→拔节期→开花期→成熟期逐渐减少；奇异指数从出苗期→分蘖期→拔节期→开花期→成熟期逐渐增加；水稻叶片锌含量与奇异幅度及奇异指数之间存在较强的相关关系，其决定系数（R2）为0.8445，均方根误差（RMSE）为5.60。光谱奇异性能有效地诊断并定量分析水稻锌污染胁迫状况，从而为作物重金属污染胁迫监测提供重要参考。

... 刘美玲^[8]等运用小波变换分析了受金属锌胁迫水稻光谱的奇异性特征 ...

2010

0.0

, LIU

Xiang-nan

, LIU

Mei-ling

(李蜜, 刘湘南, 刘美玲). Acta Scientiae Circumstantiae(环境科学学报), 2010, 30(10): 2108.

Data for the spectral reflectance of rice,chlorophyll content,leaf and soil heavy metal content were collected from three experimental rice fields with different heavy metal pollution levels in Changchun city,Jilin province,China. Based on analysis of the effect of heavy metals on chlorophyll in rice,spectral indices which are sensitive to subtle changes of the chlorophyll content were selected as the input parameters of a model using multiple stepwise regression,and the chlorophyll concentration was used as the output parameter to characterize the stress levels of heavy metal pollution. Finally,a fuzzy neural network model was established to predict the heavy metal pollution levels of rice fields. The predicted pollution levels corresponded well to the measured pollution levels. The correlation coefficient ( R 2 ) between measured and predicted chlorophyll concentration was 0.985. Therefore the fuzzy neural network model can precisely simulate the subtle changes of rice chlorophyll content under heavy metal pollution and estimate the heavy metal pollution levels in rice field.

以吉林省长春一汽厂附近3块不同重金属污染状况的水稻实验样地为例,通过地面实测获取了水稻的光谱反射率、叶片叶绿素含量、叶片及土壤重金属含量等数据.同时,在分析重金属对水稻叶片叶绿素含量影响的基础上,通过多元逐步回归分析选出对水稻叶片叶绿素含量微小变化指示灵敏的光谱参数作为模型输入层,并将水稻叶片叶绿素含量值作为输出层来表征农田重金属污染胁迫水平,最终建立了用于预测水稻农田重金属污染水平的模糊神经网络模型.结果表明,该模糊神经网络模型预测的水稻重金属污染胁迫水平与实测结果吻合度较高,预测的叶绿素含量值与实测值的拟合度较好( R 2 =0.985).表明在受重金属污染胁迫的情况下,水稻叶片叶绿素含量微小而复杂的变化可以通过构建模糊神经网络模型很好地模拟出来,从而确定出农田的重金属污染水平.

... 李蜜^[9]等建立了用于预测水稻农田重金属污染水平的模糊神经网络模型, 表征重金属胁迫下水稻叶片叶绿素含量存在微小变化, 进而确定农田重金属污染状态 ...

2016

0.0

... 且除金属含量外, 植物的供水量、有机质及养分等相互重叠的参数也会影响植被的光谱反射率^[10] ...

2013

0.0

... 3 叶片Cu含量、叶绿素含量及CPVI_m相关性分析叶绿素在光合作用中起着至关重要的作用, 且作物重金属含量与叶片中的叶绿素相对浓度密切相关^[11] ...

2017

0.0

... 计算各胁迫梯度叶片光谱对应的CPVI_m及常规植被指数NDVI, MTCI, DVI, REP^[12,13,14], 并分别与叶片中Cu²⁺含量进行相关性分析, 得到对应的相关系数r和残差平方和(the residual sum of squares, RSS) ...

2018

0.0

LIU

Yan

, NIE

Lei

, YANG

Yun

(刘艳, 聂磊, 杨耘). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2018, 34(9): 182.

针对天山牧区草地面积广、生长环境差异大等特点，选用MODIS/MOD13Q1数据，生成归一化植被指数、增强植被指数、土壤调节植被指数和差值植被指数4种从不同角度反映牧草长势的植被指数，对4个不同牧草利用类型分区分别建立了4种指数及其组合与总产草量的4类回归方程。利用留一交叉检验法评价各模型精度，最终获得适合不同草地利用类型的产草量遥感估算模型。结果表明：4种植被指数都可用于产草量估算，不同植被指数估算模型的拟合精度有区域差异性。当采用SAVI和二次多项式拟合时，RMSE最大值出现在I区，为5 857 kg/hm2，当采用NDVI和二次多项式拟合时，最小值出现在III区，仅为616.487 kg/hm2。值得注意的是采用差值植被指数、土壤调节植被指数估算产草量需考虑植被覆盖状况。其次，多个植被指数组合有信息互补的优势，采用线性回归模型估产时，多个植被指数组合精度高于单一植被指数。该研究可为天山地区草地总产草量遥感估算提供有意义的借鉴。

1995

0.0