滞尘对城市植物光谱特征及含水量、色素浓度反演的影响

发布时间：2018-06-15 17:36

本文选题：滞尘或不同滞尘量 + 植物反射光谱　；参考：《上海师范大学》2017年硕士论文

【摘要】：随着我国经济的发展,城市化进程的不断加快,城市粉尘污染日益成为居民关注的空气污染问题。植物不仅可通过覆盖裸地、吸附、降低风速等方式降低空气中粉尘的浓度,并且植物滞留粉尘的能力在一定条件下具有永久性。虽然植物具有滞留粉尘的能力,但粉尘同样会对植物产生影响。这已引起国内外学者的广泛关注,并成为了热门研究课题之一。本文的研究数据主要为2016年4月初到8月底对不同城市功能区内目标树种的冠层及叶片光谱(ASD光谱仪)、相应的滞尘量数据(万分之天平)、冠层和叶片双尺度的含水量(烘烤法)和色素浓度数据(化学溶剂法)等。基于上述数据本文首先分析了滞尘或不同滞尘量对植物反射光谱特征和红边位置的影响;接着分析了滞尘前后不同含水量水平及不同色素浓度等级下的叶片反射光谱特征变化;最后,在上述研究的基础上分析了滞尘对植物含水量反演及色素浓度反演的影响。研究结果表明:(1)可见光波段对滞尘量变化的灵敏度最高,1450-1750nm波段次之,700-1350nm波段的植物反射光谱受不同滞尘量的影响最为稳定——随着滞尘量的增加而段降低。滞尘或不同滞尘量不会引起植被“红边”位置的移动但其位置与植物种类有关;“主峰”值与“次峰”值都会随着滞尘量的增加而减小并且“主峰”位置与“次峰”位置的大小关系同样同植物种类有关。不同滞尘量对叶片反射光谱的影响相对于冠层而言更加稳定。相应的研究结果可为不同城市功能区的植物种类识别或城市不同功能区粉尘污染程度的识别提供光谱支撑。(2)滞尘对不同含水量或不同色素浓度植物光谱的影响主要体现在改变反射光谱反射值和改变相邻反射光谱间的间隔这两方面。对于不同含水量而言,在750-1350nm波段之间,滞尘前后含水量对光谱的影响规律相对稳定而且各含水量水平的反射光谱间的间隔相比较于其他波段都较大,即750-1350nm波段可很好地反映植被的含水量变化。对于不同色素含量而言,350-700nm和750-1350nm波段范围能很好地反映植物色素浓度的变化。(3)对于叶片尺度而言:滞尘不会改变叶片EWT(等效水深度)、FWC(相对含水量)同各指数之间的相关性质,但其会改变两者与各指数之间的相关系数;滞尘会降低叶片EWT同各指数之间的相关系数,而会增高叶片FWC同各指数之间的相关系数;无论在无尘还是在有尘情况下,叶片EWT同各指数之间的相关系数都高于叶片FWC同各指数之间的相关系数;这说明,EWT更能反映植物叶片的含水量情况。对于冠层尺度而言,由于冠层EWT同各指数之间的相关系数远低于FWC同各指数之间的相关系数,因此FWC更能反映植物冠层的含水量情况。(4)不同树种含水量指标与各种光谱指数间相关系数不同并且滞尘对其的影响程度也不同。各树种都存在与之含水量相关系数较高的光谱指数(“最适指数”),但在滞尘前后“最适指数”并非具有完全的一致性。除此之外,“最适指数”会随着树种的变化而变化即某一光谱指数不具备叶片含水量反演的普适性能力。对不同树种同时进行含水量反演,会增大研究的难度。(5)无论对于不同波段或是各指数而言,滞尘都不会改变各种色素同两者之间的相关性质,但都会改变各种色素同各波段或各指数间的相关系数。对于叶绿素a和类胡萝卜素而言滞尘会增高其与各波段之间的相关系数,对于叶绿素b的影响,与叶绿素a和类胡萝卜素正相反。无论是在有尘还是在无尘状态下叶绿素a同各波段之间的相关系数都高于叶绿素b同各波段之间的相关系数,叶绿素同各波段之间的相关系数高于类胡萝卜素同各波段之间的相关系数。虽然滞尘对各色素同各指数间的相关系数的影响与所选的指数有关,但就整体而言(平均变化),滞尘对其影响的规律与滞尘对波段的的影响规律相同。(6)滞尘会对叶片含水量反演和叶绿素浓度反演产生影响。在叶片含水量反演方面,滞尘对叶片含水量反演精度的影响与滞尘量有关。但总体上表现为:滞尘不仅会降低反演回归模型的决定系数,而且会增大估测模型RMSE值和RE值,并降低反演精度。而且不同建模指标对滞尘量的变化的敏感度不同。滞尘对叶绿素浓度反演的影响呈现两种变化情况:其一为增高反演精度;其二为降低反演精度;但增高程度远低于降低程度。滞尘对两者的影响程度都与所选取的建模指标有关。
[Abstract]:With the development of China's economy and the accelerating process of urbanization, urban dust pollution has become an increasingly important problem of air pollution. Plants can not only reduce the concentration of dust in the air by covering bare land, adsorbing and reducing wind speed, but the ability of plants to stay dust is permanent under certain conditions. There is the ability to detained dust, but dust also affects plants. This has aroused wide attention of scholars both at home and abroad, and has become one of the most popular research topics. The data of this paper are mainly the canopy and leaf spectrum (ASD spectrometer) of the target species in different urban functional areas from early April 2016 to the end of August, and the number of the corresponding dust amount. The water content (bake method) and pigment concentration data (chemical solvent method) on the double scale of canopy and leaf (chemical solvent method). Based on the above data, the effects of dust or different dust amount on the spectral characteristics and red edge position of plants were analyzed first, and the different water content levels and different pigment concentrations before and after the dust were analyzed. The effects of dust on the inversion of water content and pigment concentration in plants were analyzed on the basis of the above study. The results showed that: (1) the sensitivity of the visible light band to the change of dust accumulation was the highest, the 1450-1750nm band was the second, and the plant reflection spectrum in the 700-1350nm band was affected by different dust levels. The effect is most stable with the increase of dust retention. Dust or different dust amounts will not cause the movement of the "red edge" position of the vegetation, but its position is related to the plant species; the value of "main peak" and "sub peak" will decrease with the increase of the amount of dust and the relationship between the position of "main peak" and the position of "second peak" The effect of different dust amounts on leaf reflectance spectra is more stable than that of canopy. The corresponding results can provide spectral support for identification of plant species in different urban functional areas or identification of dust pollution in different functional areas of cities. (2) dust on different water content or different pigment concentrations. The effect of the spectrum is mainly reflected in the two aspects of changing the reflection spectrum of the reflection spectrum and changing the interval between adjacent reflectance spectra. For different water content, the influence of water content on the spectrum is relatively stable before and after the 750-1350nm band, and the interval between the reflectance spectra of each water content is compared to the other bands. Larger, that is, the 750-1350nm band can well reflect the changes in the water content of vegetation. For different pigments, the range of 350-700nm and 750-1350nm bands can well reflect the changes in plant pigment concentration. (3) for the leaf scale, the dust will not change the leaf EWT (equal water depth), FWC (relative water content) and the index of each index. Related properties, but it will change the correlation coefficient between the two and the index, which will reduce the correlation coefficient between the EWT and the index, and increase the correlation coefficient between FWC and the index. The correlation coefficient between the leaves EWT and the index is higher than that between the leaves FWC and the index in the dustless or dustless. Correlation coefficient; this shows that EWT can reflect the water content of plant leaves. For the canopy scale, the correlation coefficient between the canopy EWT and the index is far lower than the correlation coefficient between the FWC and the index, so FWC can reflect the water content of the plant canopy. (4) the water content index of different tree species and the various spectral indices The correlation coefficient is different and the influence degree of the dust on it is different. All the trees have a higher coefficient of spectral correlation coefficient ("the best index"), but the "optimum index" before and after the dust is not completely consistent. Numbers do not have the universal ability to invert the water content of leaves. Inversion of water content of different species will increase the difficulty of the study. (5) no matter to different bands or indices, the dust will not change the correlation between the various pigments and the two, but all of them will change the relation between the various pigments and the various bands or the indices. The correlation coefficient between chlorophyll a and carotenoids increased with each band, and the effect on chlorophyll b was opposite to chlorophyll a and carotenoid. The correlation coefficient between chlorophyll a and the various bands in dust and dust free states were higher than the correlation lines between chlorophyll b and each band. The correlation coefficient between the chlorophyll and the various bands is higher than the correlation coefficient between the carotenoids and the various bands. Although the influence of the dust on the correlation coefficient between the pigments and the various indices is related to the selected index, the regularity of the influence of the dust on the dust is the same as the influence of the dust on the band. (6) Dust can affect the inversion of water content and inversion of chlorophyll concentration. In the field of leaf water content inversion, the effect of dust retention on the accuracy of water content inversion is related to the amount of dust. However, the overall performance is that dust retention will not only reduce the decision coefficient of back regression model, but also increase the RMSE value and RE value of the estimation model, and reduce the inversion. The sensitivity of the different modeling indexes to the change of the dust amount is different. The effect of the dust on the inversion of the concentration of chlorophyll presents two changes: one is to increase the inversion accuracy, and the second is to reduce the inversion accuracy, but the degree of increase is far below the degree of reduction.
【学位授予单位】：上海师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：X173;X513

【参考文献】