小麦生物物理与生物化学参数的高光谱遥感监测

发布时间：2018-06-15 12:17

本文选题：小麦 + 冠层　；参考：《南京信息工程大学》2015年硕士论文

【摘要】：由于高光谱遥感可以提供连续的光谱信息,因此可利用这一特点探测植被的生长状况(叶面积指数、冠层光合有效辐射)或者各种生化组分(SPAD、氮、磷含量等),反演各组分含量。从而为农业生产管理提供重要依据,也为农业遥感的推广提供技术支撑。本文以小麦作为研究对象,开展了2年不同品种、不同氮素水平的田间试验。利用数理统计分析等技术手段,研究高光谱遥感(反射率、反照率)参数与小麦生理生化参数之间的相关关系,确立小麦生物物理、生物化学参数的特征光谱参数及相应监测模型,并对模型精度进行了验证,快速、准确、客观的监测小麦长势的变化,以为期提高优质小麦的产量提供理论依据。本研究得出的主要结论如下：(1)随着施氮量的增加,小麦叶面积指数(LAI)逐渐变大,且随生育期呈现先增加后减小的变化特征；对叶面积指数与小麦冠层高光谱反射特征的相关分析发现,LAI与1044nm处一阶导数光谱参数(FD1044)拟合效果最佳,决定系数为0.62。独立年际试验数据的预测精度为0.46。(2)小麦冠层截获光合有效辐射(IPAR)与可见光各波段反射率呈负相关,在近红外波段呈正相关；与小麦冠层IPAR相关性较好的高光谱参数较多,其中与水分指数(WI)拟合效果最好,线性拟合方程决定系数为0.72。(3)小麦地上部生物量随施氮量的增加而增加,不同年度、品种表现出类似规律；在此基础上了讨论了小麦生物量与高光谱参数的关系,提出了监测小麦生物量的光谱参数及其预测模型。其中基于一阶导数光谱参数(FD793)的方程拟合效果最佳。独立年际试验数据的测试表明,FD793能够对小麦生物量进行可靠的监测。(4)分析了不同氮素水平下,小麦整个生育期叶片SPAD值的变化状况,并且讨论了小麦叶片SPAD值与高光谱参数的关系,提出了小麦叶片SPAD值的敏感光谱参数及预测方程。发现叶片SPAD值与红蓝边面积比(Sdr/Sdb)、红边最小值Lo、归一化植被指数(NDVI)等关系密切,以线性方程拟合效果最好。监测模型给出了较好的检验结果,尤其是NDVI。(5)小麦氮、磷含量均随施氮量的增加而增加；红蓝边面积比(Sdr/Sdb)与小麦叶片氮含量相关性好,方程拟合精度高(R2=0.71)；导数光谱参数SD721与小麦叶片磷含量关系密切,方程拟合效果较好(R2=0.42)；利用土壤调节植被指数(SAVl2)可以较好的表达茎秆氮含量营养状况的变化,线性方程的决定系数为0.55。(6)小麦叶片、茎秆碳氮比均随施氮水平的提高而下降；光谱参数P-Depthl690和SD741与小麦叶片碳氮比关系密切,以线性方程拟合效果最好：小麦茎秆碳氮比与光化学反射指数(PRI)和ND672相关性较好,方程拟合精度高。(7)小麦冠层反照率随施氮量的增加而增加,不同生育期变化趋势一致；小麦冠层反照率与LAI、NDVI、EVI(增强型植被指数)均呈显著正相关,其中与LAI和EVI与反照率拟合方程决定系数分别为0.70和0.73,高于NDVI;小麦冠层反照率与冠层透射率呈线性负相关,拟合方程的决定系数为0.41；小麦冠层反照率与小麦叶片SPAD值之间的关系,采用分阶段的拟合方法更佳,两段线性拟合函数均呈现极显著关系；小麦叶片、茎秆氮含量与反照率的相关性好于小麦磷含量,小麦冠层反照率与叶片、茎秆氮含量均呈线性相关,拟合方程的决定系数分别为0.81和0.60；与叶片、茎秆磷含量均呈显性正相关,方程拟合效果较好,决定系数分别为0.52和0.40。本文基于高光谱遥感综合分析了小麦生物物理指标(LAI、光合有效辐射、生物量)和生物化学指标(SPAD、氮、磷含量及碳氮比)与光谱参数的关系,确定了拟合度较好的光谱监测模型,并对模型进行检验和完善,实现了小麦生物物理和生物化学参数的快速、无损监测,为高光谱遥感在农业遥感中的应用提供了技术支撑和理论依据。
[Abstract]:This paper studies the relationship between the parameters of high spectral remote sensing ( LAI ) and the physiological and biochemical parameters of wheat by means of mathematical statistics and so on . The main conclusions are as follows : ( 1 ) With the increase of nitrogen application , LAI gradually increases , and the variation of wheat leaf area index ( LAI ) decreases with the growth period .
Based on the correlation analysis of leaf area index and canopy height spectral reflectance , it is found that the best fitting effect of LAI and the first order derivative spectral parameter ( FD1044 ) at 1044nm is the best , the final coefficient is 0.62 . The prediction accuracy of the interannual test data is 0.46 . ( 2 ) There is a negative correlation between the canopy interception of the canopy and the reflectance of the visible light , and is positively correlated with the near infrared band .
The high spectral parameters of wheat crown layer IPAR were much better , and the best fit effect with water index ( WI ) was the best , the coefficient of linear fitting equation was 0.72 . ( 3 ) The biomass of the ground part of wheat increased with the increase of nitrogen application , and the varieties showed similar laws in different years .
In this paper , the relationship between wheat biomass and high spectral parameters is discussed , and the spectral parameters of wheat biomass and its prediction model are presented . The relationship between SPAD value and high spectral parameters of wheat leaves is discussed . The relationship between SPAD value and high spectral parameters of wheat leaves is discussed .
The ratio of red and blue side area ( Sdr / Sdb ) was correlated with the nitrogen content of wheat leaves , and the equation fitting precision was high ( R2 = 0.71 ) .
The spectral parameters SD721 are closely related to the phosphorus content of wheat leaves , and the equation fitting effect is better ( R2 = 0.42 ) .
By using the soil - adjusted vegetation index ( SAVl2 ) , the nutrient status of stalk nitrogen could be expressed well , and the coefficient of determination of linear equation was 0.55 . ( 6 ) The carbon - to - nitrogen ratio of wheat leaves and stem decreased with the increase of nitrogen application level ;
The spectral parameters P - Depthl690 and SD741 were closely related to the carbon - to - nitrogen ratio of wheat leaves . The best results were as follows : the carbon - to - nitrogen ratio of wheat stem was better than that of photochemical reflection index ( PRI ) and ND672 , and the fitting accuracy of equation was high . ( 7 ) The incidence of wheat canopy was increased with the increase of nitrogen application , and the variation trend of different growth periods was consistent ;
The contrast ratio of LAI , EVI ( enhanced vegetation index ) was positively correlated with LAI and EVI ( enhanced vegetation index ) .
The relationship between the canopy reflection rate and SPAD value of wheat leaves is better than that of wheat leaf SPAD value , and the two stage linear fitting functions have very significant relationship with each other .
The correlation of nitrogen content of wheat leaf and stem was better than that of leaf and stem nitrogen , and the coefficient of the fitting equation was 0.81 and 0.60 , respectively .
Based on the high - spectrum remote sensing , the relationship between LAI , photosynthetic active radiation ( biomass ) and biochemical indexes ( SPAD , nitrogen , phosphorus content and carbon / nitrogen ratio ) and spectral parameters was analyzed .
【学位授予单位】：南京信息工程大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：S512.1;S127

【参考文献】