南海近岸水域藻类叶绿素a浓度遥感反演模型
|
摘要
在海洋近岸水环境中,人类活动常常导致水质下降,生物资源减少,赤潮现象频繁发生。叶绿素a是藻类细胞内的主要光合色素,叶绿素a浓度及其动态变化反映了水体中藻类植物的丰度、生物量及其变化规律,是反映海洋水体营养状况的一个客观生物学指标,在海洋富营养状态评价中也是最为重要的指标。通过卫星遥感数据对研究区建立反演模型得到叶绿素a浓度值,这对于研究整个海域的藻类分布,进而研究海洋生态系统中的初级生产力具有重要意义,已经成为赤潮监测的有效方法。
本文以NSFC-GD联合基金项目为依托,以南海近岸水域中的藻类为研究对象,通过利用多源遥感图像对藻类叶绿素a进行分析,采用BP神经网络技术,研究海洋藻类叶绿素a浓度的遥感反演方法。基于野外实测数据、雷达影像和光学遥感影像数据,建立海洋藻类叶绿素a浓度的遥感反演模型,为监测海洋生态环境提供参考。
本文采用BP神经网络技术,结合光学和雷达遥感建立了海洋藻类叶绿素a浓度的遥感反演模型。研究工作通过以下几个步骤实现:(1)基于TM影像对藻类叶绿素及其光谱特征进行分析,选取了TM影像的特征光谱波段:TM1、TM2、TM3、TM4;(2)从雷达遥感的角度分析了雷达参数中的后向散射系数与藻类叶绿素浓度之间的潜在相关关系,并提取了雷达特征参数—HH、VV极化下的后向散射系数;(3)通过对不同输入层节点数和隐含层节点数的分析,最终确定了适合本研究的BP神经网络模型,即以上六个参数作为模型的输入参数,隐含层节点数为6,藻类叶绿素a浓度值作为模型的输出参数;(4)构建了不同输入参数组合的线性回归模型,通过实测值与预测值的比较,分析精度,证实BP神经网络的优越性。
研究结果表明:(1)当光学遥感和雷达遥感结合使用时,预测精度要比单独使用TM或SAR时高很多;(2)遥感波谱特征与叶绿素a浓度之间属于复杂的非线性问题,统计模型显得明显不足,而BP神经网络模型能取得比较好的反演效果,尤其在二类水体的叶绿素a浓度反演上具有明显的优势。
Human activities often lead to deterioration of water quality , reduction of biological resources and frequent occurrence of red tide phenomenon in the marine coastal water environment. As chlorophyll-a is the main photosynthetic pigment cells of algae,dynamics of chlorophyll-a concentration reflects the abundance of algae in water,biomass and its variation. It is an objective biological indicator to reflect the nutritional status of ocean water and also the most important indicator to evaluate the eutrophication in the ocean. To built a model for retrieving the concentration of chlorophyll-a through satellite remote sensing data in study area is important to study the distribution of algae and primary productivity of marine ecosystems,and it has become an effective way to monitor the red tide.
Relied on the NSFC-GD jointly fund project,the algae under the South China Sea coastal waters as research subjects,combined different remote sensing images to analysis the spectral of chlorophyll-a,BP neural network was used to study the remote sensing way for extracting the chlorophyll-a concentration of marine algae. Based on field observation data,radar image and satellite remote sensing data,we had established a spectral retrieval model of marine algae chlorophyll-a concentration,providing references for monitoring the marine environment.
In this study,BP neural network was used to extract remote sensing information of chlorophyll-a. Research through the following steps to achieve:(1)analyze the spectral characteristics of algae based on TM image,and then find out the spectral parameters of TM image:TM1、TM2、TM 3、TM4;(2)analyze the potential relationship between the backscattering coefficient in radar parameters and the concentration of chlorophyll-a,and then extract the feature parameters of radar:the backscatter coefficient under HH and VV polarization;(3)Through analyze the different input layer nodes and hidden nodes,the structure of BP neural network model in this study was finally fixed,that is,above six parameters as the input parameters,the hidden layer nodes is 6,and chlorophyll-a concentration as the model output parameters;(4)Constructing different input parameters of linear regression model,comparing the accuracy between the measured and predicted values,the advantages of BP neural network has been confirmed.
The results showed that:(1)when combined the optical remote sensing and radar remote sensing,the prediction accuracy was much higher than TM or SAR used alone;(2)the relationship between remote sensing spectral characteristics and chlorophyll-a was more complex nonlinear problem,so the statistical model seemed clearly insufficient,but the BP neural network model can achieve better retrieval results , especially had obvious advantages in the inversion of chlorophyll-a concentration in II water.
本文以NSFC-GD联合基金项目为依托,以南海近岸水域中的藻类为研究对象,通过利用多源遥感图像对藻类叶绿素a进行分析,采用BP神经网络技术,研究海洋藻类叶绿素a浓度的遥感反演方法。基于野外实测数据、雷达影像和光学遥感影像数据,建立海洋藻类叶绿素a浓度的遥感反演模型,为监测海洋生态环境提供参考。
本文采用BP神经网络技术,结合光学和雷达遥感建立了海洋藻类叶绿素a浓度的遥感反演模型。研究工作通过以下几个步骤实现:(1)基于TM影像对藻类叶绿素及其光谱特征进行分析,选取了TM影像的特征光谱波段:TM1、TM2、TM3、TM4;(2)从雷达遥感的角度分析了雷达参数中的后向散射系数与藻类叶绿素浓度之间的潜在相关关系,并提取了雷达特征参数—HH、VV极化下的后向散射系数;(3)通过对不同输入层节点数和隐含层节点数的分析,最终确定了适合本研究的BP神经网络模型,即以上六个参数作为模型的输入参数,隐含层节点数为6,藻类叶绿素a浓度值作为模型的输出参数;(4)构建了不同输入参数组合的线性回归模型,通过实测值与预测值的比较,分析精度,证实BP神经网络的优越性。
研究结果表明:(1)当光学遥感和雷达遥感结合使用时,预测精度要比单独使用TM或SAR时高很多;(2)遥感波谱特征与叶绿素a浓度之间属于复杂的非线性问题,统计模型显得明显不足,而BP神经网络模型能取得比较好的反演效果,尤其在二类水体的叶绿素a浓度反演上具有明显的优势。
Human activities often lead to deterioration of water quality , reduction of biological resources and frequent occurrence of red tide phenomenon in the marine coastal water environment. As chlorophyll-a is the main photosynthetic pigment cells of algae,dynamics of chlorophyll-a concentration reflects the abundance of algae in water,biomass and its variation. It is an objective biological indicator to reflect the nutritional status of ocean water and also the most important indicator to evaluate the eutrophication in the ocean. To built a model for retrieving the concentration of chlorophyll-a through satellite remote sensing data in study area is important to study the distribution of algae and primary productivity of marine ecosystems,and it has become an effective way to monitor the red tide.
Relied on the NSFC-GD jointly fund project,the algae under the South China Sea coastal waters as research subjects,combined different remote sensing images to analysis the spectral of chlorophyll-a,BP neural network was used to study the remote sensing way for extracting the chlorophyll-a concentration of marine algae. Based on field observation data,radar image and satellite remote sensing data,we had established a spectral retrieval model of marine algae chlorophyll-a concentration,providing references for monitoring the marine environment.
In this study,BP neural network was used to extract remote sensing information of chlorophyll-a. Research through the following steps to achieve:(1)analyze the spectral characteristics of algae based on TM image,and then find out the spectral parameters of TM image:TM1、TM2、TM 3、TM4;(2)analyze the potential relationship between the backscattering coefficient in radar parameters and the concentration of chlorophyll-a,and then extract the feature parameters of radar:the backscatter coefficient under HH and VV polarization;(3)Through analyze the different input layer nodes and hidden nodes,the structure of BP neural network model in this study was finally fixed,that is,above six parameters as the input parameters,the hidden layer nodes is 6,and chlorophyll-a concentration as the model output parameters;(4)Constructing different input parameters of linear regression model,comparing the accuracy between the measured and predicted values,the advantages of BP neural network has been confirmed.
The results showed that:(1)when combined the optical remote sensing and radar remote sensing,the prediction accuracy was much higher than TM or SAR used alone;(2)the relationship between remote sensing spectral characteristics and chlorophyll-a was more complex nonlinear problem,so the statistical model seemed clearly insufficient,but the BP neural network model can achieve better retrieval results , especially had obvious advantages in the inversion of chlorophyll-a concentration in II water.
引文
[1] Koponen S,Pulliainen J,Servomaa H,et al.Analysis on the feasibility of multisource remote sensing observations for Chl-a monitoring in Finnish lakes.Science of the Total Environment ,ITC,The Netherlands (CD ROM).2001,268:95-106
[2]李宝华,荒川久辛.南黄海叶绿素a与初级生产力之间的相关分析.黄渤海海洋.1998,16(2):48-53
[3]疏小舟,尹球,匡定波.内陆水体藻类叶绿素浓度与反射光谱特征的关系.遥感学报,2000年2月,4(1):41-45
[4] Liu Mei-ling,Liu Xiang-nan,Li Mi.Neural-network model for estimating leaf chlorophyll concentration in rice under stress from heavy metals using four spectral indices,biosystems engineering.2010,6:223-233
[5]曹仕,刘湘南,曹珊.成熟期水稻砷污染胁迫光谱诊断空间模型研究.中国生态农业学报.2010,18(4):234-238
[6] Wei Yu-chun,Huang Jia-zhu,Li Yun-mei.The Hyperspectral data monitoring model of chlorophyll of summer in Taihu Lake, China.Journal Of Remote Sensing,Volume 11,Issue 5,September 2007:756-762
[7]周伟华,霍文毅,袁翔城等.东海赤潮高发区春季叶绿素a和初级生产力的分布特征.应用生态学报.2003,14(7):1055-1059
[8]朱明远,毛兴华,吕瑞华.黄海海区的叶绿素a和初级生产力.黄渤海海洋.1993,11(3):38-51
[9] Pulliainen J,Kallio K,Eloheimo K.A semi-operative approach to lake water quality retrieval from remote sensing data . The Science of the Total Environment.2001,268:79-93
[10] Cracknell.Remote sensing techniques in estuaries and coastal zones—an update.Int J of Remote Sensing.1999,19(3):485-496
[11]汪小钦,陈崇成.遥感在近岸海洋环境监测中的应用.海洋环境科学.2000年11月,19(4):72-76
[12]李素菊,吴倩,王学军等.巢湖浮游植物叶绿素含量与反射光谱特征的关系.湖泊科学.2002年9月,14(3):228-234
[13]赵碧云,贺彬,朱云燕等.滇池水体中叶绿素a含量的遥感定量模型.云南环境科学.2001年9月,20(3):1-3
[14]韩秀珍,吴朝阳,郑伟等.基于水面实测光谱的太湖蓝藻卫星遥感研究.应用气象学报.2010年12月,21(6):724-731
[15] Fraser R N.Hyperspectral remote sensing of turbidity and chlorophyll-a among Nebraska Sand Hills lakes.Int J Remote Sensing.1998,19(8):1579-1589
[16] Rundquitst D C,Han L,Schalles J F,et al.Remote measurement of algal chlorophyll in surface waters: the case for the first derivative of reflectance near 690vnm.Eng Remote Sens.1996,62:195-200
[17]挥才兴.海岸带及近海卫星遥感综合应用技术.北京.海洋出版社.2005:35~40
[18]刘良明.卫星海洋遥感导论.武汉.武汉大学出版社.2005:120-142,145-153,216-220
[19]崔廷伟.渤海生物光学性质与水色遥感反演[博士学位论文].青岛.中国海洋大学,2006
[20]陈清莲,王项南.海洋水色遥感及Lnadsat-5 TM数据在海南岛东部海域水色分析中的应用.海洋技术.1995,14(3):47-60
[21]疏小舟,尹球,匡定波.内陆水体藻类叶绿素浓度与反射光谱特征的关系.遥感学报.2000年,4(1):41-45
[22] Keiner L,Yan Xiao-hai.The use of a neural network in estimating surface chlorophyll and sediments from Thematic Mapper imagery.Remote Sensing of Environment.1998,66(2):153-165
[23]唐军武,田国良.水色光谱分析与多成分反演算法.遥感学报.1997,1(4):252-256
[24]詹海刚,施平,陈楚群.利用神经网络反演海水叶绿素浓度.科学通报.2000,45(17):1879-1884
[25]曹文熙,杨跃忠.海洋光合有效辐射分布的计算模式.热带海洋学报.2002,21(3):47-54
[26]赵英时等.遥感应用分析原理与方法.北京.科学出版社.2003
[27] YuanZhi Zhang,Jouni Pulliainen,Sampsa Koponen,et al.Applicaton of an empirical neural net work to surface water quality estimation in the Gulf of Finaland using combined optical data and microwave data.Remote Sensing of Environment.2002,81:327-336
[28] Pulliainen J,Kallio K,Eloheimo K,et al.A semi-operational approach to water quality retrieval from remote sensing data.Science of the Total Environment,2001,268:79-93
[29]金铭,刘湘南.基于冠层多维光谱的水稻镉污染胁迫诊断模型研究.中国环境科学.2011,31(1):733-737
[30] Xiangnan Liu,Nan Jiang.Application of Fuzzy Reasoning to Assessment of Crop Stress Level Based on MODIS Data:A Focus on Heavy Metal Pollution.The International Archives of the Photogrammetry.Remote Sensing and Spatial Information Science.2008:347—352
[31]任敬萍,赵进平.二类水体水色遥感的主要进展与发展前景.地球科学进展,2002,17(3):363-371
[32]徐希孺.遥感物理.北京大学出版社.北京.2005:231-233
[33]孙德勇,李云梅,乐成峰等.应用水表面下辐照度比估测太湖夏季水体叶绿素a浓度.湖泊科学.2007,19(6):744-752
[34]唐军武,丁静,王其茂.大气散射对采用归一化植被指数进行赤潮遥感监测的影响研究.海洋学报.2004,26(3):136-142
[35]唐军武.海洋光学特性模拟和遥感模型.中国科学院遥感应用研究所.北京.1999
[36]疏小舟,汪骏发,沈鸣明.航空成像光谱水质遥感研究.红外和毫米波.2000,19(4):273-276
[37]刘强,柳钦火,肖青.机载多角度遥感图像的几何校正方法研究.中国科学D辑.2002,32(4):299-306
[38]毛志华,黄海清,朱乾坤.我国海区SaewiFS资料大气校正.海洋与湖沼.2001,32(6):581-587
[39]杨建洪,王锦,赵冬至.海洋水色遥感大气校正算法研究进展.海洋环境科学.2008年2月,27(1):97-100
[40]张亭禄,贺明霞.基于人工神经网络的一类水域叶绿素a浓度反演方法.遥感学报.2002,6(1):40-44
[41]巩彩兰,樊伟.海洋水色卫星遥感二类水体反演算法的国际研究进展.海洋通报.2002.21(2):77-83
[42]任敬萍,赵进平.二类水体水色遥感的主要进展和发展前景.地球科学进展.2002.17(3):363-371
[43] Zhang Xu-qin,Zhang Shi-kui,Wu Yongsen,et al.Progress inresearch on yellow-substance in seawater.Journal of Ocean-ography of Huanghai & Bohai Seas.2000,18(1):89-92
[44] Ohde T,Siegel H.Correction of bottom influence in ocean colour satellite images of shallow water areas of the Baltic Sea.International Journal of Remote Sensing.2001,22:297-313
[45] Lee Z P,Carder K L,Mobley C D,et al.Hyperspectral remote sensing for shallow waters : 2 . Deriving bottom depths and water properties by optimization .Applied Optics.1999,38:3831-3843
[46]唐军武,王晓梅,宋庆君等.黄、东海二类水体水色要素的统计反演模式.海洋科学进展.2004,22:1-7
[47]中国科学院空间科学技术中心.中国地球资源光谱信息资料汇编.北京.能源出版社.1987
[48]吕恒,江南,罗潋葱.基于TM数据的太湖叶绿素A浓度定量反演.地理科学.2006,26(4):472-476
[49] Han L,Donald C,Rundquitst D C.Comparion of NIR/RED ratio and first derivative of reflectance in estimating chlorophyll concentration:a case study in a turbid reservoir.Remote Sensing of Environment,1997,62:253-261
[50] Dekker A G,Peters S W.The use of the Thematic Mapper for analysis of entropic lakes:a case study in the Netherlands.Int J Remote Sensing,1993,14(5):799-821
[51]朱敏慧.SAR的海洋遥感探测技术综述.现代雷达.2010年2月,32(2):1-5
[52]黎夏,叶嘉安,王树功等.红树林湿地植被生物量的雷达遥感估算.遥感学报.2006年5月,10(3):387-396
[53]朴鸿泽.雷达遥感的发展及应用.价值工程.2010年:202
[54]郭华东等.雷达对地观测与应用.北京.科学出版社.2000
[55]胡著智,王慧麟,陈钦峦.遥感技术与地学应用.南京.南京大学出版社.1997
[56]舒宁.微波遥感原理.湖北.武汉大学出版社.2003,1-5
[57] Quegan S.Recent Advances in Understanding SAR.Advances in Environmental Remote Sensing.1995,5:89-104
[58]舒宁.微波遥感原理.湖北.武汉大学出版社.2003,1-5
[59]廖静娟,王庆.利用Radarsat-2极化雷达数据探测湿地地表特征与分类.国土资源遥感.2009年9月15日.3:70-73
[60]高帅,牛峥,刘晨洲.基于RADARSAT SAR估测热带人工林叶面积指数研究.国土资源遥感.2008年12月15日.4:35-57
[61]廖静娟,邵芸.多参数SAR数据森林应用潜力分析.遥感学报.2004年.4(增刊):129-134
[62] Giorgio F,Riccardo L.Synthetic Aperture Radar processing.Boca Raton.CRC Press.1999
[63]张雪虎,David McLaughlin,Elisabeth M T warog.机载真实孔径雷达在中小尺度区域海洋遥感中的应用.遥感技术与应用.2005年2月,20(1):127-132
[64]梅安新,彭望琭,秦其明等.遥感导论.高等教育出版社.2001:98-102
[65]吕恒,李新国,江南.基于反射光谱和模拟MERIS数据的太湖悬浮物遥感定量模型.湖泊科学.2005,17(2): 104-109
[66]丛爽.面向MATLAB工具箱的神经网络理论与应用.中国科学技术大学出版社.合肥.1998
[67]杜培军.Radarsat图像滤波的研究.中国矿业大学学报.2002,31(2):132-134
[68]李蜜,刘湘南,刘美玲.基于模糊神经网络的水稻农田重金属污染水平高光谱预测模型.环境科学学报.2011,30(10):2108-2115
[69] Wang Shujun,Guan Dongsheng.Remote Sensing Method of Forest Biomass Estimation by Artificial Neural Network Models . Ecology and Environment.2007,16(1):108-111
[70] Buekton MongainE.The use of neural networks for the estimation of oceaniee onstituents based on MERIS instruct.In Remote Sense,1999,20(9) :1841-1851
[71] Keiner LE , Yan XH.Aneuaral network model for estimating sea surface chlorophyll land sediment from The matie Maper imager.Remote Sensing Environment.1998,66:153-165
[72] Aure′lio Azevedo Barreto-Neto.Application of Fuzzy Logic to the Evaluation of Runoff in a Tropical Watershed.Environmental Modeling & Software 23.2007:244–253
[73] Doerffer R,Schiller H.Determination of Casez water consituents using radiative transfer simulation and its inversion by neural networks.Office of Naval Research.Washington DC.1998
[74] MongainE.The use of neural net works for the estimation of oeean iee onstituentsbasedonMERIS instruct.In Remote Sense,1999,20(9):1841-1851
[75]飞思科技产品研发中心.神经网络理论与MATLAB7实现.北京.电子工业出版社.2005
[76]施英妮.基于人工神经网络技术的高光谱遥感浅海水深反演研究.青岛.中国海洋大学.2005
[77]赵冬至,曲元,张丰收等.用TM图象估算海表面叶绿素浓度的神经网络模型.海洋环境科学.2001.20(l):16-23
[78]郑小慎,林培根.基于TM数据渤海湾叶绿素浓度反演算法研究.天津科技大学学报.2010年12月,25(6):51-53
[79]李旭文,季耿善,杨静.太湖梅梁湖湾蓝藻生物量遥感估算.国土资源遥感.1995,(2):23-28
[80]杨一鹏,王桥,肖青等.基于TM数据的太湖叶绿素a浓度定量遥感反演方法研究.地理与地理信息科学.2006,22(2):5-8
[81]吕恒,江南,罗潋葱.基于TM数据的太湖叶绿素A浓度定量反演.地理科学,2006,26(4):472-476
[2]李宝华,荒川久辛.南黄海叶绿素a与初级生产力之间的相关分析.黄渤海海洋.1998,16(2):48-53
[3]疏小舟,尹球,匡定波.内陆水体藻类叶绿素浓度与反射光谱特征的关系.遥感学报,2000年2月,4(1):41-45
[4] Liu Mei-ling,Liu Xiang-nan,Li Mi.Neural-network model for estimating leaf chlorophyll concentration in rice under stress from heavy metals using four spectral indices,biosystems engineering.2010,6:223-233
[5]曹仕,刘湘南,曹珊.成熟期水稻砷污染胁迫光谱诊断空间模型研究.中国生态农业学报.2010,18(4):234-238
[6] Wei Yu-chun,Huang Jia-zhu,Li Yun-mei.The Hyperspectral data monitoring model of chlorophyll of summer in Taihu Lake, China.Journal Of Remote Sensing,Volume 11,Issue 5,September 2007:756-762
[7]周伟华,霍文毅,袁翔城等.东海赤潮高发区春季叶绿素a和初级生产力的分布特征.应用生态学报.2003,14(7):1055-1059
[8]朱明远,毛兴华,吕瑞华.黄海海区的叶绿素a和初级生产力.黄渤海海洋.1993,11(3):38-51
[9] Pulliainen J,Kallio K,Eloheimo K.A semi-operative approach to lake water quality retrieval from remote sensing data . The Science of the Total Environment.2001,268:79-93
[10] Cracknell.Remote sensing techniques in estuaries and coastal zones—an update.Int J of Remote Sensing.1999,19(3):485-496
[11]汪小钦,陈崇成.遥感在近岸海洋环境监测中的应用.海洋环境科学.2000年11月,19(4):72-76
[12]李素菊,吴倩,王学军等.巢湖浮游植物叶绿素含量与反射光谱特征的关系.湖泊科学.2002年9月,14(3):228-234
[13]赵碧云,贺彬,朱云燕等.滇池水体中叶绿素a含量的遥感定量模型.云南环境科学.2001年9月,20(3):1-3
[14]韩秀珍,吴朝阳,郑伟等.基于水面实测光谱的太湖蓝藻卫星遥感研究.应用气象学报.2010年12月,21(6):724-731
[15] Fraser R N.Hyperspectral remote sensing of turbidity and chlorophyll-a among Nebraska Sand Hills lakes.Int J Remote Sensing.1998,19(8):1579-1589
[16] Rundquitst D C,Han L,Schalles J F,et al.Remote measurement of algal chlorophyll in surface waters: the case for the first derivative of reflectance near 690vnm.Eng Remote Sens.1996,62:195-200
[17]挥才兴.海岸带及近海卫星遥感综合应用技术.北京.海洋出版社.2005:35~40
[18]刘良明.卫星海洋遥感导论.武汉.武汉大学出版社.2005:120-142,145-153,216-220
[19]崔廷伟.渤海生物光学性质与水色遥感反演[博士学位论文].青岛.中国海洋大学,2006
[20]陈清莲,王项南.海洋水色遥感及Lnadsat-5 TM数据在海南岛东部海域水色分析中的应用.海洋技术.1995,14(3):47-60
[21]疏小舟,尹球,匡定波.内陆水体藻类叶绿素浓度与反射光谱特征的关系.遥感学报.2000年,4(1):41-45
[22] Keiner L,Yan Xiao-hai.The use of a neural network in estimating surface chlorophyll and sediments from Thematic Mapper imagery.Remote Sensing of Environment.1998,66(2):153-165
[23]唐军武,田国良.水色光谱分析与多成分反演算法.遥感学报.1997,1(4):252-256
[24]詹海刚,施平,陈楚群.利用神经网络反演海水叶绿素浓度.科学通报.2000,45(17):1879-1884
[25]曹文熙,杨跃忠.海洋光合有效辐射分布的计算模式.热带海洋学报.2002,21(3):47-54
[26]赵英时等.遥感应用分析原理与方法.北京.科学出版社.2003
[27] YuanZhi Zhang,Jouni Pulliainen,Sampsa Koponen,et al.Applicaton of an empirical neural net work to surface water quality estimation in the Gulf of Finaland using combined optical data and microwave data.Remote Sensing of Environment.2002,81:327-336
[28] Pulliainen J,Kallio K,Eloheimo K,et al.A semi-operational approach to water quality retrieval from remote sensing data.Science of the Total Environment,2001,268:79-93
[29]金铭,刘湘南.基于冠层多维光谱的水稻镉污染胁迫诊断模型研究.中国环境科学.2011,31(1):733-737
[30] Xiangnan Liu,Nan Jiang.Application of Fuzzy Reasoning to Assessment of Crop Stress Level Based on MODIS Data:A Focus on Heavy Metal Pollution.The International Archives of the Photogrammetry.Remote Sensing and Spatial Information Science.2008:347—352
[31]任敬萍,赵进平.二类水体水色遥感的主要进展与发展前景.地球科学进展,2002,17(3):363-371
[32]徐希孺.遥感物理.北京大学出版社.北京.2005:231-233
[33]孙德勇,李云梅,乐成峰等.应用水表面下辐照度比估测太湖夏季水体叶绿素a浓度.湖泊科学.2007,19(6):744-752
[34]唐军武,丁静,王其茂.大气散射对采用归一化植被指数进行赤潮遥感监测的影响研究.海洋学报.2004,26(3):136-142
[35]唐军武.海洋光学特性模拟和遥感模型.中国科学院遥感应用研究所.北京.1999
[36]疏小舟,汪骏发,沈鸣明.航空成像光谱水质遥感研究.红外和毫米波.2000,19(4):273-276
[37]刘强,柳钦火,肖青.机载多角度遥感图像的几何校正方法研究.中国科学D辑.2002,32(4):299-306
[38]毛志华,黄海清,朱乾坤.我国海区SaewiFS资料大气校正.海洋与湖沼.2001,32(6):581-587
[39]杨建洪,王锦,赵冬至.海洋水色遥感大气校正算法研究进展.海洋环境科学.2008年2月,27(1):97-100
[40]张亭禄,贺明霞.基于人工神经网络的一类水域叶绿素a浓度反演方法.遥感学报.2002,6(1):40-44
[41]巩彩兰,樊伟.海洋水色卫星遥感二类水体反演算法的国际研究进展.海洋通报.2002.21(2):77-83
[42]任敬萍,赵进平.二类水体水色遥感的主要进展和发展前景.地球科学进展.2002.17(3):363-371
[43] Zhang Xu-qin,Zhang Shi-kui,Wu Yongsen,et al.Progress inresearch on yellow-substance in seawater.Journal of Ocean-ography of Huanghai & Bohai Seas.2000,18(1):89-92
[44] Ohde T,Siegel H.Correction of bottom influence in ocean colour satellite images of shallow water areas of the Baltic Sea.International Journal of Remote Sensing.2001,22:297-313
[45] Lee Z P,Carder K L,Mobley C D,et al.Hyperspectral remote sensing for shallow waters : 2 . Deriving bottom depths and water properties by optimization .Applied Optics.1999,38:3831-3843
[46]唐军武,王晓梅,宋庆君等.黄、东海二类水体水色要素的统计反演模式.海洋科学进展.2004,22:1-7
[47]中国科学院空间科学技术中心.中国地球资源光谱信息资料汇编.北京.能源出版社.1987
[48]吕恒,江南,罗潋葱.基于TM数据的太湖叶绿素A浓度定量反演.地理科学.2006,26(4):472-476
[49] Han L,Donald C,Rundquitst D C.Comparion of NIR/RED ratio and first derivative of reflectance in estimating chlorophyll concentration:a case study in a turbid reservoir.Remote Sensing of Environment,1997,62:253-261
[50] Dekker A G,Peters S W.The use of the Thematic Mapper for analysis of entropic lakes:a case study in the Netherlands.Int J Remote Sensing,1993,14(5):799-821
[51]朱敏慧.SAR的海洋遥感探测技术综述.现代雷达.2010年2月,32(2):1-5
[52]黎夏,叶嘉安,王树功等.红树林湿地植被生物量的雷达遥感估算.遥感学报.2006年5月,10(3):387-396
[53]朴鸿泽.雷达遥感的发展及应用.价值工程.2010年:202
[54]郭华东等.雷达对地观测与应用.北京.科学出版社.2000
[55]胡著智,王慧麟,陈钦峦.遥感技术与地学应用.南京.南京大学出版社.1997
[56]舒宁.微波遥感原理.湖北.武汉大学出版社.2003,1-5
[57] Quegan S.Recent Advances in Understanding SAR.Advances in Environmental Remote Sensing.1995,5:89-104
[58]舒宁.微波遥感原理.湖北.武汉大学出版社.2003,1-5
[59]廖静娟,王庆.利用Radarsat-2极化雷达数据探测湿地地表特征与分类.国土资源遥感.2009年9月15日.3:70-73
[60]高帅,牛峥,刘晨洲.基于RADARSAT SAR估测热带人工林叶面积指数研究.国土资源遥感.2008年12月15日.4:35-57
[61]廖静娟,邵芸.多参数SAR数据森林应用潜力分析.遥感学报.2004年.4(增刊):129-134
[62] Giorgio F,Riccardo L.Synthetic Aperture Radar processing.Boca Raton.CRC Press.1999
[63]张雪虎,David McLaughlin,Elisabeth M T warog.机载真实孔径雷达在中小尺度区域海洋遥感中的应用.遥感技术与应用.2005年2月,20(1):127-132
[64]梅安新,彭望琭,秦其明等.遥感导论.高等教育出版社.2001:98-102
[65]吕恒,李新国,江南.基于反射光谱和模拟MERIS数据的太湖悬浮物遥感定量模型.湖泊科学.2005,17(2): 104-109
[66]丛爽.面向MATLAB工具箱的神经网络理论与应用.中国科学技术大学出版社.合肥.1998
[67]杜培军.Radarsat图像滤波的研究.中国矿业大学学报.2002,31(2):132-134
[68]李蜜,刘湘南,刘美玲.基于模糊神经网络的水稻农田重金属污染水平高光谱预测模型.环境科学学报.2011,30(10):2108-2115
[69] Wang Shujun,Guan Dongsheng.Remote Sensing Method of Forest Biomass Estimation by Artificial Neural Network Models . Ecology and Environment.2007,16(1):108-111
[70] Buekton MongainE.The use of neural networks for the estimation of oceaniee onstituents based on MERIS instruct.In Remote Sense,1999,20(9) :1841-1851
[71] Keiner LE , Yan XH.Aneuaral network model for estimating sea surface chlorophyll land sediment from The matie Maper imager.Remote Sensing Environment.1998,66:153-165
[72] Aure′lio Azevedo Barreto-Neto.Application of Fuzzy Logic to the Evaluation of Runoff in a Tropical Watershed.Environmental Modeling & Software 23.2007:244–253
[73] Doerffer R,Schiller H.Determination of Casez water consituents using radiative transfer simulation and its inversion by neural networks.Office of Naval Research.Washington DC.1998
[74] MongainE.The use of neural net works for the estimation of oeean iee onstituentsbasedonMERIS instruct.In Remote Sense,1999,20(9):1841-1851
[75]飞思科技产品研发中心.神经网络理论与MATLAB7实现.北京.电子工业出版社.2005
[76]施英妮.基于人工神经网络技术的高光谱遥感浅海水深反演研究.青岛.中国海洋大学.2005
[77]赵冬至,曲元,张丰收等.用TM图象估算海表面叶绿素浓度的神经网络模型.海洋环境科学.2001.20(l):16-23
[78]郑小慎,林培根.基于TM数据渤海湾叶绿素浓度反演算法研究.天津科技大学学报.2010年12月,25(6):51-53
[79]李旭文,季耿善,杨静.太湖梅梁湖湾蓝藻生物量遥感估算.国土资源遥感.1995,(2):23-28
[80]杨一鹏,王桥,肖青等.基于TM数据的太湖叶绿素a浓度定量遥感反演方法研究.地理与地理信息科学.2006,22(2):5-8
[81]吕恒,江南,罗潋葱.基于TM数据的太湖叶绿素A浓度定量反演.地理科学,2006,26(4):472-476