扎龙湿地丹顶鹤生境遥感分类方法研究
|
摘要
生物多样性是湿地生态系统重要生命特征之一。水禽,一种在生态上依赖于湿地的鸟类,是湿地生物多样性的重要组成部分。为更好保护湿地生物多样性,从保护水禽栖息地生境来保护湿地水禽进而保护湿地生物多样性已成为现今研究热点之一。
黑龙江省扎龙自然保护区是保护丹顶鹤等珍稀水禽及其赖以生存生境的著名国家级自然保护区。繁殖期丹顶鹤巢址生境选择受各种自然因素和人为因素综合影响成为一个复杂的生态过程,而其繁殖地是一个超异质性的复合环境。欲探寻繁殖期丹顶鹤生境选择过程,掌握繁殖地生境空间分布规律建立合理的生境分类系统是一项重要的基础工作。本文选取地形、水体、芦苇植被及外在干扰源作为丹顶鹤生境巢址选择主要影响因子,基于高分辨率的遥感影像对其生境进行分类研究。
本文作为理论性研究正是基于高空间分辨率ALOS影像结合其他数据,在生境这样一个空间尺度,面向繁殖期丹顶鹤生境选择特点,综合运用不同遥感分类方法及数学统计分析方法,深入分析本研究区地形、水体及芦苇植被等生境因子。同时,为更有针对性对整个研究区进行土地利用遥感分类,特选取具有代表性的试验区进行生境遥感分类方法试验。基于试验所得结论,灵活利用各种方法优势,完成扎龙湿地芦苇植被亚分类等级图并最终得出丹顶鹤生境分类图。
Biodiversity is vital for wetland ecosystem standing and developing. Waterfowls, a kind of birds relying on wetland ecologically, are the major component of wetland biodiversity. Therefore the status of wetland habitat will directly affect the waterfowls’surviving. In order to protect wetland biodiversity, protecting their habitat has been one of the hot spot in nowadays research.
Zhalong Natural Reserve in Heilongjiang Province is famous national nature reserve, which protects some rare waterfowls such as red-crowned cranes and their habitat. Zhalong Natural Reserve posseses of the largest, relative original and intact bulrush wetland in China at present. However, with the change of natural environment, high-intensity man-made interference and the change of climate in this region, the living space for red-crowned cranes has been threatened severely by both of the abrupt decreasing of wetland area and the degradation of wetland’s ecological functions in a large scale. For the purpose of protecting red-crowned cranes, it’s necessary to understand and master the spatial characteristics and the changes of the red-crowned cranes’habitat in detail from the view of their living.
The nest-site selection of red-crowned cranes’habitat during breeding period in is affected by various natural and human factors. Based on the habitat characteristics of red-crowned cranes’during breeding period in Zhalong Wetland, high spatial resolution ALOS images and other data, the paper mainly discusses the four habitat factors including terrain, water bodies, vegetation and external interference sources. With using different classifying methods on remote sensing, the classification of red-crowned cranes’habitat during breeding period in Zhalong Wetland has been accomplished eventually. The major works in this paper are as follows:
1. Describing the habitat characteristics of red-crowned cranes’habitat during breeding period in Zhalong Wetland. The four main determined habitat factors are terrain, water bodies, vegetation and external interference sources by considering the actual characteristics in the study area and the project’s conditions.
2. Determining the classification principle of red-crowned cranes’habitat considering the physical characteriscs of Zhalong Wetland and the spectral analysis of ALOS images. Based on this classification principle, the interpretation symbols for the landscape type and the landscape classification system on red-crowned cranes’habitat in the study area were established. With the interpretation symbols and classification system, the landscape map of red-crowned cranes’habitat in 2008 was completed by visual interpretation method.
3. DEM establishment. The elevation lines and elevation points are extracted from relief map with scale 1:10,000 based on MapGIS 6.7. At the same time, lakes’boundary as the control parameters are also extracted from ALOS images for DEM establishment. In order to depict the whole terrain characteristics, DEM grid map, slope and aspect map derived from it were generated by elevation interpolating using irregular triangulation network method.
4. Representative sample area was chose to do experiment using classification methods based on remote sensing. Supervision classification method, object-oriented information extraction method and decision tree based on expert knowledge method were mainly discussed. Through accuracy evaluation based on error matrix, results indicated that both object-oriented information extraction method and decision tree based on expert knowledge method showed their own advantages. So time can be saved and efficiency can be improved combining the two methods.
5. Water bodies extraction. Through analyzing the spectrum characteristics of water bodies with other objects for ALOS image, water bodies could be automatically extracted by using threshold method and decision tree method. By comparing with water bodies by visual interpretation, the accuracy of automatic results could reach 85%.
6. Realizing the vegetation information extraction and bulrush sub-classification. The NDVI grid image has been accomplished by using the modeling function based on ERDAS 9.1. And based on the NDVI grid image, bulrush sub-classification was completed by using statistic analysis method. A qualitative description about the relationship among bulrush, terrain and slope was made. On the theory basis of the habitat characteristics of nest-site selection, combined with various data sources, another qualitative description about the rationality and validity of bulrush sub-classification was done.
7. Based on the conclusion of experiments about remote sensing classification, combined with all kinds of existing remote sensing data, the habitat classification map of red-crowned cranes’habitat during breeding period in Zhalong Wetland was eventually accomplished by using object-oriented information extraction method, decision tree based on expert knowledge method, also including necessary visual interpretation method.
Finally, the shortages in this paper were listed and further improvement were discussed which will be further rearched in future.
黑龙江省扎龙自然保护区是保护丹顶鹤等珍稀水禽及其赖以生存生境的著名国家级自然保护区。繁殖期丹顶鹤巢址生境选择受各种自然因素和人为因素综合影响成为一个复杂的生态过程,而其繁殖地是一个超异质性的复合环境。欲探寻繁殖期丹顶鹤生境选择过程,掌握繁殖地生境空间分布规律建立合理的生境分类系统是一项重要的基础工作。本文选取地形、水体、芦苇植被及外在干扰源作为丹顶鹤生境巢址选择主要影响因子,基于高分辨率的遥感影像对其生境进行分类研究。
本文作为理论性研究正是基于高空间分辨率ALOS影像结合其他数据,在生境这样一个空间尺度,面向繁殖期丹顶鹤生境选择特点,综合运用不同遥感分类方法及数学统计分析方法,深入分析本研究区地形、水体及芦苇植被等生境因子。同时,为更有针对性对整个研究区进行土地利用遥感分类,特选取具有代表性的试验区进行生境遥感分类方法试验。基于试验所得结论,灵活利用各种方法优势,完成扎龙湿地芦苇植被亚分类等级图并最终得出丹顶鹤生境分类图。
Biodiversity is vital for wetland ecosystem standing and developing. Waterfowls, a kind of birds relying on wetland ecologically, are the major component of wetland biodiversity. Therefore the status of wetland habitat will directly affect the waterfowls’surviving. In order to protect wetland biodiversity, protecting their habitat has been one of the hot spot in nowadays research.
Zhalong Natural Reserve in Heilongjiang Province is famous national nature reserve, which protects some rare waterfowls such as red-crowned cranes and their habitat. Zhalong Natural Reserve posseses of the largest, relative original and intact bulrush wetland in China at present. However, with the change of natural environment, high-intensity man-made interference and the change of climate in this region, the living space for red-crowned cranes has been threatened severely by both of the abrupt decreasing of wetland area and the degradation of wetland’s ecological functions in a large scale. For the purpose of protecting red-crowned cranes, it’s necessary to understand and master the spatial characteristics and the changes of the red-crowned cranes’habitat in detail from the view of their living.
The nest-site selection of red-crowned cranes’habitat during breeding period in is affected by various natural and human factors. Based on the habitat characteristics of red-crowned cranes’during breeding period in Zhalong Wetland, high spatial resolution ALOS images and other data, the paper mainly discusses the four habitat factors including terrain, water bodies, vegetation and external interference sources. With using different classifying methods on remote sensing, the classification of red-crowned cranes’habitat during breeding period in Zhalong Wetland has been accomplished eventually. The major works in this paper are as follows:
1. Describing the habitat characteristics of red-crowned cranes’habitat during breeding period in Zhalong Wetland. The four main determined habitat factors are terrain, water bodies, vegetation and external interference sources by considering the actual characteristics in the study area and the project’s conditions.
2. Determining the classification principle of red-crowned cranes’habitat considering the physical characteriscs of Zhalong Wetland and the spectral analysis of ALOS images. Based on this classification principle, the interpretation symbols for the landscape type and the landscape classification system on red-crowned cranes’habitat in the study area were established. With the interpretation symbols and classification system, the landscape map of red-crowned cranes’habitat in 2008 was completed by visual interpretation method.
3. DEM establishment. The elevation lines and elevation points are extracted from relief map with scale 1:10,000 based on MapGIS 6.7. At the same time, lakes’boundary as the control parameters are also extracted from ALOS images for DEM establishment. In order to depict the whole terrain characteristics, DEM grid map, slope and aspect map derived from it were generated by elevation interpolating using irregular triangulation network method.
4. Representative sample area was chose to do experiment using classification methods based on remote sensing. Supervision classification method, object-oriented information extraction method and decision tree based on expert knowledge method were mainly discussed. Through accuracy evaluation based on error matrix, results indicated that both object-oriented information extraction method and decision tree based on expert knowledge method showed their own advantages. So time can be saved and efficiency can be improved combining the two methods.
5. Water bodies extraction. Through analyzing the spectrum characteristics of water bodies with other objects for ALOS image, water bodies could be automatically extracted by using threshold method and decision tree method. By comparing with water bodies by visual interpretation, the accuracy of automatic results could reach 85%.
6. Realizing the vegetation information extraction and bulrush sub-classification. The NDVI grid image has been accomplished by using the modeling function based on ERDAS 9.1. And based on the NDVI grid image, bulrush sub-classification was completed by using statistic analysis method. A qualitative description about the relationship among bulrush, terrain and slope was made. On the theory basis of the habitat characteristics of nest-site selection, combined with various data sources, another qualitative description about the rationality and validity of bulrush sub-classification was done.
7. Based on the conclusion of experiments about remote sensing classification, combined with all kinds of existing remote sensing data, the habitat classification map of red-crowned cranes’habitat during breeding period in Zhalong Wetland was eventually accomplished by using object-oriented information extraction method, decision tree based on expert knowledge method, also including necessary visual interpretation method.
Finally, the shortages in this paper were listed and further improvement were discussed which will be further rearched in future.
引文
[1]殷康前,倪晋仁.湿地研究综述[J].生态学报,1998,18(5):539~546.
[2]赵魁义,等.湿地生物多样性保护[M].中国林业出版社,2008-9-1.
[3]刘红玉,杨青,等.湿地景观变化对水禽生境影响研究进展[J].湿地科学,2003,1 (2):115~121.
[4]盛连喜,何春光,等.中国水禽的保护生物学研究进展[J].湿地科学,2003,1(1):26 ~32.
[5]余新晓,牛健植,等.景观生态学[M].北京:高等教育出版社,2006.
[6]逢世良.扎龙自然保护区丹顶鹤的航空调查报告[C].高师理科学刊,2000,20(2) :59~60.
[7]吴庆明.基于“3S”技术的扎龙湿地恢复初期丹顶鹤繁殖期生境选择研究[D].东北林业大学,2005.
[8]马逸清,李晓民.我国丹顶鹤的数量分布.国际鹤类学术讨论论文摘要[C].1987, 12.
[9]张艳红,邓伟,等.向海自然保护区丹顶鹤生境结构空间特征[J].生态学报,2006, 26(11):3725~3731.
[10]张柏.遥感技术在中国湿地研究中的应用[J].遥感技术与应用,1996(3):67~71.
[11]王伟,陆健健.三垟湿地生态系统服务功能及其价值[J].生态学报,2005,25.
[12]李天宏,赵智杰,等.深圳河河口红树林变化的多时相遥感分析[J].遥感学报, 2002,6(5):364~370.
[13]Intermountain Aquatica INC. Environmental conserving and habitat restoration [Z]. Http//www.Intermountain aquatics.com.2002.
[14]李枫,杨红军,等.扎龙湿地丹顶鹤巢址选择研究[J].东北林业大学学报,1999,27 (6):57~60.
[15]孔博,张树清,等.遥感和GIS技术的水禽栖息地适宜性评价中的应用[J].遥感学报,2008,12(6):1001~1009.
[16]Dissanska M. Object-based classification of very high resolution panchromatic images for evaluating recent change in the structure of patterned peatlands[J]. CANADIAN JOURNAL OF REMOTE SENSING,2009,32(2):189~215.
[17]王建强,吴连喜,张岩岩.基于3S技术湿地遥感信息分类的研究[J].水利科技与经济,2006,12(10):718~720.
[18]汪闽,骆建承,周成虎,等.结合高斯马尔可夫随机场纹理模型与支撑向量机在高分辨率图像上提取道路网[J].遥感学报,2005,9(3):271~276.
[19]Haack,B. Monitoring wetland changes with remote sensing: an east Africa example[J].Environmental Management.1996,20:411~419.
[20]Lunetta,R.S.,M.E.Balogh. Application of Multi-Temporal Landsat-5 TM Imagery for Wetland Identification[J].Photogrammetric Engineering and Remote Sensing, 1999,65(11):1303~1310.
[21]Andrew ME, Habitat suitability modelling of an invasive plant with advanced remote sensing data[J]. DIVERSITY AND DISTRIBUTIONS, 2009,14(4): 627~640.
[22]Jin-kang, FENG Xue-zhi, ect. The methods of extracting water information from SPOT image[J]. CHINESE GEOGRAPHICAL SCIENCE,2002,12(1):68~72.
[23] Bridget S H,Waiters D,Powell C,Shrubsele D. Wetland management:An analysis of past practice and recent policy changes in Ontario[J].Journal of Environmental Management.In Press,Available online 20 March. 2006.
[24]骆建承,王钦敏,马江洪,等.遥感图像最大似然分类分类方法的EM改进算法[J].测绘学报,2002,31(3):234~239.
[25]Kindscher K.,Fraser A.,Jakubauskas M.E.Debinski D.M..Identifying wetland meadows in Grand Teton National Park using remote sensing and average wetalnd values.Wetlands Ecology and Management,1998,5:265~273.
[26]Bronge L B, Naslund-Landenmark B. Wetland classification for Swedish CORINE Land Cover adopting a semi-automatic interactive approach. Canadian Jounral of Remote Sensing,2002,28(2):139~155.
[27]Rebecca L. Phillips, Ofer Beeri, Edward Shawn DeKeyser. Remote wetland assessment for Missouri coteau prairie glacial basins .Weltands,2005,25(2):335~ 349.
[28]Corey Baker, Rick Lawrence, MAPPING WETLANDS AND RIPARIAN AREAS USING LANDSAT ETM+ IMAGERY AND DECISION-TREE-BASE- D MODELS[J]. The Society of Wetland Scientist. 2006,26(2):465~474.
[29]雍国玮,石承仓,等.高原若尔盖草地沙化及湿地萎缩动态遥感监测[J].山地学报,2003,21(6):758~762.
[30]张树清.三江平原湿地遥感分类模式研究[J].遥感技术与应用,1999,14(1):54~ 58.
[31]衣伟宏,杨柳,等.基于ETM+影像的扎龙湿地遥感分类研究[J].湿地科学,2004, 2(3):208~212.
[32]牛明香.湿地资源遥感动态监测及其对生态环境的影响研究[D].山东农业大学硕士毕业论文,2004.
[33]张杰,沈芳,刘志国.长江口潮滩湿地植被光谱分析与遥感检测[J].华东师范大学学报(自然科学版),2007,7(4):42~63.
[34]陈定贵,周德民,等.三江平原洪河自然保护区湿地遥感分类研究[J].遥感技术与应用,2007,22(4):485~491.
[35]田庆久,等.江苏近海岸水深遥感研究[J].遥感学报,2007,11(3):373~379.
[36]张寿选,段洪涛,等.基于水体透明度反演的太湖水生植被遥感信息提取[J],湖泊科学,2008,20(2):184~190.
[37]刘丽,等.图像纹理特征提取方法探讨综述[J].中国图像图形学报,2009,64[4]:622~734.
[38]张树清,等. 3S支持下的中国典型沼泽湿地景观时空动态变化研究[M].中国长春:吉林大学出版社,2008,8.
[39]李长友.扎龙湿地生态系统保护对策研究[R].黑龙江省科学技术委员会软科学课题研究报告,黑龙江扎龙国家级自然保护区管理局课题组, 2002.
[40]李长友.走进扎龙湿地[Z].欧盟小项目便捷基金项目资助,黑龙江扎龙国家级自然保护区管理局,.2005.
[41]汪青雄.扎龙湿地丹顶鹤(Grus japonensis)栖息地利用,东北林业大学,2008.
[42]李文发,赵英杰,李金华.兴凯湖自然保护区的丹顶鹤数量调查初报[J].黑龙江八一农垦大学学报,1984,(2):87- 88.
[43]马逸清等.三江平原丹顶鹤的数量分布[J].自然资源研究,1985,(2):38~46.
[44]具诚,蒋劲松.吉林省鹤类的分布与数量[J].吉林林业科技,1988,(6):27.
[45]王凌,李秀珍.辽河三角洲湿地以丹顶鹤生境保护为核心的退耕预案研究[J].生态学报,2004,24(8):1710~1719.
[46]胡远满,舒莹,等.辽宁双台河口自然保护区丹顶鹤繁殖生境变化及其繁殖容量分析[J].生态学杂志,2004,23(5):7~12.
[47]马国恩,等.丹顶鹤数量调查方法的初步探讨[J].动物学杂志,1980,15(1):4~7.
[48]李晓民,李凤山,杨锡臣.挠力河流域环境变化对鹤类数量分布影响研究.见:王岐山主编.鹤类研究现状及二十一世纪展望,2002,(22):64.
[49]邢莲莲,杨贵生,潘艳秋.内蒙古的鹤类资源与保护.见:王岐山主编.中国鹤类研究[R].昆明:云南教育出版社,2005.
[50]李晓民.黑龙江省又发现一个丹顶鹤繁殖地[J].野生动物,1999,20(5):45.
[51]孔凡哲,芮孝芳.基于地形特征的流域水文相似性[J].地理研究,2003,22(6):709 ~715.
[52]袁金国.遥感图像数字处理[M],中国环境科学出版社,2005.
[53]党安荣,王晓林,等. ERDAS IMAGINE遥感图像处理方法[M].北京:清华大学出版社, 2002.
[54]申晋利.吉林省莫莫格地区面向对象湿地遥感分类方法研究[D].中国地质大学,北京.2006.
[55]李小娟,刘晓萌,等. ENVI遥感影像处理教程[M].北京:中国环境科学出版社, 2008.
[56]余晶,蒋平安,高敏华.基于决策树的土地利用分类方法研究[J].新疆农业科学,2009,46(2):430~434.
[57]李爽,张二勋.基于决策树的遥感影像分类方法研究[J].地域研究与开发, 2003,22(1):17~21.
[58].http://www.gisjc.com/a/ENVI/2010/0414/44.html [J/OL],2010.4.14.
[59]赵英时,等.遥感应用分析原理与方法[M].北京:科学出版社, 2002.
[60]李开丽.东亚飞蝗生境的遥感分类研究[D].南京:南京师范大学, 2005.
[61]王庆光,等.基于人工神经网络的湿地遥感分类研究[D].北京:中国地质大学, 2006.
[62]朱丽娟,刘红玉.挠力河流域丹顶鹤繁殖期生境景观连接度分析[J].生态与农村环境学报,2008,24(2):12~16.
[63]万冬梅,高玮,等.生境破碎化对丹顶鹤巢位选择的影响[J].应用生态学报,2002, 13(5): 581~584.
[64]田新光.面向对象高分辨率遥感影像信息提取[D].中国测绘科学研究院,2007.
[65]王新辉.面向对象的高分辨率影像香榧分布信息提取研究[D].浙江大学,2008.
[66]孙艳玲,等.基于决策树和MODIS数据的土地利用分类[J].资源科学,2007,29 (5):169~174.
[67]邬伦,刘瑜,张晶,等.地理信息系统原理、方法和应用[M].科学出版社,2001.
[68]郑晓燕,张艳红,刘万崧.扎龙湿地水体遥感提取过程中云阴影去除方法研究[J].湿地科学,2010,8(1):21~27.
[69]郭凯,孙培新,刘卫国.利用ERDAS IMAGINE从遥感影像中提取植被指数[J],西部探矿工程, 2005,109(6):210~212.
[70]姜玲,黄家柱. ALOS数据湿地植被信息提取研究[J].河北遥感,2007,(2): 23~26.
[71]韩芳,鸟日娜,李兴华.湿地资源的遥感监测方法[J]内蒙古气象,2006,4:33~35.
[72]延昊,王长耀,牛铮等.遥感植被指数对多时相AVHRR数据主成分分析的影响[J]遥感学报,2002,1:30~34.
[73]谭继强,孙丽梅.基于SPOT5遥感影像的湿地资源调查方法研引[J].测绘与空间地理信息,2007,2:19-26.
[74]李明师,齐家,彭世奎等.基于TM和ASTER的植被指数建模性能对比[J].林业资源管理,2006,6:7~80.
[75]刘义娜.面向对象的高分辨率遥感影像信息提取与尺度效应分析[D].东北师范大学, 2009.
[2]赵魁义,等.湿地生物多样性保护[M].中国林业出版社,2008-9-1.
[3]刘红玉,杨青,等.湿地景观变化对水禽生境影响研究进展[J].湿地科学,2003,1 (2):115~121.
[4]盛连喜,何春光,等.中国水禽的保护生物学研究进展[J].湿地科学,2003,1(1):26 ~32.
[5]余新晓,牛健植,等.景观生态学[M].北京:高等教育出版社,2006.
[6]逢世良.扎龙自然保护区丹顶鹤的航空调查报告[C].高师理科学刊,2000,20(2) :59~60.
[7]吴庆明.基于“3S”技术的扎龙湿地恢复初期丹顶鹤繁殖期生境选择研究[D].东北林业大学,2005.
[8]马逸清,李晓民.我国丹顶鹤的数量分布.国际鹤类学术讨论论文摘要[C].1987, 12.
[9]张艳红,邓伟,等.向海自然保护区丹顶鹤生境结构空间特征[J].生态学报,2006, 26(11):3725~3731.
[10]张柏.遥感技术在中国湿地研究中的应用[J].遥感技术与应用,1996(3):67~71.
[11]王伟,陆健健.三垟湿地生态系统服务功能及其价值[J].生态学报,2005,25.
[12]李天宏,赵智杰,等.深圳河河口红树林变化的多时相遥感分析[J].遥感学报, 2002,6(5):364~370.
[13]Intermountain Aquatica INC. Environmental conserving and habitat restoration [Z]. Http//www.Intermountain aquatics.com.2002.
[14]李枫,杨红军,等.扎龙湿地丹顶鹤巢址选择研究[J].东北林业大学学报,1999,27 (6):57~60.
[15]孔博,张树清,等.遥感和GIS技术的水禽栖息地适宜性评价中的应用[J].遥感学报,2008,12(6):1001~1009.
[16]Dissanska M. Object-based classification of very high resolution panchromatic images for evaluating recent change in the structure of patterned peatlands[J]. CANADIAN JOURNAL OF REMOTE SENSING,2009,32(2):189~215.
[17]王建强,吴连喜,张岩岩.基于3S技术湿地遥感信息分类的研究[J].水利科技与经济,2006,12(10):718~720.
[18]汪闽,骆建承,周成虎,等.结合高斯马尔可夫随机场纹理模型与支撑向量机在高分辨率图像上提取道路网[J].遥感学报,2005,9(3):271~276.
[19]Haack,B. Monitoring wetland changes with remote sensing: an east Africa example[J].Environmental Management.1996,20:411~419.
[20]Lunetta,R.S.,M.E.Balogh. Application of Multi-Temporal Landsat-5 TM Imagery for Wetland Identification[J].Photogrammetric Engineering and Remote Sensing, 1999,65(11):1303~1310.
[21]Andrew ME, Habitat suitability modelling of an invasive plant with advanced remote sensing data[J]. DIVERSITY AND DISTRIBUTIONS, 2009,14(4): 627~640.
[22]Jin-kang, FENG Xue-zhi, ect. The methods of extracting water information from SPOT image[J]. CHINESE GEOGRAPHICAL SCIENCE,2002,12(1):68~72.
[23] Bridget S H,Waiters D,Powell C,Shrubsele D. Wetland management:An analysis of past practice and recent policy changes in Ontario[J].Journal of Environmental Management.In Press,Available online 20 March. 2006.
[24]骆建承,王钦敏,马江洪,等.遥感图像最大似然分类分类方法的EM改进算法[J].测绘学报,2002,31(3):234~239.
[25]Kindscher K.,Fraser A.,Jakubauskas M.E.Debinski D.M..Identifying wetland meadows in Grand Teton National Park using remote sensing and average wetalnd values.Wetlands Ecology and Management,1998,5:265~273.
[26]Bronge L B, Naslund-Landenmark B. Wetland classification for Swedish CORINE Land Cover adopting a semi-automatic interactive approach. Canadian Jounral of Remote Sensing,2002,28(2):139~155.
[27]Rebecca L. Phillips, Ofer Beeri, Edward Shawn DeKeyser. Remote wetland assessment for Missouri coteau prairie glacial basins .Weltands,2005,25(2):335~ 349.
[28]Corey Baker, Rick Lawrence, MAPPING WETLANDS AND RIPARIAN AREAS USING LANDSAT ETM+ IMAGERY AND DECISION-TREE-BASE- D MODELS[J]. The Society of Wetland Scientist. 2006,26(2):465~474.
[29]雍国玮,石承仓,等.高原若尔盖草地沙化及湿地萎缩动态遥感监测[J].山地学报,2003,21(6):758~762.
[30]张树清.三江平原湿地遥感分类模式研究[J].遥感技术与应用,1999,14(1):54~ 58.
[31]衣伟宏,杨柳,等.基于ETM+影像的扎龙湿地遥感分类研究[J].湿地科学,2004, 2(3):208~212.
[32]牛明香.湿地资源遥感动态监测及其对生态环境的影响研究[D].山东农业大学硕士毕业论文,2004.
[33]张杰,沈芳,刘志国.长江口潮滩湿地植被光谱分析与遥感检测[J].华东师范大学学报(自然科学版),2007,7(4):42~63.
[34]陈定贵,周德民,等.三江平原洪河自然保护区湿地遥感分类研究[J].遥感技术与应用,2007,22(4):485~491.
[35]田庆久,等.江苏近海岸水深遥感研究[J].遥感学报,2007,11(3):373~379.
[36]张寿选,段洪涛,等.基于水体透明度反演的太湖水生植被遥感信息提取[J],湖泊科学,2008,20(2):184~190.
[37]刘丽,等.图像纹理特征提取方法探讨综述[J].中国图像图形学报,2009,64[4]:622~734.
[38]张树清,等. 3S支持下的中国典型沼泽湿地景观时空动态变化研究[M].中国长春:吉林大学出版社,2008,8.
[39]李长友.扎龙湿地生态系统保护对策研究[R].黑龙江省科学技术委员会软科学课题研究报告,黑龙江扎龙国家级自然保护区管理局课题组, 2002.
[40]李长友.走进扎龙湿地[Z].欧盟小项目便捷基金项目资助,黑龙江扎龙国家级自然保护区管理局,.2005.
[41]汪青雄.扎龙湿地丹顶鹤(Grus japonensis)栖息地利用,东北林业大学,2008.
[42]李文发,赵英杰,李金华.兴凯湖自然保护区的丹顶鹤数量调查初报[J].黑龙江八一农垦大学学报,1984,(2):87- 88.
[43]马逸清等.三江平原丹顶鹤的数量分布[J].自然资源研究,1985,(2):38~46.
[44]具诚,蒋劲松.吉林省鹤类的分布与数量[J].吉林林业科技,1988,(6):27.
[45]王凌,李秀珍.辽河三角洲湿地以丹顶鹤生境保护为核心的退耕预案研究[J].生态学报,2004,24(8):1710~1719.
[46]胡远满,舒莹,等.辽宁双台河口自然保护区丹顶鹤繁殖生境变化及其繁殖容量分析[J].生态学杂志,2004,23(5):7~12.
[47]马国恩,等.丹顶鹤数量调查方法的初步探讨[J].动物学杂志,1980,15(1):4~7.
[48]李晓民,李凤山,杨锡臣.挠力河流域环境变化对鹤类数量分布影响研究.见:王岐山主编.鹤类研究现状及二十一世纪展望,2002,(22):64.
[49]邢莲莲,杨贵生,潘艳秋.内蒙古的鹤类资源与保护.见:王岐山主编.中国鹤类研究[R].昆明:云南教育出版社,2005.
[50]李晓民.黑龙江省又发现一个丹顶鹤繁殖地[J].野生动物,1999,20(5):45.
[51]孔凡哲,芮孝芳.基于地形特征的流域水文相似性[J].地理研究,2003,22(6):709 ~715.
[52]袁金国.遥感图像数字处理[M],中国环境科学出版社,2005.
[53]党安荣,王晓林,等. ERDAS IMAGINE遥感图像处理方法[M].北京:清华大学出版社, 2002.
[54]申晋利.吉林省莫莫格地区面向对象湿地遥感分类方法研究[D].中国地质大学,北京.2006.
[55]李小娟,刘晓萌,等. ENVI遥感影像处理教程[M].北京:中国环境科学出版社, 2008.
[56]余晶,蒋平安,高敏华.基于决策树的土地利用分类方法研究[J].新疆农业科学,2009,46(2):430~434.
[57]李爽,张二勋.基于决策树的遥感影像分类方法研究[J].地域研究与开发, 2003,22(1):17~21.
[58].http://www.gisjc.com/a/ENVI/2010/0414/44.html [J/OL],2010.4.14.
[59]赵英时,等.遥感应用分析原理与方法[M].北京:科学出版社, 2002.
[60]李开丽.东亚飞蝗生境的遥感分类研究[D].南京:南京师范大学, 2005.
[61]王庆光,等.基于人工神经网络的湿地遥感分类研究[D].北京:中国地质大学, 2006.
[62]朱丽娟,刘红玉.挠力河流域丹顶鹤繁殖期生境景观连接度分析[J].生态与农村环境学报,2008,24(2):12~16.
[63]万冬梅,高玮,等.生境破碎化对丹顶鹤巢位选择的影响[J].应用生态学报,2002, 13(5): 581~584.
[64]田新光.面向对象高分辨率遥感影像信息提取[D].中国测绘科学研究院,2007.
[65]王新辉.面向对象的高分辨率影像香榧分布信息提取研究[D].浙江大学,2008.
[66]孙艳玲,等.基于决策树和MODIS数据的土地利用分类[J].资源科学,2007,29 (5):169~174.
[67]邬伦,刘瑜,张晶,等.地理信息系统原理、方法和应用[M].科学出版社,2001.
[68]郑晓燕,张艳红,刘万崧.扎龙湿地水体遥感提取过程中云阴影去除方法研究[J].湿地科学,2010,8(1):21~27.
[69]郭凯,孙培新,刘卫国.利用ERDAS IMAGINE从遥感影像中提取植被指数[J],西部探矿工程, 2005,109(6):210~212.
[70]姜玲,黄家柱. ALOS数据湿地植被信息提取研究[J].河北遥感,2007,(2): 23~26.
[71]韩芳,鸟日娜,李兴华.湿地资源的遥感监测方法[J]内蒙古气象,2006,4:33~35.
[72]延昊,王长耀,牛铮等.遥感植被指数对多时相AVHRR数据主成分分析的影响[J]遥感学报,2002,1:30~34.
[73]谭继强,孙丽梅.基于SPOT5遥感影像的湿地资源调查方法研引[J].测绘与空间地理信息,2007,2:19-26.
[74]李明师,齐家,彭世奎等.基于TM和ASTER的植被指数建模性能对比[J].林业资源管理,2006,6:7~80.
[75]刘义娜.面向对象的高分辨率遥感影像信息提取与尺度效应分析[D].东北师范大学, 2009.