基于粗糙集理论的北京山区森林健康预警研究
|
摘要
保护森林就是保护人类的生存环境和社会的持续发展,北京作为全国的政治、经济和文化中心,对首都的森林生态的保护有着深远的意义。本文以森林健康理论、预警理论和粗糙集理论为研究基础,并将三种理论有机结合,以森林健康评价为基础,森林健康预警为主体,粗糙集理论为途径构建森林健康预警系统：结合森林健康理论和预警理论,论述了森林健康预警研究的必要性和重要性；分析影响森林健康的因素及其作用机理,基于PSR(压力-状态-响应)框架深入分析了森林健康预警的过程；按照森林健康预警系统的构建思路,构建森林健康预警系统；基于粗糙集理论构建了森林健康预警计算模型,并进行指标属性的约简和权重的计算；通过建立预警调控措施专家库和各预警管理系统,实现了预警调控子系统的功能。
森林健康预警研究是森林健康理论与预警理论相结合的产物,本研究首次将粗糙集理论应用于森林健康预警中,以粗糙集理论方法作为森林健康预警计算模型,从指标的约简到指标权重的确定,都为森林健康预警提供了新的方法。
通过将粗糙集约简理论和BP(误差反向传递)神经网络相集合,扩展了粗糙集理论在预警预测领域的应用。将粗糙集方法作为BP神经网络的前置预处理系统,可以简化神经网络的构造,减小神经网络的训练集并缩短网络的训练学习时间。本研究将这一混合模型应用于鹫峰国家森林公园的健康预测中,分析并验证了这一模型的合理性和可靠性。
利用建立的森林生态系统健康预警系统,结合北京山区森林健康发展状况,对北京山区的森林生态系统健康进行预警。预警诊断子系统输出结果表明,北京市山区15块研究样地中,处于无警状态的有4块,轻警4块,中警3块,重警3块,巨警1块。整体警情较严重,森林管理者通过利用预警调控子系统分析森林“病变”的根源,结合专家库,提出相应的森林健康恢复措施,可及时、有效的排除警患。
Protection of forests is to protect the human living environment and sustainable development of society. Protection of forest ecosystems is very necessary for Beijing, due to the center of political, economic and cultural of China. In this thesis, an early warning system are built, based on theory of forest health, early warning and rough set theory, as well as evaluation of forest health and forest health warning. We study the need for forest health, the importance of early warning research, analysis of factors affecting forest health and the role of incentives systematically. And also the process of forest health warning is analyzed in depth based on the PSR framework. According to the idea of constructing of forest health warning system, a forest health early warning systems was built; based on rough set model, a model of forest health warning was constructed and calculate the indicator properties and weight reduction calculation; Through the establishment of early warning and the warning expert database management system we achieve the Early Warning Control Subsystem function.
Research on early warning of forest health is combination of forest health and early warning theory. In this paper, we first apply the rough set method to early warning of forest health. The rough set theory method is used as early warning model of forest health from reducing to weighting the indicators. Both of them provide a new approach to the early warning of forest health.
By integrating the reduction of rough set and BP neural network, we expand the rough set theory in the field of forest health early warning and forecast. Taking the rough set as a preconditioning system of BP neural network, we can simplify the neural network's structure and reduce the training set of neural network so as to save a lot of training time to learn. The hybrid model is applied to the health forecast of Jiufeng National Forest Park, which proves that the model is reasonable and reliable.
We use the established forest ecosystem and combine the development status of Beijing forest health for early warning on Beijing forest ecosystem health. The outputs of early warning diagnostic subsystem showed that there are four no warning blocks, four light warning blocks, three heavy warning blocks and one huge warning block in the fifteen forest areas of Beijing. The whole warning of Beijing forest areas is more serious. Forest mangers analyzed the source of forest disease through combining warning controlled subsystem with expert databases. At last, the corresponding forest health restoration measures can be timely and effective provided in order to suffering from exclusions.
森林健康预警研究是森林健康理论与预警理论相结合的产物,本研究首次将粗糙集理论应用于森林健康预警中,以粗糙集理论方法作为森林健康预警计算模型,从指标的约简到指标权重的确定,都为森林健康预警提供了新的方法。
通过将粗糙集约简理论和BP(误差反向传递)神经网络相集合,扩展了粗糙集理论在预警预测领域的应用。将粗糙集方法作为BP神经网络的前置预处理系统,可以简化神经网络的构造,减小神经网络的训练集并缩短网络的训练学习时间。本研究将这一混合模型应用于鹫峰国家森林公园的健康预测中,分析并验证了这一模型的合理性和可靠性。
利用建立的森林生态系统健康预警系统,结合北京山区森林健康发展状况,对北京山区的森林生态系统健康进行预警。预警诊断子系统输出结果表明,北京市山区15块研究样地中,处于无警状态的有4块,轻警4块,中警3块,重警3块,巨警1块。整体警情较严重,森林管理者通过利用预警调控子系统分析森林“病变”的根源,结合专家库,提出相应的森林健康恢复措施,可及时、有效的排除警患。
Protection of forests is to protect the human living environment and sustainable development of society. Protection of forest ecosystems is very necessary for Beijing, due to the center of political, economic and cultural of China. In this thesis, an early warning system are built, based on theory of forest health, early warning and rough set theory, as well as evaluation of forest health and forest health warning. We study the need for forest health, the importance of early warning research, analysis of factors affecting forest health and the role of incentives systematically. And also the process of forest health warning is analyzed in depth based on the PSR framework. According to the idea of constructing of forest health warning system, a forest health early warning systems was built; based on rough set model, a model of forest health warning was constructed and calculate the indicator properties and weight reduction calculation; Through the establishment of early warning and the warning expert database management system we achieve the Early Warning Control Subsystem function.
Research on early warning of forest health is combination of forest health and early warning theory. In this paper, we first apply the rough set method to early warning of forest health. The rough set theory method is used as early warning model of forest health from reducing to weighting the indicators. Both of them provide a new approach to the early warning of forest health.
By integrating the reduction of rough set and BP neural network, we expand the rough set theory in the field of forest health early warning and forecast. Taking the rough set as a preconditioning system of BP neural network, we can simplify the neural network's structure and reduce the training set of neural network so as to save a lot of training time to learn. The hybrid model is applied to the health forecast of Jiufeng National Forest Park, which proves that the model is reasonable and reliable.
We use the established forest ecosystem and combine the development status of Beijing forest health for early warning on Beijing forest ecosystem health. The outputs of early warning diagnostic subsystem showed that there are four no warning blocks, four light warning blocks, three heavy warning blocks and one huge warning block in the fifteen forest areas of Beijing. The whole warning of Beijing forest areas is more serious. Forest mangers analyzed the source of forest disease through combining warning controlled subsystem with expert databases. At last, the corresponding forest health restoration measures can be timely and effective provided in order to suffering from exclusions.
引文
[1]程晓燕.北京山区森林景观格局分析[D].北京：北京林业大学,2009.
[2]陈高,代力民,范竹华,等.森林生态系统健康及其评估监测[J].应用生态学报,2002,13(5)：605-610.
[3]陈军飞.城市生态系统诊断预警研究[D].江苏：河海大学,2004.
[4]陈治谏,陈国阶.环境影响评价的预警系统研究[J].环境科学,1992,13(4)：20-26.
[5]陈国阶,何锦峰.生态环境预警的理论和方法探讨[J].重庆环境科学,1999,21(4)：8-11.
[6]樊邦勇.城市房地产预警系统研究[D].四川：西南交通大学,2006.
[7]范燕敏.天山北坡中断伊犁绢嵩荒漠退化草地土壤质量的演变与评价及预警系统的研究[D].新疆：新疆农业大学,2009.
[8]谷建才.华北土石山区典型区域主要类型森林健康分析与评价[D].北京：北京林业大学,2006.
[9]甘敬.北京山区森林健康评价研究[D].北京：北京林业大学,2007.
[10]高均凯.森林健康基本理论及评价方法研究[D].北京：北京林业大学,2007：18-21.
[11]甘敬,张振明,余新晓,等.森林健康监测与评价研究[J].水土保持研究,2006,13(3)：177-178.
[12]高志亮.北京典型区域森林生态系统健康评价与预警研究[D].北京：北京林业大学,2009.
[13]高志亮,余新晓,等.北京八达岭林场森林健康评价研究[J].林业资源管理,2008,(4)：78-79.
[14]谷建才,陆贵巧,等.森林健康评价指标及应用研究[J].河北农业大学报.2006,29(2)：68-69.
[15]关欣,衣晓,何友.一种新的粗糙集属性约简方法及其应用[J].控制与决策,2009,24(3)：464-467.
[16]黄小磊.粗糙集属性约简及在电力系统中的若干应用[D].浙江：浙江大学,2008.
[17]胡淑恒.巢湖流域的生态安全预警研究[D].合肥：合肥工业大学,2004.
[18]胡和兵.生态敏感地区生态安全评价与预警研究[D].上海：华东师范大学,2006.
[19]贾忠奎,马履一,徐程扬,等.北京市森林资源动态及可持续经营对策[J].干旱区资源与环境,2006,20(3)：30-36.
[20]刘邵权,陈国阶,陈治谏.三峡库区山地生态系统预警[J].山地学报,2002,20(3)：302-306.
[21]刘传正,李铁锋,程凌鹏,等.区域地质灾害评价预警的递进分析理论与方法[J].水文地质工程地质,2004,(4)：1-8.
[22]刘琛,张贵,肖化顺.“3S”在森林病虫害预警中的应用与展望[J].湖南环境生物职业技术学院学报,2004,(4)：8-13.
[23]吕光辉.中国西部干旱区生态安全评价、预警与调控研究[D].乌鲁木齐：新疆大学,2005.
[24]吕昭智,马桂龙,李永龙,等.新疆北部棉铃虫越冬代羽化物候学模型[J].石河子大学学报(自然科学版),2004,22(6)：461-464.
[25]李升.地下水环境健康预警研究—以黄河下游悬河段(河南)为例[D].吉林大学,2008.
[26]鲁绍伟,刘凤芹,余新晓,等.北京市八达岭林场森林生态系统健康性评价[J].水土保持学报,2006a,20(3)：79-81,105.
[27]鲁绍伟,余新晓,刘凤芹,等.北京八达岭林场森林燃烧性及防火措施研究[J].北京林业大学学报,2006b,28(3)：109-114.
[28]李静锐,余新晓,张振明,等.森林生态系统健康评价指标体系的建立[J].水土保持研究,2007,14(3)：173-175.
[29]马立.北京山地森林健康综合评价体系的构建与应用[D].北京：北京林业大学,2007：24-32.
[30]马立,韩海荣,马钦彦,等.森林生态系统健康的研究进展[J].林业调查规划,2007,32(1)：103-110.
[31]彭道黎,张志华,靳云燕.北京市生态公益林经营目标及指标体系的研究[J].林业调查规划,2006,31(6)：16-19.
[32]钱锋.基于粗糙集知识的离散化方法和约简算法的研究[D].福建：厦门大学,2007.
[33]邵东国,李元红,王忠静,等.基于神经网络的干旱内陆河流域生态环境预警方法研究[J].中国农村水利水电,1996,(6)：10-12.
[34]苏维词,李久林.乌江流域生态环境预警评价初探[J].贵州科学,1997,15(3)：207-214.
[35]中磊,邓华锋,等.北京市高保护价值森林状况[J].安徽农业科学,2010,38(16)：8779-8781.
[36]谭旭,唐云岚,等.杂合数据的粗糙集属性约简方法[J].国防科技大学学报：2008,30(6)：83-88.
[37]吴延熊.区域森林资源预警系统的研究[D].北京：北京林业大学,1998.
[38]武巧英,陈丽华,于景金,等.北京鹫峰国家森林公园健康评价研究[J].中国农学通报,2010,26(12)：90-93.
[39]王懿祥.天然林监测预警系统研究[D].昆明：西南林学院,2004.
[40]王雄.赤峰市森林资源—环境—经济复合系统可持续发展动态评价及预警[D].内蒙：内蒙古农业大学,2007.
[41]王龙.山西煤炭开发与生态环境预警初探[J].生态经济,1995,(5)：32-36.
[42]王兵,郭浩,王燕,等.森林生态系统健康评估研究进展[J].中国水土保持科学2007,5(3)：114-121.
[43]文传甲.三峡库区农业生态经济系统的预警分析[J].山地研究,1998,16(1)：13-20.
[44]肖劲风,欧阳华,等.森林生态系统健康评价指标及其在中国的应用[J].地理学报,2003,58(6)：803-808.
[45]徐强.区域环境经济与预警[M].北京：中国环境科学出版社,2001.
[46]余新晓,甘敬,李金海,等.森林健康评价、监测与预警[M].北京：科学出版社,2010.
[47]尹华军,刘庆.森林生态系统健康诊断研究进展及亚高山针叶林健康诊断的思考[J].世界科技研究与发展,2003,25(5)：56-61.
[48]余亮,边馥苓.粗糙神经网络在森林火灾预警中的应用[J].武汉大学学报(信息科学版),2006,31(8)：720-723.
[49]邹长新.内陆河流域生态安全研究——以黑河为例[D].南京：南京气象学院,2003.
[50]钟来元,刘黎明.区域综合食物安全风险预警指标体系[J].农业工程学报,2010,26(3)：374-379.
[51]曾文英,王明文.3S技术的发展、现状及其趋势综述[J].新余高专学报,2003,8(2)：16-24.
[52]张建设.基于粗糙集的多属性决策问题[D].重庆：重庆大学,2006.
[53]张大任.洞庭湖生态环境预警[J].地理学与国土研究,1991,7(2)：42-44.
[54]张东.浑善达克沙地荒漠化灾害预警指标体系的研究[D].北京：北京林业大学,2005.
[55]张华.区域特色农业可持续发展诊断预警研究[D].北京：中国农业科学研究院,2009.
[56]张秋根,王桃云,钟全林.森林生态环境健康评价初探[J].水土保持学报,2003,17(5)：16-18.
[57]Angus J. Carnegie.Forest health condition in New South Wales, Australia,1996-2005. Ⅱ. Fungal damage recorded in eucalypt plantations during forest health surveys and their management[J]. Australasian Plant Pathology,2007,36(3):225-239.
[58]Ali Asgary, Jason K.Levy, Nader Mehregan.Estimating willingness to pay for a hypothetical earthquake early warning system[J]. Environmental Hazards.2007.7(4):312-320.
[59]Costanza R, Norton BG, Haskell BD.Ecosystem Health: New Goals for Environmental Management[M]. Washington DC:Island Press,1992
[60]Chen.K, Pettorossi A., Ras Z., Skowron A. Attributes and rough properties in information system. International Journal of Approximate Reasoning,1998:365-376.
[61]Ivanova Petya I, Tagarev Todor D. Indicator space configuration for early warning of violent political conflicts by genetic algorithms[J]. Annals of Operations Research.1997,1(4):287-311.
[62]Lai P H, Nepal S K. Local perspectives of ecotourism development in Tawushan Nature Reserve, Taiwan[J]. Tourism Management,2006,27:1117-1129.
[63]McMichael A J, Bolin B, Costanza R. Globalization and the sustainability of human health: an ecological perspective [J]. Bioscience,1999,49: 205-210.
[64]Macro Ferretti. Forest Health Assessment and Monitoring-Issues for Consideration [J]. Environmental Monitoring and Assessment,1997,48:45-72.
[65]O'Neill K. P, Amacher M C, Perry C H. Soils as an indicator of forest health: a guide to the collection, analysis, and interpretation of soil indicator data in the Forest Inventory and Analysis program [R]. In:St. Paul, MN:U.S. Department of Agriculture, Forest Service, North Central Research Station.2005,53.
[66]Pawlak Z. Rough sets — theoretical aspects of reasoning about data[M]. Dordrecht:Kluwer Academic Publishers.1991:9-30.
[67]Pawlak Z. Rough set theory and it s application to data analysis [J]. Int J of Cybernetics and Systems,1998,29:661-688.
[68]R.Neil Sampson,David Lewis Adams. Assessing forest ecosystem health in the Inland West[M]. Binghamton:The Haworth Press,1994.
[69]Samuel A. Alexander,Craig J. Palmer. Forest Health Monitoring in the United States: First Four Years[J]. Environmental monitoring and assessment,1999,55(2):267-277.
[70]S.Greco, B.Matarazzo. Rough set approach to decisions under Risk[J]. W.Ziarko and Yao:RSSCTC 2000, LNAI 2005:160-169.
[71]Wunder S. Ecotourism and economic incentives-an empirical approach[J]. Ecological Economics, 2000,32:465-479.
[72]Zarnoch S J, Bechtold W A, Stolte K W. Using crown condition variables as indicators of forest health [J].Canadian Journal of Forest Research,2004,34(5):1057-1070.
[2]陈高,代力民,范竹华,等.森林生态系统健康及其评估监测[J].应用生态学报,2002,13(5)：605-610.
[3]陈军飞.城市生态系统诊断预警研究[D].江苏：河海大学,2004.
[4]陈治谏,陈国阶.环境影响评价的预警系统研究[J].环境科学,1992,13(4)：20-26.
[5]陈国阶,何锦峰.生态环境预警的理论和方法探讨[J].重庆环境科学,1999,21(4)：8-11.
[6]樊邦勇.城市房地产预警系统研究[D].四川：西南交通大学,2006.
[7]范燕敏.天山北坡中断伊犁绢嵩荒漠退化草地土壤质量的演变与评价及预警系统的研究[D].新疆：新疆农业大学,2009.
[8]谷建才.华北土石山区典型区域主要类型森林健康分析与评价[D].北京：北京林业大学,2006.
[9]甘敬.北京山区森林健康评价研究[D].北京：北京林业大学,2007.
[10]高均凯.森林健康基本理论及评价方法研究[D].北京：北京林业大学,2007：18-21.
[11]甘敬,张振明,余新晓,等.森林健康监测与评价研究[J].水土保持研究,2006,13(3)：177-178.
[12]高志亮.北京典型区域森林生态系统健康评价与预警研究[D].北京：北京林业大学,2009.
[13]高志亮,余新晓,等.北京八达岭林场森林健康评价研究[J].林业资源管理,2008,(4)：78-79.
[14]谷建才,陆贵巧,等.森林健康评价指标及应用研究[J].河北农业大学报.2006,29(2)：68-69.
[15]关欣,衣晓,何友.一种新的粗糙集属性约简方法及其应用[J].控制与决策,2009,24(3)：464-467.
[16]黄小磊.粗糙集属性约简及在电力系统中的若干应用[D].浙江：浙江大学,2008.
[17]胡淑恒.巢湖流域的生态安全预警研究[D].合肥：合肥工业大学,2004.
[18]胡和兵.生态敏感地区生态安全评价与预警研究[D].上海：华东师范大学,2006.
[19]贾忠奎,马履一,徐程扬,等.北京市森林资源动态及可持续经营对策[J].干旱区资源与环境,2006,20(3)：30-36.
[20]刘邵权,陈国阶,陈治谏.三峡库区山地生态系统预警[J].山地学报,2002,20(3)：302-306.
[21]刘传正,李铁锋,程凌鹏,等.区域地质灾害评价预警的递进分析理论与方法[J].水文地质工程地质,2004,(4)：1-8.
[22]刘琛,张贵,肖化顺.“3S”在森林病虫害预警中的应用与展望[J].湖南环境生物职业技术学院学报,2004,(4)：8-13.
[23]吕光辉.中国西部干旱区生态安全评价、预警与调控研究[D].乌鲁木齐：新疆大学,2005.
[24]吕昭智,马桂龙,李永龙,等.新疆北部棉铃虫越冬代羽化物候学模型[J].石河子大学学报(自然科学版),2004,22(6)：461-464.
[25]李升.地下水环境健康预警研究—以黄河下游悬河段(河南)为例[D].吉林大学,2008.
[26]鲁绍伟,刘凤芹,余新晓,等.北京市八达岭林场森林生态系统健康性评价[J].水土保持学报,2006a,20(3)：79-81,105.
[27]鲁绍伟,余新晓,刘凤芹,等.北京八达岭林场森林燃烧性及防火措施研究[J].北京林业大学学报,2006b,28(3)：109-114.
[28]李静锐,余新晓,张振明,等.森林生态系统健康评价指标体系的建立[J].水土保持研究,2007,14(3)：173-175.
[29]马立.北京山地森林健康综合评价体系的构建与应用[D].北京：北京林业大学,2007：24-32.
[30]马立,韩海荣,马钦彦,等.森林生态系统健康的研究进展[J].林业调查规划,2007,32(1)：103-110.
[31]彭道黎,张志华,靳云燕.北京市生态公益林经营目标及指标体系的研究[J].林业调查规划,2006,31(6)：16-19.
[32]钱锋.基于粗糙集知识的离散化方法和约简算法的研究[D].福建：厦门大学,2007.
[33]邵东国,李元红,王忠静,等.基于神经网络的干旱内陆河流域生态环境预警方法研究[J].中国农村水利水电,1996,(6)：10-12.
[34]苏维词,李久林.乌江流域生态环境预警评价初探[J].贵州科学,1997,15(3)：207-214.
[35]中磊,邓华锋,等.北京市高保护价值森林状况[J].安徽农业科学,2010,38(16)：8779-8781.
[36]谭旭,唐云岚,等.杂合数据的粗糙集属性约简方法[J].国防科技大学学报：2008,30(6)：83-88.
[37]吴延熊.区域森林资源预警系统的研究[D].北京：北京林业大学,1998.
[38]武巧英,陈丽华,于景金,等.北京鹫峰国家森林公园健康评价研究[J].中国农学通报,2010,26(12)：90-93.
[39]王懿祥.天然林监测预警系统研究[D].昆明：西南林学院,2004.
[40]王雄.赤峰市森林资源—环境—经济复合系统可持续发展动态评价及预警[D].内蒙：内蒙古农业大学,2007.
[41]王龙.山西煤炭开发与生态环境预警初探[J].生态经济,1995,(5)：32-36.
[42]王兵,郭浩,王燕,等.森林生态系统健康评估研究进展[J].中国水土保持科学2007,5(3)：114-121.
[43]文传甲.三峡库区农业生态经济系统的预警分析[J].山地研究,1998,16(1)：13-20.
[44]肖劲风,欧阳华,等.森林生态系统健康评价指标及其在中国的应用[J].地理学报,2003,58(6)：803-808.
[45]徐强.区域环境经济与预警[M].北京：中国环境科学出版社,2001.
[46]余新晓,甘敬,李金海,等.森林健康评价、监测与预警[M].北京：科学出版社,2010.
[47]尹华军,刘庆.森林生态系统健康诊断研究进展及亚高山针叶林健康诊断的思考[J].世界科技研究与发展,2003,25(5)：56-61.
[48]余亮,边馥苓.粗糙神经网络在森林火灾预警中的应用[J].武汉大学学报(信息科学版),2006,31(8)：720-723.
[49]邹长新.内陆河流域生态安全研究——以黑河为例[D].南京：南京气象学院,2003.
[50]钟来元,刘黎明.区域综合食物安全风险预警指标体系[J].农业工程学报,2010,26(3)：374-379.
[51]曾文英,王明文.3S技术的发展、现状及其趋势综述[J].新余高专学报,2003,8(2)：16-24.
[52]张建设.基于粗糙集的多属性决策问题[D].重庆：重庆大学,2006.
[53]张大任.洞庭湖生态环境预警[J].地理学与国土研究,1991,7(2)：42-44.
[54]张东.浑善达克沙地荒漠化灾害预警指标体系的研究[D].北京：北京林业大学,2005.
[55]张华.区域特色农业可持续发展诊断预警研究[D].北京：中国农业科学研究院,2009.
[56]张秋根,王桃云,钟全林.森林生态环境健康评价初探[J].水土保持学报,2003,17(5)：16-18.
[57]Angus J. Carnegie.Forest health condition in New South Wales, Australia,1996-2005. Ⅱ. Fungal damage recorded in eucalypt plantations during forest health surveys and their management[J]. Australasian Plant Pathology,2007,36(3):225-239.
[58]Ali Asgary, Jason K.Levy, Nader Mehregan.Estimating willingness to pay for a hypothetical earthquake early warning system[J]. Environmental Hazards.2007.7(4):312-320.
[59]Costanza R, Norton BG, Haskell BD.Ecosystem Health: New Goals for Environmental Management[M]. Washington DC:Island Press,1992
[60]Chen.K, Pettorossi A., Ras Z., Skowron A. Attributes and rough properties in information system. International Journal of Approximate Reasoning,1998:365-376.
[61]Ivanova Petya I, Tagarev Todor D. Indicator space configuration for early warning of violent political conflicts by genetic algorithms[J]. Annals of Operations Research.1997,1(4):287-311.
[62]Lai P H, Nepal S K. Local perspectives of ecotourism development in Tawushan Nature Reserve, Taiwan[J]. Tourism Management,2006,27:1117-1129.
[63]McMichael A J, Bolin B, Costanza R. Globalization and the sustainability of human health: an ecological perspective [J]. Bioscience,1999,49: 205-210.
[64]Macro Ferretti. Forest Health Assessment and Monitoring-Issues for Consideration [J]. Environmental Monitoring and Assessment,1997,48:45-72.
[65]O'Neill K. P, Amacher M C, Perry C H. Soils as an indicator of forest health: a guide to the collection, analysis, and interpretation of soil indicator data in the Forest Inventory and Analysis program [R]. In:St. Paul, MN:U.S. Department of Agriculture, Forest Service, North Central Research Station.2005,53.
[66]Pawlak Z. Rough sets — theoretical aspects of reasoning about data[M]. Dordrecht:Kluwer Academic Publishers.1991:9-30.
[67]Pawlak Z. Rough set theory and it s application to data analysis [J]. Int J of Cybernetics and Systems,1998,29:661-688.
[68]R.Neil Sampson,David Lewis Adams. Assessing forest ecosystem health in the Inland West[M]. Binghamton:The Haworth Press,1994.
[69]Samuel A. Alexander,Craig J. Palmer. Forest Health Monitoring in the United States: First Four Years[J]. Environmental monitoring and assessment,1999,55(2):267-277.
[70]S.Greco, B.Matarazzo. Rough set approach to decisions under Risk[J]. W.Ziarko and Yao:RSSCTC 2000, LNAI 2005:160-169.
[71]Wunder S. Ecotourism and economic incentives-an empirical approach[J]. Ecological Economics, 2000,32:465-479.
[72]Zarnoch S J, Bechtold W A, Stolte K W. Using crown condition variables as indicators of forest health [J].Canadian Journal of Forest Research,2004,34(5):1057-1070.