地位指数信息图谱的多尺度分析模型
|
摘要
立地质量是指在某一立地上既定森林或者其他植被类型的生产潜力,所以立地质量与树种关联,并有高低之分。立地质量的评价分为直接评价和间接评价,直接评价法包括根据林分蓄积或收获量进行评定和利用林分树高进行评价,间接评价法包括多元地位指数法和植被指数法等。因树高易于测定,而且受林分密度影响较小,因此,利用林分高编制地位指数表的是林业上最常采用的评定方法。
但是定量化的方法进行森林生态系统立地分类还存在许多困难,例如野外样点的采集和调查非常耗时和耗力,许多森林类型缺乏长期定位观测的样地资料,大部分结果主要还是来源于以点为基础的小范围调查的工作。所以寻找一个简便、直观和高精度的方法来估计地位指数是非常有意义的。
遥感技术的发展和数学方法的渗透,为地位指数的研究提供了很好的平台,卫星遥感为大面积研究森林生态系统生产力和及其空间分布提供了一条重要的途径,从遥感角度分析地表植被覆盖情况最有研究价值的是植被指数。20世纪80年代以来,卫星资料在国内外植被研究中得到广泛应用,例如：利用NDVI研究叶片面积指数的空间分布特点及其变化,研究森林生态系统净初级生产力的空间分布等。利用典型的植被指数和地位指数的关系估计和分析森林地位指数及其动态变化对于制定合理的造林方案和经营管理措施具有十分重要的意义。但由于理论、技术与人力和物力等条件的限制,该领域的研究还处于起步阶段,相关报道很少。
“地学信息图谱”概念提出给了作者启发,它是按照一定指标递变规律或分类规律排列的一组能够反映地球科学时空信息规律的数字形式的地图、图表、曲线或图像。于是提出了地位指数信息图谱这个概念,根据树种的不同,按照一定的时间规律排列的能够反映研究区域地位指数时空规律的数字形式地图、图表、曲线或者图像。
本文选择建立一个分析模型对其进行初步的探索研究,为提高模型拟合的精度,采用多尺度分析的方法,它也是地学信息图谱研究中一个重要方法之一。考虑到数据的收集以及地域的限制等问题,综合考虑将实验区域定为湖南攸县黄丰桥林场。
利用TMM遥感图像光谱信息良好的综合性以及归一化指数(NDVI)是植物生长状态以及植被空间分布密度的最佳指示因子这一特性,借助遥感软件(ENVI)和地理信息系统软件(GIS),结合MATLAB小波分析和SPSS逐步回归分析方法,对湖南省株洲攸县黄丰桥林场杉木地位指数遥感估测的尺度进行选择,并对拟合的模型进行分析和精度评估。
研究发现,植被指数在进行5重分解重构后基本达到分解末端,在进行逐步回归分析后,只有植被指数原始数据以及一级分解重构的数据保留,其余4个尺度的数据都被剔除。拟合的模型中包括两个自变量,一是植被指数原始数据,个其一级分解重构后的数据,即一尺度数据。
最后对两个拟合模型的精度评估,检验模型的使用价值。因在林业上的地位指数是一个正整数,而拟合的模型不可能精确的算出这个数字,只能算出一个近似的数字,结合小班的具体情况来得出最终的正整数,所以本文采用建立1：1的比例分析图对模型拟合结果进行精度分析,在SPSS中建立1：1的精度分析图,结果表明两个模型拟合的相关度都在0.9以上,这说明利用遥感指数预估地位指数可以达到一定的精度,具有较高的使用价值。但由于国内少有此类研究,理论尚不成熟,有待深化。
Site quality is established in a stand on the floor of forests or other vegetation types of production potential, so the associated site quality and tree line, and high and low points. Evaluation of site quality assessment is divided into direct and indirect assessment, direct assessment method including the Stand or harvest were evaluated and assessed by stand height, indirect assessment of the status of law, including multiple index and vegetation index method. Easy to determine because of tree height, and less affected by stand density, therefore, prepared using high-index table Stand is the most commonly used assessment of forestry methods.
But the quantitative methods of site classification of forest ecosystems, there are still many difficulties, such as sample collection and field surveys are time consuming and labor-intensive, and many forest types in the lack of long-term observation plots targeting information, most of the results mainly from A point-based small-scale investigations. So looking for a simple, intuitive and accurate method to estimate the position index is very meaningful.
Remote sensing technology and the penetration of mathematical methods for the study site index provides a good platform for the large area of satellite remote sensing of forest ecosystem productivity and its spatial distribution provides an important way, from the perspective of remote sensing of vegetation coverage of most of the vegetation index value.20th century,80 years, satellite data of vegetation at home and abroad has been widely used, for example:the use of leaf area index of NDVI spatial distribution and changes of forest net primary productivity of ecosystems such as the spatial distribution. Use of vegetation index and status of the typical relationship between the estimation and analysis of Indices of forest status index and its dynamics program for the development of the Planting and management measures is of great significance. However, theory, technology and human and material resources and other conditions, researchers in the field still in its infancy, very few relevant reports.
"Geo-information map of put forward the concept of inspiration to the author, which is in accordance with the classification of certain indicators of Homologous or a group of regular arrangement to reflect the laws of earth science spatial information in digital form of maps, charts, curves or images. So I think the concept of site index map information, according to the different species, in accordance with the laws of a certain time arranged to reflect the status of the study area in digital form the law of exponential time and space maps, charts, curves or images.
However, theory, technology and human and material resources and other conditions, use of remote sensing vegetation indices reflect the site quality and its spatial and temporal dynamics research in the field still in its infancy, few relevant reports, so this model was chosen to establish a preliminary analysis The exploration and research, to improve the accuracy of model fitting, this multi-scale analysis method chosen, it is also the study of geo-information spectrum method is an important one. Taking into account data collection, and geographical constraints and other issues, this selection of You County, Hunan Toyohashi yellow cedar forest data as experimental data.
Spectral information of TM remote sensing images using well-integrated and normalized difference index (NDVI) is a plant growth status, and spatial distribution of vegetation density factor of this feature the best indication, with remote sensing software (ENVI) and geographical information systems software (GIS), Combined with MATLAB wavelet analysis and SPSS stepwise regression analysis method, You County, Zhuzhou, Hunan Toyohashi yellow fir forest site index to choose the scale of remote sensing estimation, and model fitting analysis and accuracy assessment.
Found that decomposition of vegetation index during the 5 re-decomposition of the reconstructed basically reached the end, during the stepwise regression analysis, only the raw data of vegetation index and the level decomposition and reconstruction of data retention, and the remaining four scales of data have been excluded. Fitting model includes two independent variables, namely the raw data of vegetation index, a level of decomposition and reconstruction of their data after that data of a scale.
Finally, the accuracy assessment of the two simulation models to test the model using the value. Because of the position index in forestry is a positive integer, and can not fit the model calculates the exact figure can only calculate an approximate figure, combined with the specific circumstances of small classes to come to the end of the positive integers, so this is used to establish Figure 1:1 ratio of precision of the results of model fitting analysis in SPSS to establish the accuracy of analysis of Figure 1:1, the results show that the correlation of two model fitting are above 0.9, indicating that the use of remote sensing index forecast Position index can reach a certain precision, with high use value. But because few such studies, theory is not yet mature, to be deepened.
但是定量化的方法进行森林生态系统立地分类还存在许多困难,例如野外样点的采集和调查非常耗时和耗力,许多森林类型缺乏长期定位观测的样地资料,大部分结果主要还是来源于以点为基础的小范围调查的工作。所以寻找一个简便、直观和高精度的方法来估计地位指数是非常有意义的。
遥感技术的发展和数学方法的渗透,为地位指数的研究提供了很好的平台,卫星遥感为大面积研究森林生态系统生产力和及其空间分布提供了一条重要的途径,从遥感角度分析地表植被覆盖情况最有研究价值的是植被指数。20世纪80年代以来,卫星资料在国内外植被研究中得到广泛应用,例如：利用NDVI研究叶片面积指数的空间分布特点及其变化,研究森林生态系统净初级生产力的空间分布等。利用典型的植被指数和地位指数的关系估计和分析森林地位指数及其动态变化对于制定合理的造林方案和经营管理措施具有十分重要的意义。但由于理论、技术与人力和物力等条件的限制,该领域的研究还处于起步阶段,相关报道很少。
“地学信息图谱”概念提出给了作者启发,它是按照一定指标递变规律或分类规律排列的一组能够反映地球科学时空信息规律的数字形式的地图、图表、曲线或图像。于是提出了地位指数信息图谱这个概念,根据树种的不同,按照一定的时间规律排列的能够反映研究区域地位指数时空规律的数字形式地图、图表、曲线或者图像。
本文选择建立一个分析模型对其进行初步的探索研究,为提高模型拟合的精度,采用多尺度分析的方法,它也是地学信息图谱研究中一个重要方法之一。考虑到数据的收集以及地域的限制等问题,综合考虑将实验区域定为湖南攸县黄丰桥林场。
利用TMM遥感图像光谱信息良好的综合性以及归一化指数(NDVI)是植物生长状态以及植被空间分布密度的最佳指示因子这一特性,借助遥感软件(ENVI)和地理信息系统软件(GIS),结合MATLAB小波分析和SPSS逐步回归分析方法,对湖南省株洲攸县黄丰桥林场杉木地位指数遥感估测的尺度进行选择,并对拟合的模型进行分析和精度评估。
研究发现,植被指数在进行5重分解重构后基本达到分解末端,在进行逐步回归分析后,只有植被指数原始数据以及一级分解重构的数据保留,其余4个尺度的数据都被剔除。拟合的模型中包括两个自变量,一是植被指数原始数据,个其一级分解重构后的数据,即一尺度数据。
最后对两个拟合模型的精度评估,检验模型的使用价值。因在林业上的地位指数是一个正整数,而拟合的模型不可能精确的算出这个数字,只能算出一个近似的数字,结合小班的具体情况来得出最终的正整数,所以本文采用建立1：1的比例分析图对模型拟合结果进行精度分析,在SPSS中建立1：1的精度分析图,结果表明两个模型拟合的相关度都在0.9以上,这说明利用遥感指数预估地位指数可以达到一定的精度,具有较高的使用价值。但由于国内少有此类研究,理论尚不成熟,有待深化。
Site quality is established in a stand on the floor of forests or other vegetation types of production potential, so the associated site quality and tree line, and high and low points. Evaluation of site quality assessment is divided into direct and indirect assessment, direct assessment method including the Stand or harvest were evaluated and assessed by stand height, indirect assessment of the status of law, including multiple index and vegetation index method. Easy to determine because of tree height, and less affected by stand density, therefore, prepared using high-index table Stand is the most commonly used assessment of forestry methods.
But the quantitative methods of site classification of forest ecosystems, there are still many difficulties, such as sample collection and field surveys are time consuming and labor-intensive, and many forest types in the lack of long-term observation plots targeting information, most of the results mainly from A point-based small-scale investigations. So looking for a simple, intuitive and accurate method to estimate the position index is very meaningful.
Remote sensing technology and the penetration of mathematical methods for the study site index provides a good platform for the large area of satellite remote sensing of forest ecosystem productivity and its spatial distribution provides an important way, from the perspective of remote sensing of vegetation coverage of most of the vegetation index value.20th century,80 years, satellite data of vegetation at home and abroad has been widely used, for example:the use of leaf area index of NDVI spatial distribution and changes of forest net primary productivity of ecosystems such as the spatial distribution. Use of vegetation index and status of the typical relationship between the estimation and analysis of Indices of forest status index and its dynamics program for the development of the Planting and management measures is of great significance. However, theory, technology and human and material resources and other conditions, researchers in the field still in its infancy, very few relevant reports.
"Geo-information map of put forward the concept of inspiration to the author, which is in accordance with the classification of certain indicators of Homologous or a group of regular arrangement to reflect the laws of earth science spatial information in digital form of maps, charts, curves or images. So I think the concept of site index map information, according to the different species, in accordance with the laws of a certain time arranged to reflect the status of the study area in digital form the law of exponential time and space maps, charts, curves or images.
However, theory, technology and human and material resources and other conditions, use of remote sensing vegetation indices reflect the site quality and its spatial and temporal dynamics research in the field still in its infancy, few relevant reports, so this model was chosen to establish a preliminary analysis The exploration and research, to improve the accuracy of model fitting, this multi-scale analysis method chosen, it is also the study of geo-information spectrum method is an important one. Taking into account data collection, and geographical constraints and other issues, this selection of You County, Hunan Toyohashi yellow cedar forest data as experimental data.
Spectral information of TM remote sensing images using well-integrated and normalized difference index (NDVI) is a plant growth status, and spatial distribution of vegetation density factor of this feature the best indication, with remote sensing software (ENVI) and geographical information systems software (GIS), Combined with MATLAB wavelet analysis and SPSS stepwise regression analysis method, You County, Zhuzhou, Hunan Toyohashi yellow fir forest site index to choose the scale of remote sensing estimation, and model fitting analysis and accuracy assessment.
Found that decomposition of vegetation index during the 5 re-decomposition of the reconstructed basically reached the end, during the stepwise regression analysis, only the raw data of vegetation index and the level decomposition and reconstruction of data retention, and the remaining four scales of data have been excluded. Fitting model includes two independent variables, namely the raw data of vegetation index, a level of decomposition and reconstruction of their data after that data of a scale.
Finally, the accuracy assessment of the two simulation models to test the model using the value. Because of the position index in forestry is a positive integer, and can not fit the model calculates the exact figure can only calculate an approximate figure, combined with the specific circumstances of small classes to come to the end of the positive integers, so this is used to establish Figure 1:1 ratio of precision of the results of model fitting analysis in SPSS to establish the accuracy of analysis of Figure 1:1, the results show that the correlation of two model fitting are above 0.9, indicating that the use of remote sensing index forecast Position index can reach a certain precision, with high use value. But because few such studies, theory is not yet mature, to be deepened.
引文
[1]陈燕,齐清文,杨桂山.地学信息图谱的基础理论探讨[J].地理科学。
[2]李慧珍.地学信息图谱的应用探讨[D].福州：福建师范大学,2006.
[3]齐清文,池天河.地学信息图谱的理论和方法[J].地理学报,2001,56(B09)：8-18.
[4]张洪岩,王钦敏,励惠国.地学信息图谱方法研究的框架[J].地球信息科学,2003(4)：101-103.
[5]陈述彭,岳天祥,励惠国.地学信息图谱研究及其应用[J].地理研究,2000,19(4)：337-343.
[6]周江评,崔功豪,张京祥,等.城镇交通网络信息图谱研究刍议[J].地理研究,2001(4)：397-406.
[7]唐中实,李小乙.澳门城市遥感信息图谱[J].地理学报,2001(B09)：38-42.
[8]周俊,徐建刚.小城镇信息图谱初探[J].地理科学,2002,22(3);324-330.
[9]陈毓芬,廖克.中国自然景观综合信息图谱研究[J].地球信息科学,2003(3)：97-102.
[10]张百平,周成虎,陈述彭.中国山地垂直带信息图谱的探讨[J].地理学报,2003(2)：163-171.
[11]齐清文.地学信息图谱的最新进展[J].测绘科学,2004,29(6)：15-23.
[12]武国胜,林惠花,廖克.福建省生态环境空间格局图谱分析[J].地球信息科学,2005(1)：116-121.
[13]廖克,秦建新,张青年.地学信息图谱与数字地球[J].地理研究,2001,20(1)：55-61.
[14]Goodchild M F, Quattrochi D A. Scale, multi-scaling, remote sensing andGIS[A]. In:Quattrochi D A, Goodchild M F, eds. Scale in Remote Sensing and GIS Ration[M]. New York:Lewis Publishers,1997:1-11
[15]李霖,吴凡.空间数据多尺度表达模型及其可视化[M].北京：科学出版社,加05：31-33
[16]李霖,吴凡著.空间数据多尺度表达模型及其可视化[M].北京：科学出版社,2005：1
[17]Lam N, Quattrochi D A. On the issues of scale, resolution, and fractral analysis in the mapping sciences[J]. Prof. Geogr.,1992,44(1):88-98
[18]贾海峰,刘雪华等.环境遥感原理与应用[M].北京：清华大学出版社,2006：44,131-140
[19]Woodcock C E, StrahlerAH. The factor of scale in remote sensing [J]. Remote Sensing of Environment,1987,21:311-332.
[20]申卫军,邻建国,林永标.空间粒度变化对景观格局分析的影响[J].生态学报,2003,23(12)：2506-2519.
[21]McNully S G, Vose J M, Swank W T. Scaling predicted pine forest hydrology and productivity across the southern United States [A]. In: Quattrochi D A, Goodchild M F, eds. Scale in Remote Sensing and GIS[C]. Boca Raton CRC Lewis Publishers,1997:197-209.
[22]Friedl M A. Exanining the effects of sensor resolution and sub-pixel heterogeneity of spectral vegetation inidices:Implications for bilphysical modeling[A]. In:QuattrochiD A, Goodchild M F, eds. Scale in Remote Sensing and GIS[C]. Boca Raton CRC Lewis Publishers,1997:113-140.
[23]GP. patil, J. Balbus, et al.Multiscale advanced raster map analysis system:Definition, Design and development[J]. Environmental and Ecological Statisties,2004,11:113-138
[24]王家耀,成毅.空间数据的多尺度特征与自动综合[J].海洋测绘,2004,24(4)：1-3
[25]彭晓鹃,邓孺孺等.遥感尺度转换研究进展[J].地理与地理信息科学,2004,20(5)：6-10
[26]Tuner M G, Gardner R H,0'Neill R V. landscape Eeology in Theory and Practice[M].New York:Springer-Verlag,2001:22-34
[27]Farina A. Principles and Methods in Landscape Ecology [M]. London: Chapman & Hall,1998:35-49
[28]David Peterson, V.Thomas Parker. Ecological scale:Theory and Appl ication[M]. New York:Columbia University Press,1998:5-16
[29]吴凡.地理空间数据的多尺度处理和表示研究[M].武汉大学博士学位论文,2002：61-63
[30]Wu J. Effects of changing scale on landscape pattern analysis:Scaling relations[J]. Landscape Ecology,2004,19 (2):125-138
[31]张彤,蔡永立.生态学研究中的尺度问题[J].生态科学,2004,23(2)：175-178
[32]张娜.生态学中的尺度问题：内涵与分析方法[J].生态学报,2006,26(7)：T2340-2355T
[33]吕一河,傅伯杰.生态学中的尺度及尺度转换方法[J].生态学报,2001,21(12)：2096-2105
[34]张娜.生态学中的尺度问题——尺度上推[J].生态学报,2007,27T(10)：T4252-4266
[35]詹昭宁.中国森林立地分类[M].北京：中国林业出版社,1989
[36]张万儒.中国森林立地[M].北京：科学出版社,1997
[37]徐化成.国外森林立地分类系统发展综述[J].世界林业研究,1988
[38]卢卫峰.基于RS与GIS的森林立地信息管理系统的研究[M].西北农林大学硕士学位论文,2004：1
[39]孟宪宇.测树学[M].中国林业出版社.1996
[40]廖显春,张新华,杨祖达,等.宜昌县防护林立地分类与评价的系统研究[J].南京林业大学学报,1998,22(2)：79-82.
[41]张金池,林文棣.苏北淤泥质海岸森林立地分类和评价方法研究[J].南京林业大学学报,1991,15(1)：23-28
[42]顾云春,李永武,杨承栋著.森林立地分类与评价的立地要素原理与方法[M].北京：科学出版社,1993
[43]徐化成.关于我国森林立地分类的发展问题[M].林业科学,1988
[44]JiangH, Strittholt James R, Frost P A, Slosser N C. The classification of late seral forests in the Pacific Northwest, USA using Landsat ETM+imagery. Remote Sensing of Environment,2004,91:320-331.
[45]Running S W, Peterson D L, Spanner M A, Teuber K B. Remote sensing of coniferous forest leaf area. Ecology,1986,67:273-276.
[46]Jensen J R. Remote sensing of the environment. Prentice Hall. New Jersey,2000.
[47]JiangH, Apps Michael J. Zhang Y L, Peng C H, Woodard Paul M. Modeling the spatial pattern of net primary productivity in Chinese forests. Ecological Modeling,1999,122:275.
[48]Chen J M. Derivation and validation of Canada-wide coarse-resolution leaf area index maps using high-resolution satellite imagery and ground measurements. Remote Sens Ennviron,2001,80:1-20.
[49]Chen J M, Cihlar J. Retrieving Leaf Area Index of Boreal Conifer Forests Using Landsat TM Images. Remote Sens Environ,1996,55:153-162.
[50]Qi J. Leaf Area Index Estimates Using Remotely Sensed Data and BRDF Models in a Semiarid Region. Remote Sens Environ,2000,73:18-30.
[51]JiangH. Population ecology of Spruce. Beijing:Chinese Forestry Press, 1992.
[52]Ma M D, Liu Y J. Comparative analysis on the environmental factors of Spruce stand through four mathematical methods. Chinese Journal of Eco-Agriculture,2006,14(1):59-65.
[53]郭达志,方涛,杜培军等.论复杂系统研究的等级结构与尺度推绎[J].中国矿业大学学报.2003,32(3)：213-217
[54]成礼智,王红霞,罗永.小波的理论与应用.科学出版社.2006.10
[55]Grossman A, Morlet J. Decomposition of Hardy Functions into Square Integable Wavelets of Constant Shape [J]. Siam. T J. Math. Anal T,1984, 15(4):723-736
[56]Y. Meyer. Pseudo differential Operators[J]. Proc. Symp. Pure. Math, 1985,43(1):215-235
[57]I Daubechies. T The wavelet transform, time-frequency localization and T signal analysis[J]. IEEE Trans.Inf. Theory,1990,36(5):961-1005
[58]赵雨耿.小波分析在图像处理中的应用研究[M].北京理工大学硕士学位论文,2003：2,11-16
[59]张德丰.MATLAB小波分析.机械工业出版社.2009.1
[60]刘贵忠,邸双亮.小波分析极其应用.西安电子科技大学出版社.1992
[61]熊正丰.金融时间序列分形维估计的小波方法[J].系统工程理论与实践,2002,12：48-54
[62]陈建裕,郭德方.基于连续小波变换的NOAA影像尺度分析[J].中国图象图形学报,2004,9(7)：837-840
[63]Nakken M. Wavelet a Talysis of rainfall-runoff variability isolating climatic from anthropogenic Patterns[J]. Environ. Model Software,1999, 14(4):283-295
[64]秦前清,杨宗凯.实用小波分析.西安电子科技大学出版社1992
[65]唐远炎,王玲.小波分析与文本文字识别.科学出版社.2004.5
[66]李双成,高伟明,周巧富等.基于小波变换的NDVI与地形因子多尺度空间相关分析[J].生态学报,2006,26(12)：4198-4203
[67]张彩霞,杨勤科,段建军.高分辨率数字高程模型的构建方法[J].水利学报,2006,37(8)：1008-1014.
[68]郝振纯,池宸星,王玲,等.DEM空间分辨率的初步分析[J].地球科学进展,2005,20(5)：499-504.
[69]Allaier. G, Homogeniaztion and two-scale convegrence. SIAM Jounral on Mathematic Analysis,1992,23(6):1482-1518.
[70]Weinna E, Bjorn Engquist. The heterogeneous multi-scale method. Communications in Mathematical Sciences,2003,1(1):87-132
[71]Jun-zhi. Cui, T. M. Shin, The two-scale analysis method for bodies with small Periodic configurations. Struct Engineering and Mechanics,1999, 7(6):601-614
[72]牟凤云,朱博勤,贺华中等.基于小波变换的多源遥感数据融合方法研究[J].国土资源遥感,2003(4)：30-34
[73]强赞霞,彭嘉雄,王洪群.基于小波变换局部方差的遥感图像融合[J].华中科技大学学报(自然科学版),2003,31(6)：89-91
[74]薛天民,张玮.基于小波变换的RS与TM遥感卫星影像融合[J].中国农业大学学报,2003,25(3)：47-50
[75]易美华.小波变换在遥感图像压缩中的应用及MATLAB实现[J].T物探化探计算技术,2003,25(3)：270-273
[76]黄玲.基于自适应矢量化的遥感图像压缩方法研究[J].信息技术,2004,28(6)：1-4
[77]裴克明,郑成勇.T基于纹理能量分析的自适应图像融合方法[J].红外,2004,4：10-13
[78]黄昕,张良培,李翠琳.基于小波变换的影像纹理特征提取试验[J].测绘信 息与工程,2005,30(6)：7-9
[79]孟帆,叶正麟,潘璐璐.小波域多尺度TARMA模型的纹理分割方法[J].计算机工程与应用,2005,29：45-47
[80]王贤敏,关泽群.小波在遥感图像分析中的应用综述[J].遥感技术与应用,2002,17(5)：276283
[81]吴纪桃,王桥.复杂地貌形态多比例尺表达的二维小波分析研究[J].遥感学报,2003,7(2)：93-97
[82]Brenner C. Building reconstruction from images and laser scanning [J]. International Journal of Applied Earth Observation and Geoinformation,2005,6(3-4):187-198
[83]Civco D L. Topographic Normalization of Landsat Thematic Mapper Digital Imagery
[84]陈燕,齐清文,杨桂山.地学信息图谱时空维的诠释与应用[J].地球科学进展,2006,21(1)：10-13.(ciVcO)
[85]孟宪宇.测树学[M].北京：中国林业出版社,1996：97-107.
[86]张明锋.生态坏境综合信息图谱的空间信息模型及其可视化分析[M].福建师范大学博士学位论文,2007
[2]李慧珍.地学信息图谱的应用探讨[D].福州：福建师范大学,2006.
[3]齐清文,池天河.地学信息图谱的理论和方法[J].地理学报,2001,56(B09)：8-18.
[4]张洪岩,王钦敏,励惠国.地学信息图谱方法研究的框架[J].地球信息科学,2003(4)：101-103.
[5]陈述彭,岳天祥,励惠国.地学信息图谱研究及其应用[J].地理研究,2000,19(4)：337-343.
[6]周江评,崔功豪,张京祥,等.城镇交通网络信息图谱研究刍议[J].地理研究,2001(4)：397-406.
[7]唐中实,李小乙.澳门城市遥感信息图谱[J].地理学报,2001(B09)：38-42.
[8]周俊,徐建刚.小城镇信息图谱初探[J].地理科学,2002,22(3);324-330.
[9]陈毓芬,廖克.中国自然景观综合信息图谱研究[J].地球信息科学,2003(3)：97-102.
[10]张百平,周成虎,陈述彭.中国山地垂直带信息图谱的探讨[J].地理学报,2003(2)：163-171.
[11]齐清文.地学信息图谱的最新进展[J].测绘科学,2004,29(6)：15-23.
[12]武国胜,林惠花,廖克.福建省生态环境空间格局图谱分析[J].地球信息科学,2005(1)：116-121.
[13]廖克,秦建新,张青年.地学信息图谱与数字地球[J].地理研究,2001,20(1)：55-61.
[14]Goodchild M F, Quattrochi D A. Scale, multi-scaling, remote sensing andGIS[A]. In:Quattrochi D A, Goodchild M F, eds. Scale in Remote Sensing and GIS Ration[M]. New York:Lewis Publishers,1997:1-11
[15]李霖,吴凡.空间数据多尺度表达模型及其可视化[M].北京：科学出版社,加05：31-33
[16]李霖,吴凡著.空间数据多尺度表达模型及其可视化[M].北京：科学出版社,2005：1
[17]Lam N, Quattrochi D A. On the issues of scale, resolution, and fractral analysis in the mapping sciences[J]. Prof. Geogr.,1992,44(1):88-98
[18]贾海峰,刘雪华等.环境遥感原理与应用[M].北京：清华大学出版社,2006：44,131-140
[19]Woodcock C E, StrahlerAH. The factor of scale in remote sensing [J]. Remote Sensing of Environment,1987,21:311-332.
[20]申卫军,邻建国,林永标.空间粒度变化对景观格局分析的影响[J].生态学报,2003,23(12)：2506-2519.
[21]McNully S G, Vose J M, Swank W T. Scaling predicted pine forest hydrology and productivity across the southern United States [A]. In: Quattrochi D A, Goodchild M F, eds. Scale in Remote Sensing and GIS[C]. Boca Raton CRC Lewis Publishers,1997:197-209.
[22]Friedl M A. Exanining the effects of sensor resolution and sub-pixel heterogeneity of spectral vegetation inidices:Implications for bilphysical modeling[A]. In:QuattrochiD A, Goodchild M F, eds. Scale in Remote Sensing and GIS[C]. Boca Raton CRC Lewis Publishers,1997:113-140.
[23]GP. patil, J. Balbus, et al.Multiscale advanced raster map analysis system:Definition, Design and development[J]. Environmental and Ecological Statisties,2004,11:113-138
[24]王家耀,成毅.空间数据的多尺度特征与自动综合[J].海洋测绘,2004,24(4)：1-3
[25]彭晓鹃,邓孺孺等.遥感尺度转换研究进展[J].地理与地理信息科学,2004,20(5)：6-10
[26]Tuner M G, Gardner R H,0'Neill R V. landscape Eeology in Theory and Practice[M].New York:Springer-Verlag,2001:22-34
[27]Farina A. Principles and Methods in Landscape Ecology [M]. London: Chapman & Hall,1998:35-49
[28]David Peterson, V.Thomas Parker. Ecological scale:Theory and Appl ication[M]. New York:Columbia University Press,1998:5-16
[29]吴凡.地理空间数据的多尺度处理和表示研究[M].武汉大学博士学位论文,2002：61-63
[30]Wu J. Effects of changing scale on landscape pattern analysis:Scaling relations[J]. Landscape Ecology,2004,19 (2):125-138
[31]张彤,蔡永立.生态学研究中的尺度问题[J].生态科学,2004,23(2)：175-178
[32]张娜.生态学中的尺度问题：内涵与分析方法[J].生态学报,2006,26(7)：T2340-2355T
[33]吕一河,傅伯杰.生态学中的尺度及尺度转换方法[J].生态学报,2001,21(12)：2096-2105
[34]张娜.生态学中的尺度问题——尺度上推[J].生态学报,2007,27T(10)：T4252-4266
[35]詹昭宁.中国森林立地分类[M].北京：中国林业出版社,1989
[36]张万儒.中国森林立地[M].北京：科学出版社,1997
[37]徐化成.国外森林立地分类系统发展综述[J].世界林业研究,1988
[38]卢卫峰.基于RS与GIS的森林立地信息管理系统的研究[M].西北农林大学硕士学位论文,2004：1
[39]孟宪宇.测树学[M].中国林业出版社.1996
[40]廖显春,张新华,杨祖达,等.宜昌县防护林立地分类与评价的系统研究[J].南京林业大学学报,1998,22(2)：79-82.
[41]张金池,林文棣.苏北淤泥质海岸森林立地分类和评价方法研究[J].南京林业大学学报,1991,15(1)：23-28
[42]顾云春,李永武,杨承栋著.森林立地分类与评价的立地要素原理与方法[M].北京：科学出版社,1993
[43]徐化成.关于我国森林立地分类的发展问题[M].林业科学,1988
[44]JiangH, Strittholt James R, Frost P A, Slosser N C. The classification of late seral forests in the Pacific Northwest, USA using Landsat ETM+imagery. Remote Sensing of Environment,2004,91:320-331.
[45]Running S W, Peterson D L, Spanner M A, Teuber K B. Remote sensing of coniferous forest leaf area. Ecology,1986,67:273-276.
[46]Jensen J R. Remote sensing of the environment. Prentice Hall. New Jersey,2000.
[47]JiangH, Apps Michael J. Zhang Y L, Peng C H, Woodard Paul M. Modeling the spatial pattern of net primary productivity in Chinese forests. Ecological Modeling,1999,122:275.
[48]Chen J M. Derivation and validation of Canada-wide coarse-resolution leaf area index maps using high-resolution satellite imagery and ground measurements. Remote Sens Ennviron,2001,80:1-20.
[49]Chen J M, Cihlar J. Retrieving Leaf Area Index of Boreal Conifer Forests Using Landsat TM Images. Remote Sens Environ,1996,55:153-162.
[50]Qi J. Leaf Area Index Estimates Using Remotely Sensed Data and BRDF Models in a Semiarid Region. Remote Sens Environ,2000,73:18-30.
[51]JiangH. Population ecology of Spruce. Beijing:Chinese Forestry Press, 1992.
[52]Ma M D, Liu Y J. Comparative analysis on the environmental factors of Spruce stand through four mathematical methods. Chinese Journal of Eco-Agriculture,2006,14(1):59-65.
[53]郭达志,方涛,杜培军等.论复杂系统研究的等级结构与尺度推绎[J].中国矿业大学学报.2003,32(3)：213-217
[54]成礼智,王红霞,罗永.小波的理论与应用.科学出版社.2006.10
[55]Grossman A, Morlet J. Decomposition of Hardy Functions into Square Integable Wavelets of Constant Shape [J]. Siam. T J. Math. Anal T,1984, 15(4):723-736
[56]Y. Meyer. Pseudo differential Operators[J]. Proc. Symp. Pure. Math, 1985,43(1):215-235
[57]I Daubechies. T The wavelet transform, time-frequency localization and T signal analysis[J]. IEEE Trans.Inf. Theory,1990,36(5):961-1005
[58]赵雨耿.小波分析在图像处理中的应用研究[M].北京理工大学硕士学位论文,2003：2,11-16
[59]张德丰.MATLAB小波分析.机械工业出版社.2009.1
[60]刘贵忠,邸双亮.小波分析极其应用.西安电子科技大学出版社.1992
[61]熊正丰.金融时间序列分形维估计的小波方法[J].系统工程理论与实践,2002,12：48-54
[62]陈建裕,郭德方.基于连续小波变换的NOAA影像尺度分析[J].中国图象图形学报,2004,9(7)：837-840
[63]Nakken M. Wavelet a Talysis of rainfall-runoff variability isolating climatic from anthropogenic Patterns[J]. Environ. Model Software,1999, 14(4):283-295
[64]秦前清,杨宗凯.实用小波分析.西安电子科技大学出版社1992
[65]唐远炎,王玲.小波分析与文本文字识别.科学出版社.2004.5
[66]李双成,高伟明,周巧富等.基于小波变换的NDVI与地形因子多尺度空间相关分析[J].生态学报,2006,26(12)：4198-4203
[67]张彩霞,杨勤科,段建军.高分辨率数字高程模型的构建方法[J].水利学报,2006,37(8)：1008-1014.
[68]郝振纯,池宸星,王玲,等.DEM空间分辨率的初步分析[J].地球科学进展,2005,20(5)：499-504.
[69]Allaier. G, Homogeniaztion and two-scale convegrence. SIAM Jounral on Mathematic Analysis,1992,23(6):1482-1518.
[70]Weinna E, Bjorn Engquist. The heterogeneous multi-scale method. Communications in Mathematical Sciences,2003,1(1):87-132
[71]Jun-zhi. Cui, T. M. Shin, The two-scale analysis method for bodies with small Periodic configurations. Struct Engineering and Mechanics,1999, 7(6):601-614
[72]牟凤云,朱博勤,贺华中等.基于小波变换的多源遥感数据融合方法研究[J].国土资源遥感,2003(4)：30-34
[73]强赞霞,彭嘉雄,王洪群.基于小波变换局部方差的遥感图像融合[J].华中科技大学学报(自然科学版),2003,31(6)：89-91
[74]薛天民,张玮.基于小波变换的RS与TM遥感卫星影像融合[J].中国农业大学学报,2003,25(3)：47-50
[75]易美华.小波变换在遥感图像压缩中的应用及MATLAB实现[J].T物探化探计算技术,2003,25(3)：270-273
[76]黄玲.基于自适应矢量化的遥感图像压缩方法研究[J].信息技术,2004,28(6)：1-4
[77]裴克明,郑成勇.T基于纹理能量分析的自适应图像融合方法[J].红外,2004,4：10-13
[78]黄昕,张良培,李翠琳.基于小波变换的影像纹理特征提取试验[J].测绘信 息与工程,2005,30(6)：7-9
[79]孟帆,叶正麟,潘璐璐.小波域多尺度TARMA模型的纹理分割方法[J].计算机工程与应用,2005,29：45-47
[80]王贤敏,关泽群.小波在遥感图像分析中的应用综述[J].遥感技术与应用,2002,17(5)：276283
[81]吴纪桃,王桥.复杂地貌形态多比例尺表达的二维小波分析研究[J].遥感学报,2003,7(2)：93-97
[82]Brenner C. Building reconstruction from images and laser scanning [J]. International Journal of Applied Earth Observation and Geoinformation,2005,6(3-4):187-198
[83]Civco D L. Topographic Normalization of Landsat Thematic Mapper Digital Imagery
[84]陈燕,齐清文,杨桂山.地学信息图谱时空维的诠释与应用[J].地球科学进展,2006,21(1)：10-13.(ciVcO)
[85]孟宪宇.测树学[M].北京：中国林业出版社,1996：97-107.
[86]张明锋.生态坏境综合信息图谱的空间信息模型及其可视化分析[M].福建师范大学博士学位论文,2007