【作者】 赖日文；

【导师】 孟宪宇； 冯仲科；

【作者基本信息】 北京林业大学 ， 森林经理学， 2007， 博士

【摘要】 本文主要借助遥感和地理信息系统及其它集成技术，收集并处理闽江流域自然条件和社会经济状况以及相关资料信息，对福建省闽江流域森林生产力空间分异及分区状况、森林立地类型的划分、森林资源的可利用数量、自然品质等方面进行专题分析研究，在此基础上对闽江流域森林资源利用的适宜度进行评价和调整，为闽江流域森林资源合理利用提供决策和技术支持。主要研究结果：(1)对森林类型信息提取通过建立专家分类知识库，选取适当的光谱阈值，结合分层分类法对闽江流域森林类型进行划分。对光谱值比较接近的森林类型采用监督分类方法实现专题信息的提取，对局部区域采用人机交互目视解译，以提高分类精度。(2)本研究以反距离权重插值法(IDW)对降雨线的进行内插处理，获取闽江流域各季度的降雨量，温度因子栅格空间数据库采用泛克里格插值获得，经模外检验精度较高，总体精度都达到95％以上。(3)选用中式通用土壤流失方程式进行闽江流域水土流失定量评价，运用ArcGIS和ERDAS的空间数据管理和分析功能分别生成降水侵蚀力R因子栅格专题图、土壤可蚀性K因子栅格专题图、地形因子LS因子栅格专题图、植被管理C因子专题图，侵蚀防治措施P因子专题图，再通过ERDAS将各因子图叠加连乘，获得闽江流域的土壤侵蚀模型，生成闽江流域土壤侵蚀模数图。(4)依据国家及福建省的生态公益林区划标准，采用3S技术在遥感影像图上区划出了闽江流域生态公益林基础上区划出了闽江流域商品林。(5)以闽江流域林地1981～2005年平均温度、平均降水量栅格数据库及主要土壤质量因子栅格数据库为基础，建立森林气候生产力、森林生产力和林木产量栅格数据库，运用ERDAS建模功能生成闽江流域林地森林气候生产力、森林生产力、林木产量分布图。通过对各生产力空间分异及分区分析，全流域各生产力类型以高、中生产力面积占绝对优势，总体空间分布上大致呈从东南向西北略有递增的趋势，中部和西北部较高，东南和西部相对较低的状态。(6)本研究借助RS和GIS技术，根据森林立地类型分类的原则，选择土壤有机质含量、土层厚度、坡向和土壤侵蚀模数作为闽江流域森林立地类型划分的主导因子，通过ERDAS建模功能实现主导因子的图形空间叠加分析获得多立地因子综合叠加类型图，利用计算机实现立地类型的自动分类及编号。分类结果闽江流域共划分为88种立地类型，其中以极丰富有机质中土层阴坡微度侵蚀土类型面积占明显优势，其次依次为极丰富有机质中土层阳坡微度侵蚀土、丰富有机质中土层阴坡微度侵蚀土和丰富有机质中土层阳坡微度侵蚀土，一般有机质薄／厚土层阴／阳坡强度侵蚀土立地类型为最少，面积均在20hm~2以下。(7)以少量地面样地对应的遥感和GIS信息为基础，建立以象元为单位的含定性变量的蓄积估测方程，可用于预报森林总蓄积。(8)本研究采用主成分分析、相关分析和因子分析结合理论分析有效地筛选了包含地面曲率、林地利用率和年均气温等在内的39个森林资源利用适宜度评价指标，构建了评价指标体系。(9)本研究选用灰色关联度分析确定森林资源利用适宜度评价指标权重，按加权求和方法计算各评价单元的总分值通过遥感处理软件建模功能实现闽江流域森林资源利用适宜度评价得分，经模糊聚类获得闽江流域森林资源利用适宜度评价等级分布图。更多还原

【Abstract】 By collecting and treating the natural, social and economic conditions as well as other related information based on RS, GIS and other integrated technology. Analyzing the variance and division status of forest productivity, division of forest site, available quantities and natural quality of forest resources in Minjiang watershed. On the above researches, the evaluation was launched to the utilization suitability and adjusting of forest resources, which provided the stagy and technology about reasonably utilizing the forest resources in Minjiang watershed. The main research results were as followed:(1) Extracting the thematic information of forest types by established the knowledge base of expert classification, select appropriate spectral threshold and combine the stratified classification to plot the forest types in Minjiang watershed. Extracting thematic information was realized for the forest types with adjacent spectrum by the supervised classification and the interactive visualization interpretation was adopted to improve the classified precision for the partial region.(2) In this study adopting the inverse distance weighted (IDW) to deal with the rain gauge by interpolation, obtaining the rain capacity in every quarter in Minjiang watershed. Kriging interpolation was employed to get the spatial date base of temperature factor and the overall precision reached 95% by external module examination.(3) The soil erosion module was obtained and the modulus figure was generated in Minjiang watershed by choosing universal water and soil loss equation to quantitatively evaluate the soil and water losses in Minjiang watershed, applying the spatial data management and analysis function of ARCGIS and ERDAS to respectively create the special grid figure of R factor of erosive force of rainfall, K. factor of soil erosion and LS factor of topographic feature, the spatial figure of R factor, C factor of plant management, P factor of erosive prevention measure, as well as using ERDAS to superimposing multiplication continuously.(4) According to the regionalizing standard of ecological forest of nation and Fujian province, on the whole the ecological forest in Minjiang watershed was regionalized based on RS image by employing geomatics. Founded on the mask works of ecological forest, the commercial forest was regionalized in Minjiang watershed.(5) This study constructed the grid data base of forest climate productivity, forest productivity and forest yield founded on the many years average temperature and rainfall data in each county and city in Minjiang watershed. The distribution map of forest climate productivity, forest productivity and forest yield in Minjiang watershed was generated by ERDAS model. Through analyzed the variance and division of all productivity, the result showed the middle and high type area occupied absolute superiority. The overall spatial distribution of them presented the slightly increasing trend from southeast to northwest in general and it was higher in middle and northwest parts and relative lower in southeast and west parts.(6) With the aid of RS-GIS and according to the rule of forest site classification, choosing the soil organic matter content, soil layer thickness, aspect and soil erosion modulus as the leading factor of the forest site types classification in Minjiang watershed, sequentially, the superimposing analysis on the figure space was realized, the superimposing synthesis map of multi-site-factors was obtained and the auto-classification and number to site types were implemented with the computer by the modeling function of ERDAS soft. From the result of classification, we partitioned the Minjiang watershed into 88 site types. Among them, the area of the slightly eroded middle thickness layer soil types of shady slope with extreme rich organic matter accounted for the obvious superiority, then in turn that of the slightly eroded middle thickness layer soil types of sun slope with extreme rich organic matter, the slightly eroded middle thickness layer soil types of shady slope with rich organic matter and the slightly eroded middle thick layer soil types of sun slope with rich organic matter. Generally that was the least area in the barren organic matter, thick layer with strongly eroded soil in shady or sun slopes, the all areas were under 20hm~2.(7) Based on the RS and GIS information corresponding with a few sample plots, the estimating equations of forest stock including qualitative variable with pixel unit was set up to forecast the total stock effectively.(8) The principal components analysis, relative analysis and factor analysis were combined with theory analysis to effectively select 39 expedient degree evaluation indexes of forest resources including the land surface curve, forested land utilization and average annual temperature, which were constructed assessment index system.(9) The analysis on gray correlative degree was adopted to determine the expedient degree evaluation index weight of forest resources utilization, and the total values of each appraised unit were calculated according to the weighting summation method. For the expedient degree evaluation of forest resources in Minjiang watershed, the scores were achieved by the modeling function of RS software and the rank distribution figure were obtained by fuzzy clustering analysis.更多还原