节点文献
GIS技术支持下的海岸带遥感动态监测分析
Using Remote Sensing Dynamic Monitoring and Analysis of Coastal Zone Changes Supported by GIS Technology
【作者】 侯志华;
【导师】 吴泉源;
【作者基本信息】 山东师范大学 , 自然地理学, 2006, 硕士
【副题名】以龙口市海岸带为例
【摘要】 海岸带地处海陆之交,凭借其自身丰富的自然资源和优越的地理位置成为人类竞争和开发的重要区域,同时又由于受到强烈的海陆作用,是全球自然环境和生态最为复杂和脆弱的地域之一。近年来,随着经济的发展和对土地资源的不合理利用,海岸带资源和环境遭到极大破坏,对海岸带变化的动态监测研究成为全球变化研究的重要领域。在科技迅速发展的今天,遥感技术和地理信息系统技术为全球变化研究提供了主要数据源和技术手段,也成为海岸带变化研究的一种重要的现代化方法。本文以龙口市海岸带为研究区,在国家自然科学基金项目“山东海岸带陆海作用及人类活动的环境影响与生态效应”和“龙口市土地利用数据库更新”课题支撑下,完成了基于遥感技术和GIS技术的海岸带专题信息提取、动态变化分析,并对其变化的驱动因素进行了详细的阐述。论文的主要研究内容和研究成果有:1.本文在遥感影像海岸带专题信息增强处理基础上,完成海岸带各种信息的提取,其中,低潮线由于长期被海水淹没,难以直接提取,本文以水深遥感理论为指导,对海水深度与影像灰度进行相关回归分析,通过建立遥感水深模型,反演水深,再结合一定潮汐资料实现低潮线的提取。研究表明采用遥感定量技术能迅速地获取大量海水资料,重复获取数据的周期短,在海岸带动态变化监测中具有重要的研究价值。2.通过计算机自动识别的方法提取海岸带陆域地物信息,并采用BP神经网络分类器,将影像光谱特征、纹理结构特征和海岸地物区位指数一并加入判别准则进行识别,与传统的单一根据地物光谱特征统计模式识别的最大似然法相比,分类精度得到极大提高。3.构建1984-2004年间的龙口市海岸带动态变化信息图谱,并根据图谱从时间和空间两个角度,对海岸带岸线的进退变化和土地利用动态变化规律做了系统分析。结果表明,20年来龙口市海岸带是一个变化过程较为活跃的地域。岸线
【Abstract】 Coastal zone, located in the connection belt of sea and land, has become a keyfield of human competition and exploration owing to its abundant natural resourcesand advantageous geographical location. However, affected by the strong interactionbetween ocean and continent, it has the most complicated natural environment and themost fragile ecology. In recent years, with the rapid development of economy, theresources and environment of coastal zone have been greatly damaged due to theunreasonable utility of land resource. So the study of dynamic monitoring coastalzone changes has became a hot topic of Global Changes. Nowadays, as the rapiddevelopment of science and technology, remote sensing and geography informationsystem have offered the main data sources and technological means for studyingglobal changes, and it is also an important advanced way to studying coastal zonechanges. Taking Longkou coastal zone as a case, the paper, on the basis of RS andGIS technology, extracted the thematic information of the coastal zone, then analyzedthe changes from space and time and expatiated on driving factors on the changes indetail, supported by national natural science fund “The function of land, sea andenvironment impact and ecological effect of human activity” and the subject of “Therenewal database of land use of Longkou city ”. The major contents and researchresults are as following:1. The need of information for the research was obtained based on remote-sensingimage processing to enhance thematic information of the coastal zone. It was difficultespecially to been drew out directly for the low tide line because of being submergedby sea long time. So the study based on the water depth remote sensing theory carriedon the statistical correlation analysis on the single band or multi-band combinationdata and created the RS-fathoming model, then inversed the fathoming model to drewout the low tide line according as the synchronous tide data. It is obvious that this newmethod to obtain seawater information quickly and efficiently. So it is value for thestudy of dynamic monitoring coastal zone changes using remote-sensing technology.2. The land information of the coastal zone was extracted by computationalauto-classification. Compared with the conventional Maximum Likelihood Classifierin the process of which only the spectral information of objects is used, the ArtificialNeural Network (ANN) has been developed and applied to remote sensing dataclassification problem, which can overcome the limitation stated above. It can makeclassification by integrating RS data (spectral values, texture) with the geographicdata (DEM, slope etc.) resulting from GIS. In the paper, the need of land coverclassification was made though BP Neural Network classifier, a sort of ANN bewidely used for classification of remote image data nowadays, integrating spectralvalues with texture data and the distance from coastline. The analysis result showsthat this method can improve the classification accuracy greatly.3. Geographic Information Tupu was created to study the Longkou coastal zonechanges that included the advance and retreat of coast and the features of land-use inboth spatial and temporal from1984 to 2004. The result showed that the coastal zoneof Longkou city is an active area with changes obviously during the course of study.There were notable difference of advance and retreat in each coast section fromdifferent years. The general trend was moving to land obviously in the north coast,pushing to sea in the coast between Longkou Pot and Jimu Island, and keeping steadyrelatively in the west coast in addition to having a little advance latter. The changingextent was more remarkable in the 1990s especially in 1992 to 1998, and the coasttrended to be steady after 2000. The land use also changed greatly in the coastal zoneduring 1984 – 2004. There was a great decrease in sand land and cultivated land and aremarkable increase in water area, garden land and build-up land. The speed of theland use change in the region was up to 1.53% per year, and the sand land and waterarea changes had the highest speed in all land-use type. As to the spatial change,55.65% of the land use in the region was under the changing progress, major type ofwhich were cultivated land turned to garden land, and cultivated land and garden landturned to build-up land. But the active land-use conversion was the sand land turningto other land-use type in the region less than 500 meters distance from the coastline.4. Based on the macroscopic analysis of the driving mechanism of the coastalzone changes, it is clear that the changes of coatal zone were caused by both naturaland human factors. However, for Longkou coatal zone, human activity was the mainfactor, and the natural drove the changes more by giving a feedback to the humanactivity. Canonical Correlation Analysis was applied to identify the human drivingfactors quantitatively, and it was concluded that human activities drove the changes ofland use by the economic development, agriculture structure change, marine culture,and exploration on sand and coal resources. Nature gave a feedback to the humanactivity by means of hydrodynamic condition change at coastwise, seawater intrusionand mining subsidence.
【Key words】 coastal zone; remote-sensing image; water depth remote sensing; BP Neural Network; Geographic Information Tupu; driving force;
- 【网络出版投稿人】 山东师范大学 【网络出版年期】2006年 09期
- 【分类号】P208;P237
- 【被引频次】35
- 【下载频次】1602