图像挖掘技术在林火影像中的应用

提出了林火遥感图像挖掘的概念,描述了图像挖掘技术在林火遥感图像上的具体应用及实现,针对林火部门每天产生并存储的有价值的林火遥感图像,结合林火上所特有的领域知识,提出了一种通过对林火遥感图像进行图像挖掘,可以将林火遥感图像分成有火或无火两类图像的方法。该方法的实现对帮助林火工作人员及早发现火情,做出火灾预测并采取相应措施具有重大的意义。     

Understanding large-scale deforestation in southern Jinotega, Nicaragua from 1978 to 1999 through the examination of changes in land use and land cover

Nicaragua, home to the largest remaining extent of rainforest in Central America (total surface area) and to a significant indigenous population, has lost approximately half of its forest cover since 1950. This major and rapid loss of forest cover has been explained as the consequence of an eastward moving agricultural frontier that cuts through the region of Jinotega. If the current deforestation rate continues, the country could lose its remaining forest cover over the course of the next two decades; therefore, it is essential that the dynamics and relationships of land-use and land-cover change (LUCC) in this region are understood. To examine LUCC in Nicaragua over time, Landsat imagery from the southern portion of the region of Jinotega, taken in 1978, 1987, and 1999 was utilized. A remote-sensing method, supervised classification, which allows for the grouping of spectrally similar values for each year, followed by an image change detection analysis (postclassification comparison) was conducted. Groundtruthing (field validation) was conducted in 2006 to validate the data, which yielded increasing overall accuracy rates of 71.68% for 1978, 82.35% for 1987, and 84.38% for 1999. The classification and change detection results showed that if the agricultural cultivation overtook this region, it happened before 1978. Therefore, the possibility that either deforestation did not actually occur along an agricultural frontier or that it was located further east exists; this would be an interesting subject for future studies. There was, however, clear evidence of increased forest cover from 1987 to 1999 near the urban center, correlating with the enforced reforestation law in the city of Jinotega.     

利用混合水体指数模型(CIWI)提取滨海湿地水体的信息

根据对滨海湿地区域光谱特性分析和背景信息的反射特点,对主要的水体信息提取方法进行比较分析,将混合水体指数模型（CIWI）应用于TM遥感影像,针对滨海湿地区域水体信息进行提取实验,并与近年得到广泛应用和承认的水体指数模型：归一化差异水体指数（NDWI）、改进的归一化差异水体指数（MNDWI）等进行比较。实验证明CIWI的区分度最高,水体与植被、水体与城镇、植被与城镇之间的区分度分别为505%、292%与267%;水体、植被、城镇像元之间的差异比NDWI和MNDWI增大,减少了混淆信息的影响,对江苏盐城滨海湿地研究区的水体信息提取精度达到949%,实现了滨海湿地区域水体信息的准确快速提取。     

杭州城市热岛空间分布及缓减对策

以夏季的杭州市为例,利用Landsat 5 TM热红外遥感数据提取并反演了杭州市地表的覆盖类型与地表温度,通过热场分级分析和剖面分析的方法,对杭州城市尺度下热岛的空间分布特征进行了分析,并揭示了杭州热岛的分布格局与城市用地类型、功能布局之间的空间对应关系,从城市用地类型的演化、城市空间格局的改变、城市用地布局的合理性以及人为热排放等方面剖析了杭州城市热岛的产生与影响机理.最后,从城市规划视角提出了合理控制城市用地扩展、构建开敞的城市生态空间、优化城市用地功能布局以及开展城市下垫面规划设计等城市热岛缓减对策,供未来杭州城市热环境改善工作参考.     

城市复杂地表TM温度反演及其与MODIS产品的比较

在利用决策树模型将城市复杂地表分为植被、建筑物、水体、裸土等4类进行地表比辐射率估算和利用MODIS近红外数据反演大气水分含量的基础上,采用Jimenez-Munoz单通道算法进行了城市复杂地表TM温度的反演,并将反演结果与EOS MODIS地表温度产品进行了比较分析.结果表明:(1)基于地表信息分类提取后的地表比辐射率和MODIS反演的大气水汽含量得到的地表温度接近于实际状况;(2)反演结果与EOS MODIS地表温度产品对比发现Landsat TIM的陆表温度均值和标准差高于EOS MODIS的陆表温度产品.     

嫩江下游土地沙化的时空变化——基于MODIS数据的遥感分析

以2000,2004和2008年的MODIS数据为主要数据源,以嫩江流域下游为研究对象,在Arc-GIS9.0和ERDAS IMAGE9.0等软件支持下,建立沙化模型,并对沙化进行了信息提取与分析;借助ArcGIS强大的空间分析功能,完成了土地沙漠化的定量分析及其时空变化规律,在此基础上进一步探讨了土地沙化的过程、机理和成因.结果表明:研究区域沙漠化土地以轻度和中度沙漠化土地为主;重度和极重度沙漠化土地的年增长率较大,水域的年减少率最大;重度、极重度沙漠化土地主要向较低等级的沙漠化土地和水域转变,呈现一定的逆转移.     

滑坡灾害遥感解译不确定性研究

滑坡灾害遥感解译是遥感技术在滑坡灾害应用中的关键技术.研究表明,滑坡易发性和危险性区划图的误差主要来源于滑坡编目,其中大部分的误差则是源于滑坡编目时航卫片的解译,尤其是小比例尺航卫片的解译.从滑坡灾害解译、解译标志、解译过程等论述了滑坡灾害遥感解译,探讨了影响滑坡灾害遥感解译的不确定性因素,为减少滑坡编目和区划的误差提供参考.     

Using LiDAR Data Analysis to Estimate Changes in Insolation Under Large‐Scale Riparian Deforestation1

Greenberg, Jonathan Asher, Erin L. Hestir, David Riano, George J. Scheer, and Susan L. Ustin, 2012. Using LiDAR Data Analysis to Estimate Changes in Insolation Under Large‐Scale Riparian Deforestation. Journal of the American Water Resources Association (JAWRA) 48(5): 939‐948. DOI: 10.1111/j.1752‐1688.2012.00664.x Abstract: Riparian vegetation provides shade from insolation to stream channels. A consequence of removing vegetation may be an increase in insolation that can increase water temperatures and negatively impact ecosystem health. Although the mechanisms of riparian shading are well understood, spatially explicit, mechanistic models of shading have been limited by the data requirements of precisely describing the three‐dimensional structure of a riparian corridor. Remotely acquired, high spatial resolution LiDAR data provide detailed three‐dimensional vegetation structure and terrain topography over large regions. By parameterizing solar radiation models that incorporate terrain shadowing with LiDAR data, we can produce spatially explicit estimates of insolation. As a case study, we modeled the relative change in insolation on channels in the Sacramento‐San Joaquin River Delta under current conditions and under a hypothesized deforested Delta using classified LiDAR, rasterized at a 1‐m resolution. Our results suggest that the removal of levee vegetation could result in a 9% increase in solar radiation incident on Delta waters, and may lead to water temperature increases. General, coarse‐scale channel characteristics (reach width, azimuth, levee vegetation cover, and height) only accounted for 72% of the variation in the insolation. This indicates that the detailed information derived from LiDAR data has greater explanatory power than coarser reach‐scale metrics often used for insolation estimates.     

边远少数民族贫困地区民族村寨旅游开发研究——以贵州黔东南西江苗寨为例

以贵州省黔东南州西江苗寨为例,通过田野调查了解西江苗寨的旅游发展历程,在分析旅游开发对西江苗寨民族文化产生积极作用的基础之上,探析旅游快速发展背景下的西江苗寨民族文化旅游开发问题。针对西江苗寨旅游开发过程中出现的问题,提出西江苗寨民族文化保护性旅游开发对策。     

基于遥感影像的丹江口水库水域面积动态变化与原因研究

丹江口水库是南水北调中线工程的水源地,水库水域面积的动态变化研究,对保障调水有重要意义。针对丹江口水库水域面积变化较快、短期波动明显,对南水北调中线工程稳定调水有较大影响的问题,采用2000年3月~2015年10月逐月遥感影像,基于归一化差异水体指数(NDWI)提取水体,研究了丹江口水库水域面积的动态变化。结果表明:丹江口水库近16 a来月际平均水域面积均超过400 km2,2~7月份为枯水期,8月份~翌年1月份为丰水期。按季节来看,夏季至秋季水库水域面积呈增大趋势,冬季至次年春季水库水域面积呈减小趋势,同时水库水域在秋季波动最为剧烈,冬季较为平缓。自2014年12月~2015年10月,受大坝加高蓄水影响,水库水域面积均维持在550 km2以上。水库的水位涨落形成的消落区主要集中在丹江水库北部、东部和汉江水库的西部。