多源数据协同下的林地提取研究

发布时间：2018-01-09 05:08

本文关键词：多源数据协同下的林地提取研究　出处：《四川师范大学》2017年硕士论文　论文类型：学位论文

【摘要】：本文通过开展多源数据协同下的林地提取研究,达到监测森林动态变化的目的。本研究对维护和提升区域生态系统服务功能具有重要意义,是提高林地监管能力、开展生态空间规划、兑现生态效益补偿、加强林地保护利用管理和科学预测植被长势的重要基础和科学支撑,并为区域经济社会发展、城乡建设、土地利用及生态环境保护等奠定基础。本文主要探讨了以下内容:1、利用GF1、GF2、ZY3三种高分数据源开展基于知识发现的2015年林地提取研究。首先,探索高分影像的光谱、遥感派生、纹理和形状特征。第二,依据特征知识,建立知识库进行林地资源提取与目视解译修正。最后,采用google earth影像和实地调查数据进行精度验证。结果表明GF1、GF2、ZY3和总体提取结果的精度分别达到95.1%、94.5%、91.4%、92%,Kappa系数分别为0.9、0.89、0.83、0.84。说明本次提取结果较好,并可以用于接下来的研究。2、开展生物量估算研究。首先,在提取结果的基础之上,结合2015年林地变更数据中的公顷蓄积量和主要树种信息,建立生物量与蓄积量之间的线性回归模型,开展研究区内的生物量估算工作。第二,通过DEM、坡度、坡向数据对生物量开展空间分析研究。结果表明在中起伏山地即海拔500-1000m、缓的即坡度15-25o、东南即坡向112.5-157.5o范围内分布的林地资源生物量较多。3、开展2011—2015年林地动态变化分析。首先,将2011年林保数据、2013年国土数据和本次2015年提取结果分为三个时段(2011—2013、2013—2015、2011—2015)进行三种数据源的衔接分析工作。分析发现2011—2013年林地面积减少115.53hm2,2013—2015年增加8701.17hm2,2011—2015年增加7546.02hm2。变化原因主要分成人为和自然因素,变化的面积基本符合实际。第二,通过DEM、坡度、坡向数据开展地形因子对植被分布影响分析。结果表明,区内林地在海拔200-1000m、坡度8-25o、坡向112.5-202.5o的范围内分布面积较多。
[Abstract]:In this paper, the purpose of monitoring forest dynamic change is achieved through the study of forest land extraction under the cooperation of multi-source data. This study is of great significance to maintain and enhance the regional ecosystem service function. It is an important foundation and scientific support for improving the ability of forestland supervision, carrying out ecological space planning, realizing ecological benefit compensation, strengthening forestland conservation and utilization management and scientifically forecasting vegetation growth, and also for the development of regional economy and society. Urban and rural construction, land use and ecological environment protection laid the foundation. This paper mainly discusses the following content: 1, using GF1 and GF2. ZY3 three kinds of high score data sources based on knowledge discovery in 2015 woodland extraction research. First, explore the high score image spectrum, remote sensing derivation, texture and shape features. Second, based on feature knowledge. Finally, google earth images and field survey data were used to verify the accuracy. The results showed that GF1 and GF2. The accuracy of ZY3 and total extraction was 95.1 and 91.4, respectively, and the coefficient of Kappa was 0.90.89 and 0.83, respectively. 0.84.The result of this extraction is good, and can be used in the following research. 2, to carry out biomass estimation research. First, on the basis of the extraction results. Based on the data of forest land change in 2015, the linear regression model between biomass and biomass was established to estimate the biomass in the study area. Based on the data of DEM, slope and direction, the spatial analysis of biomass was carried out. The results showed that the gradient was 15-25o in the middle undulating mountain area, that is, 500-1000m above sea level. The biomass of forestland resources distributed in the range of 112.5-157.5o in southeastern China was more than .3. the dynamic change analysis of forest land was carried out from 2011 to 2015. Firstly, the data of forest protection in 2011 were analyzed. The data of 2013 and 2015 are divided into three periods: 2011-2013 / 2013-2015. The analysis shows that the forest land area decreased 115.53 hm ~ 2 in 2011-2013. An increase of 8701.17hm2hm2in 2013-2015 and an increase of 7546.02hm2between 2011-2015. The main causes of change are adult and natural factors. The change of the area is basically in line with the reality. Second, through DEM, slope, slope data to analyze the impact of terrain factors on the vegetation distribution. The results show that the woodland in the area is 200-1000m above sea level. The slope is 8 to 25 o. and the distribution area is more in the range of 112.5-202.5 o.
【学位授予单位】：四川师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S771

【参考文献】