【作者】 郭云开；

【导师】 陈正阳； 朱建军；

【作者基本信息】 中南大学 ， 大地测量学与测量工程， 2008， 博士

【摘要】 土地整理是人类文明用地的产物,是人类利用自然和改造自然的措施,是社会经济发展到一定阶段解决土地利用问题的必然选择。从目前我国开展土地开发整理的实践看,较普遍的存在重数量、轻质量和生态,开发整理规划、可行性研究、整理工程量计算、整理后的土壤质量评价等事项主要以图纸、人工为主,较少采用遥感、地理信息系统等先进高效的技术方法。对土壤的质量验收鉴定及整理后的成果管理等方面缺乏先进有效的方法,而开发整理出的耕地质量较差很难形成一定的生产能力甚至造成丢荒,这不仅没有实现真正意义上的耕地整理开发,而且造成了资源的破坏和开发资金的浪费。随着土地整理实践的开展,目前已有大量项目竣工验收进入运行阶段,这些项目决策是否正确、预期目标和效益是否实现等尚不明确,急需进行后评价。遥感(RS,Remote Sense)用于大范围、高分辨率、快速信息获取,将“RS”技术应用于土地开发整理项目信息采集、处理和分析,可以进一步规范土地开发整理规划编制,对土地开发整理工程设计与实施动态监测、工程验收、特别是评价工程效果,可以极大提高技术含量,从而提高项目质量和效率。为此,我们以湖南省国土资源厅重点资助项目“基于遥感反演的土地开发整理评价”研究课题为带动,开启了国内用遥感这一当今先进手段评价土地开发整理之先河。综合运用多种研究方法,包括静态分析与发展动态预测相结合、政策规程与实际工程需要相结合的评价体系研究、遥感定性分析与定量计算相结合、传统技术与现代新技术方法相结合等科学的综合分析方法。通过较为系统的研究主要取得了如下成果:(1)依据《中华人民共和国土地管理法》及国土资源部《土地开发整理项目验收规程》等相关法律、法规,结合湖南省土地开发整理之省情和试验区实际情况,在研究高分辨率遥感信息特点和定量遥感解译理论的基础上,提出了用定量遥感方法,按土地平整工程、农田水利工程、田间道路工程、其他工程(防风林工程等)等四个方面的指标评价土地开发整理工程的工程量。项目区实验表明,该方法速度快、效率好、精度高、劳动强度极低、反应工程完成情况全面,分类提取的指标信息与实际情况情况很吻合。(2)在研究了土壤含水量、土壤有机质含量、地表温度、植被指数及其植被状态指数和条件温度植被指数等方面用遥感直接或间接评价了土壤质量,充分说明了其具有较好的、开创性的、发掘性的应用遥感技术能够评估或评价土壤质量和土壤农作潜力的基础上。指出,目前还没有建立土地开发整理工程及其土壤质量改善等方面有机结合系统评价土地开发整理工程的评价体系,更是没有建立用遥感方法评价土地开发整理工程评价体系。本研究在前人遥感应用的基础上,从多个遥感数据源预处理、分析入手,提出了用遥感方法评价开发整理工程的工程量,采用修正土壤调整植被指数、植被覆盖度、土壤含水量、土壤潜育化指标、LBV变换等比较关键性的指标对土地开发整理工程影响土壤质量情况进行较为科学客观的评价评价。(3)通过分析遥感信息物理机理、遥感图像处理解译原理及土壤遥感信息特征和提取特点之后,首先研究了对用于整理工程量评价的高分辨率Quickbird影像,在国产软件TITAN Image 6.5进行切割(根据评价的区域进行)和二维三次卷积内插法(4个单色波段及1个全色波段)融合;其次,为了保证精度,采用了GPS实测地面控制点,以多项式为几何纠正模型,对融合后的影像进行影像空间像元位置的变换,再用双三次插值法计算变换后各像元的亮度值,即像元亮度值的重采样。计算结果表明:点位误差均小于1m,,若换算成像元,点位误差在1个像元左右,完全满足整理工程量信息提取要求;其三,提出并采用ISODATA(自组织分类)方法,有效提取项目区内不同地块的面积值,沟渠和道路等线状物的长度值。面积提取精度在97%以上,线状物长度提取精度在98%以上。利用分类图及结合现场数据,还可以有效的进行灌排系统布局合理性分析、防护林分布合理性分忻、道路网布局合理性分析。(4)分析研究了土壤含水量与土壤质量的相关关系、土壤水分的光谱特征、土壤湿度与相对反射率,探讨了作物缺水指数法、距平植被指数法、微波遥感监测土壤水分、热惯量法、数据同化法等遥感方法提取土壤含水量的适应性,提出用光学植被盖度遥感分析土壤水分,利用样本统计和相关分析相结合的方法,建立了光学植被盖度的TM模型及不同深度土壤水分与遥感植被绿度指数之间的回归模型,对试验区各个时像的TM影像进行了土壤三个不同层次的含水量信息提取和分析。同时用其模型提取了项目区IKONOS影像、中巴资源一号卫星影像三个层次的土壤含水量信息。依据提取的信息分析了项目区土地开发整理后土壤质量改善提高情况,分析成果与实地情况一致,量化了土地整理效果。(5)研究了植被指数、植被覆盖度与土壤质量的定量关系。综合分析了比值植被指数RVI、归一化植被指数NDVI、差值植被指数DVI、土壤调节植被指数SAVI、增强植被指数EVI以及植被覆盖度等与土壤质量之间的关系特征;提出了MSAVI(称为“修正型土壤调整植被指数”)模型,像元二分植被盖度模型;通过试验区多种遥感影像MSAVI指数、像元二分模型参数提取分析,证实了它们是适合评价试验区土壤质量的指标。(6)通过研究土壤潜育化与土壤质量相关关系的特性,潜育化土壤热力学特征及其遥感解译机理,指出了涝渍灾害土壤的热惯量概念模型、数值解算模型,得出了用遥感的方法研究农田涝渍灾害致使土壤潜育化理论和实际可行性。研究指出,卫星遥感影像上存在着三个重要的地物遥感特征,即地物的总辐射水平L、可见光-红外光辐射平衡B和地物辐射随波段变化的方向和速度—向量V,进而提出用遥感图像LBV数据变换法进行试验区IKONOS等遥感影像的变换与信息提取,试验和证实了所建立的LBV变换方程。成果证明试验区开发复垦之前土地荒芜、积水严重、潜育化厉害,与实际情况相同。通过本文的系统研究和项目区实际检验,证实了定量遥感能很好地应用于土地丌发整理的工程量评价和土壤质量评价,其研究成果,能为国土局以及相关部门对土地复垦项目的验收提供有力支持;充分说明了遥感反演技术用于评价土地丌发整理项目的工程量完成情况真实、可靠、可行,为遥感的应用开辟了新途径;有力证明了应用定量遥感----遥感反演监测评价土地整治质量准确、有效,动态性强、时效性高。本研究在大规模的土地开发整理工程评价中大有发展潜力。更多还原

【Abstract】 Land consolidation is a result of civilized land use of human beings, an approach to utilize and transform the nature by human, and an inevitable choice to solve the problem of land use when economy developed into a certain degree. From the point of view of the practice of land development and land consolidation in our country, there commonly exist problems of preferring quantity to quality and ecologic environment; consolidation planning, feasibility study, calculation of consolidation engineering volume, soil quality assessment after consolidation are done by hand, rarely with advanced and effective technologies and methods such as remote sense, geographic information system etc. Soil quality acceptance and result management after consolidation lack advanced and effective methods; and if the consolidated land is of inferior quality, it will not own certain production capacity and in the contrary will result in land lying waste. This kind of consolidation has not realized the real consolidation and development of cultivated land and even resulted in destruction of resources and wasting of development fund.With the development of land consolidation practice, at present, the acceptance of a great number of completed projects has entered the operational phase. It is still unclear whether these projects are correct and whether the anticipatory goals and benefits can be achieved. Hence, post-evaluation is urgently needed.Remote Sense provides quick access to vast amounts of high-resolution information. Application of "RS" in the acquisition, processing and analysis of land development information and consolidation information can further standardize the compiling of land development and consolidation plan, and greatly increase the technical contents in land development and consolidation design, implementation of dynamic monitoring, acceptance of projects, particularly, appraisal of project achievements, thus improving the quality and efficiency of the projects. Motivated by the research subject of "Evaluation of Land Development and Consolidation Based on RS", a key project sponsored by The Land and Resources Department of Hunan Province, we applied advanced method-RS to evaluate land development and consolidation for the first time in China.By comprehensively employing multi research approaches, including evaluation system study which combines static analysis with development forecasting, policy regulations with practical engineering needs, and scientific integrated analysis approach which combines RS qualitative analysis with quantitative analysis, traditional technology with modern new technology, the systematic study comes to fruition:(1) In conformity with the relevant laws and regulations, such as the Land Administration Law of the PRC and The Acceptance Codes of Land Development and Consolidation Project, combined with the actual situation of experimental areas and the land development and consolidation of Hunan, the research, based on the features of the high-resolution remote sensing information and the remote sensing interpretation system, proposes quantitative remote sensing method, which can evaluate the engineering quantity of the land development and consolidation project according to land-leveling project, water conservancy and irrigation, crisscross footpaths project and other projects (windbreak forest project etc.). The experiment in project area shows that this method has such characteristics as high speed, high efficiency, high degree of accuracy, extremely low labor intensity, etc; also it can reflect the performance of the project in a complete way. And the extracted index information of classification closely meets the actual conditions.(2) When studying soil moisture, soil organic matter content, land surface temperature, vegetation index, vegetation condition index and vegetation-temperature condition index through remote sensing, soil quality is evaluated directly or indirectly. This fact can speak volumes for the ability of inaugurated applicable remote sensing technology to evaluate or assess the quality and agriculture potential of soil. Based on this conclusion, the reserch shows that there is neither any evaluation system which can systematically assess the land development and consolidation project through the organic combination of the land development and consolidation project with the improvement of the soil quality, nor any evaluation system to evaluate land development and consolidation project by means of remote sensing. Based on the previous application of remote sensing, this research proposes engineering quantity evaluation of land development and consolidation project by using remote sense, starting with the pretreatment and analysis of multi remote sensing data sources. This research makes relatively scientific and objective evaluation on the influence of the land development and consolidation project towards the soil quality through the utilization of the relatively critical indexes such as the MSAVI, vegetation coverage, soil moisture, soil gleying index, LBV transformation.(3) After analyzing information mechanism, interpretation theory of remote sensing image processing, features and extraction characteristics of remote sensing information of soil, this research first studies the high-resolution Quickbird image, which is used to sort the evaluation of the engineering quantity, through cutting the images on the home-produced software TITAN Image 6.5 (according to the regions of evaluation) and blending the images through the 2-D cubic convolution (4 monochromatic bands and 1 panchromatic band ); second, in order to assure the accuracy, this research adopts the GPS-measured ground control points, uses the polynomial as the geometric correction model to change the images spatial pixel position of the blended images, then utilizes the bi-cubic interpolation to calculate the pixel lightness value after the change of position, namely resampling of pixel lightness value. According to the results of calculation, the position errors of points are less than 1m; after converted into pixel, the errors are about 1 pixel, so they can meet the requirements of the information extraction of consolidation engineering quantity. Third, the research proposes and adopts the method of ISODATA (self-organizing classification) to effectively extract the area value of different areas in the project area and the length of threads such as ditches and roads. The accuracy of extracted area is above 97%; the accuracy of the extracted length is above 98%. Using classification map and onsite data, rationality analysis of irrigation and drainage systems layout, rationality analysis of windbreak forest distribution and rationality analysis of road network distribution can be made.(4) The paper analyzes the correlation between soil water content and soil quality, spectral signatures of soil moisture, soil moisture and relative reflection; probes into adaptability of extracting soil moisture content by using such remote sensing methods as crop water stress index method, average vegetation index method, microwave remote sensing orientation method, thermal inertia approach and data assimilation method; puts forward the idea of analyzing soil moisture with optical vegetation coverage remote sense; builds TM model of optical vegetation coverage and regression model between soil moistures at different depths and greenness indexes of remote sensing vegetation by using the approach of combining sample statistics with correlation analysis; makes information extraction and analysis of soil moisture content in three different layers on TM image of experimental area at different time. Meanwhile, IKONOS image in project area and soil moisture content information in three different layers of CBERS-1 satellite image are extracted with the help of this model. According to the information extracted, the improvement and promotion state of soil quality after land development and consolidation are analyzed. And the results of analysis are corresponding with the actual situation, thus quantizing the soil consolidation effects.(5) This paper studies the quantitative relationship among vegetation index, vegetation coverage and soil quality ; comprehensively analyzes the relationship characteristic between soil quality and such indexes as ratio vegetation index (RVI), normalized difference vegetation index (NDVI), difference vegetation index (DVI), soil adjusted vegetation index (SAVI), enhanced vegetation index (EVI) and vegetation coverage ; establishes MSAVT (called ’modified soil adjusted vegetation index’) model and dimidiate pixel model of vegetation coverage. By analyzing MSAVI index of various remote sensing images in experimental area and dimidiate pixel model parameter, this paper testifies that these indexes are suitable for soil quality evaluation in experimental area.(6) By studying the nature of correlation between gleying process of soil and soil quality, thermodynamic characteristics of gleyed soil as well as remote interpretation mechanism, this paper puts forward thermal inertia conceptual model of water-logging disaster soil, number resolution model, and points out the theoretical and practical feasibility of using remote sense to study the phenomenon of water-logging disaster leading to soil gleying.The study points out that there are three important remote characteristics of surface features on satellite remote sensing imaging map, namely overall radiation level of surface features—L, the direction and speed of change of visible light-infrared radiation balance B and surface features radiation with band change—vector V; further proposes to adopt IKONOS and other remote sensing images transformation and information extraction in experimental area through remote sensing image LBV data transformation. The established LBV transformation equation is also testified by experiments. The serious degrees of land barrenness, standing water and gleying of soil before the reclamation in the experimental area are identical with the actual situation.Through systematic research in this paper and actual inspection in project area, it is verified that quantitative RS is well applicable for the evaluation of project quantity of land development and consolidation as well as soil quality assessment. The research findings can be forceful support of the acceptance of land reclamation project run by Land and Resources Bureau or relevant departments. This paper fully proves that applying RS retrieve technique to evaluate the performance of the engineering quantity of land development and consolidation project is true, reliable and feasible. This approach is a new way for RS application. This paper also forcefully proves that the approach of employing quantitative RS--RS retrieve monitoring to evaluate land consolidation quality is characterized by accuracy, effectiveness, strong dynamics and timeliness. This study has great development potential in the evaluation of large-scaled land development and consolidation projects.更多还原