基于GIS的黄土高原南部地区土地资源利用与优化配置研究

【作者】 李强；

【导师】 任志远；

【作者基本信息】 陕西师范大学 ， 地图学与地理信息系统， 2012， 博士

【摘要】 土地是一种非常特殊的资源,它既是自然环境要素发生和变化的重要载体,也是人类自身生存和繁衍的物质保障。同时,作为具有自然属性和社会经济属性的共同体,土地资源在经济建设与社会发展中也起着非常重要的基础性、战略性作用。土地资源利用是土地在人类活动的持续或周期性干预下,进行自然再生产和经济再生产的复杂的社会经济过程,而土地资源优化配置是缓解人地矛盾、实现土地资源可持续利用的重要手段。本文在黄土高原综合自然指数分区研究的基础上,以黄土高原南部地区及其内部典型区为研究对象,通过野外实地调查,采用遥感(RS)和地理信息系统(GIS)集成技术与数理模型、空间统计分析等方法,利用黄土高原南部地区1980年、2000年和2010年30年跨度的TM影像遥感解译结果,在不同的时空尺度上分析了黄土高原南部地区及其典型区土地利用的时空差异性特征；并将空间抽样方法与Binary Logistic回归模型相结合,在分析不同研究区土地利用特征尺度的基础上,对黄十高原南部地区及四个典型区不同土地利用类型与其影响因素之间的关系作了定量分析；进而根据不同类别典型区土地利用存在的问题,构建土地承载力评测模型,从定性与定量的角度对每个典型区的土地承载力进行动态测评；最后针对基于流域单元与基于都市经济区的两类不同典型区,提出将灰色线性规划模型(GLP)分别与CLUE-S模型及CA-Markov模型相结合的两种方案,在对不同典型区土地利用进行数量控制与结构优化的基础上,利用两种方案将优化结果配置到空间单元上。研究表明：(1)黄土高原南部地区土地利用类型比较完全,耕地比重最大,各个时期所占比重均大于40%,草地、林地次之。从总体上看,黄土高原南部地区三十年来(1980-2010年)各十地利用类型的面积都有所变化,其中耕地和未利用地呈现减少趋势,其他各类土地利用类型的面积均呈增加趋势,其中建设用地增长趋势最为明显。土地利用转移矩阵的分析结果表明,30年间黄土高原南部地区土地利用变化最为剧烈的是耕地,从数量变化上看有2577.72km2耕地转移为其他类别,其中转移为建设用地的面积最多,达到1157.07km2;建设用地的扩张、生态脆弱区退耕还林还草政策实施、农业产业结构调整是导致耕地面积减少的主要原因。从空间分布上看耕地变化较为剧烈的区域主要分布在陕西关中平原、山西汾河谷地和豫西北地区。建设用地30年间也有较明显的增加态势,面积累计增加了1275.73km2。研究区建设用地面积增加区域小于耕地面积减少区域,说明此时段有相当部分耕地转移为林地和草地,可见研究区退耕还林还草成效较明显。(2)从选取的两类典型区的土地利用变化情况可以看出,在研究时段内各个典型区的土地利用变化总体上都是呈现耕地面积减少,建设用地增加的趋势,但是耕地减少的原因不尽相同,基于流域单元的两个典型区耕地面积减少主要是在生态环境建设政策(如退耕还林还草)主导下的结果,其次是由于建设用地的扩展,如延河流域下游段耕地面积从研究期初的1818.1932km2,到2010年减少到了1739.5037km2,总面积接近于同期草地面积。减少的耕地按面积大小依次流向林地、草地和建设用地。相比之下,两个都市经济区耕地面积的减少主要是由于建设用地的增长速度较快的原因造成的。其中晋东南经济区耕地总面积到2010年从期初的10329.35km2减少到了10183.79km2,减少的耕地主要流向建设用地及林地。(3)土地利用与影响因素定量分析显示,影响黄土高原南部地区不同土地利用类型的自然和人文因素相对集中,主要包括地形坡度和人口密度,这说明该区域土地利用的人为干扰较强烈,人类活动是研究区土地资源和土地生态系统变化速率不断加快的重要因素之一。而自然因素中的地形坡度也是影响研究区土地利用的重要因子,不同土地利用类型,特别是农业用地和城镇建设用地与坡度呈现明显的负相关,说明地形坡度是研究区土地开发与利用的限制条件。(4)基于典型区的土地承载力时空变化测评结果显示,在2000年到2010年间,四个典型区各个区县土地承载力都呈现出两级分化的趋势,虽然两级分化的程度有差异,但总的趋势是一致的,而研究区整体土地承载力水平下降,需要借助于高效合理、科学的土地优化配置来提高土地承载力。对于整个研究区两极分化的情况,应当保证在土地承载力富余地区经济稳定发展的情况下,将富余地区的粮食外调给土地承载力超载的地区,实现整个区域的土地承载力平衡。(5)通过典型区不同土地优化配置方案的实验和对比发现,基于CLUE-S模型的配置方案对黄土高原南部地区的都市经济区拟合效果较好,其空间模拟结果的Kappa系数分别为0.8016(晋东南经济区)和0.7902(大西安规划区),而CA-Markov模型针对研究区流域单元效果较好,其kappa系数延河流域为0.7951,千河流域达到0.8069,两套方案针对各自典型区空间模拟效果的kappa系数均高于0.79。(6)基于CLUE-S模型的优化配置方案研究显示,晋东南经济区与延河流域下游段的建设用地均明显增加,而耕地均表现为减少,其中晋东南经济区建设用地增加的要多一些,这说明此区域经济发展和城市化建设潜力要大于延河流域下游段。两个典型区林地面积和草地面积都有不同程度地增加,主要是由于植树造林及退耕还林还草政策发挥了显著的成效,使得生态环境得以改善。基于土地适宜性和CA-Markov模型相结合的优化配置方案研究显示,大西安规划区和千河流域土地利用现状中的主要地类斑块需要进行调整,其中大西安规划区耕地的调整主要位于秦岭山地北麓,包括周至县、户县和长安区的南部以及蓝天县的部分区域,此外在三原县及富平县的北部边界地带也有少量需要进一步调整的耕地。从调整类型上看,主要是将不适合耕地分布的区位根据土地适宜性类型调整成林地或草地。而千河流域需要调整的耕地主要集中在流域东南部的千阳县,包括千阳县中部,东北部以及西南部靠近宝鸡市辖区的位置。凤翔县与千阳县交界地带的耕地也需要进行调整,此外流域中上游的陇县也有部分耕地需要调整的区域。本文在以下方面有所创新和改进：①以小网格推算模型为理论基础,构建了黄土高原主导气象要素关于经纬度和高度的数学模型及用于区域划分的综合指数模型,在GIS的支持下提出了基于自然区划与行政单元相结合的黄土高原南部地区的范围界线；②以黄土高原南部地区为研究对象,系统、全面地分析了过去30年间的土地利用变化过程,并采取对典型地区土地利用对比分析的方法,剖析了不同自然环境和社会经济发展水平下四个典型区的土地利用变化特点及时空差异；③将空间抽样方法引入到Binary Logistic回归分析中,对研究区不同土地利用类型特征尺度、时空差异及其影响因素进行定量分析,尝试构建基于空间约束性Logistic回归定量分析土地利用变化及其影响因素关系的新方案；④针对基于流域单元与基于都市经济区的两类不同典型区,提出将灰色线性规划模型(GLP)分别与CLUE-S模型及CA-Markov模型相耦合的两种方案,结合土地适宜性评价研究,在对不同典型区土地利用进行数量控制与结构优化的基础上,利用GIS软件将优化结果配置到空间单元上,由此为研究区土地资源开发利用、生态环境保护和社会经济发展提供统一的研究思路。更多还原

【Abstract】 Land is a very special resource. It is not only the important carrier for generation and change of natural environmental element but also the material assurance for subsistence and reproduction of human being, in addition, as the community characterized by natural quality and social economic property, the land as valuable resource plays an important fundamental and strategic role in economic construction and social development. Utilization of land resource is the complex social economic process in natural reproduction and economic reproduction under the persistent or periodic intervention of human activity, while the optimum distribution of land resource should be an important means to mitigate contradiction between human and land and implement sustainable utilization of land resource.The paper presents the characteristic of space time discrepancy of land utilization in southern area and typical area on loess plateau from different time space scales, adopting interpretation results of TM image remote sensing in30years from1980,2000to2010through RS and GIS integration and mathematical model, spatial statistical analysis and field investigation, taking southern area and internal typical area on loess plateau as research object based on the research of natural region division on loess plateau and shows the quantitative analysis of the relationship between various land use types and their influencing factors in southern area and four typical areas on loess plateau through combination of spatial sampling with Binary Logistic regression model based on research of the characteristic and scale of land utilization.And according to the problems existing in land utilization in various typical areas to build evaluation model to perform dynamic evaluation on land bearing capacity in every typical area in qualitative and quantitative manner. Finally, the two schemes that are to combine the grey linear programming model (GLP) with CLUE-S model CA-Markov model respectively are put forward for different typical areas based on valley unit and based on urban economic zone so as to apply the optimum results to the spatial unit adopting the two schemes based on research and appraisal of land suitability as well as quantitative control and structure optimization in terms of land utilization in different typical areas and implement the ecology oriented optimum distribution scheme for land resource to maximize ecological service value and economic benefit of land resource system through research, experiment and comparison of different schemes in different typical areas. The research findings are as follow:(1) The land use types in southern area on loess plateau are relatively complete, with the largest percentage of arable land, its percentages in all periods are all more than40%; meadow and woodland are next to it in this respect. In general, the area of land use types for the past30years (1980-2010) in southern area on loess plateau has changed to a certain extent, of which the arable land and unused land area have been on the decrease, while the area of other land use types have been on the increase, of which the land used for building has been on sharp increase. The land use type that plays a leading role in structural changes of land use in southern area of loess plateau is arable land, land used for building and meadow. The analysis results of transition matrix for land use indicate that the most significant change in land use in southern area on loess plateau for the past30years has been in arable land, the quantitative change shows that there has been2577.72km2of arable land changed into other types, of which the land used for building accounts for the largest percentage, being up to1157.07km2; the expansion of land used for building, implementation of the policy on returning farmland to forest and grassland in ecologically fragile area and agricultural restructuring are the primary cause for reduction in arable area. It is observed from spatial distribution that Guanzhong Plain in Shaanxi, Fenhe River valley in Shanxi and northwest in Henan have changed markedly in arable land use utilization, with significant increase trend in land used for building in the past30years. The area of land used for building has increased by1275.73km2. The region with area of land used for building increased is less than the region with area of arable land decreased, indicating that some arable lands have been changed into woodland and grassland in the period and it is observed that the policy on returning farmland to forests or grassland (Especially returning farmland to grassland) have achieved remarkable results.(2) It is observed from the analysis of land use change in two typical areas selected that the change of land use in every typical area shows reduction in arable land and enlargement in land used for building, but the causes for reduction in arable land are different, the reduction in arable land in two typical areas based on valley unit are as a result of the policy on ecological environment improvement (Such as policy on returning farmland to forest or grassland), the second cause is the extension of land used for building, for example, the arable area in lower reach of Yehe River valley has reduced from1818.1932km2at the beginning of research to1739.5037km2in2010, its total area is similar to the meadow area in the corresponding period.The reduced arable area has become forest land, grassland and land used for building in turn, the area of forest land changed from arable land is40.9832km2, and area of grassland changed from arable land is36.851km2, while the reduced arable area in Qianhe River valley has become grassland, land used for building and forest land in turn; from1996to2010,34.34km2of arable land became grassland,6.07km2of arable land became land used for building; in contrast, the reduction in arable land in two urban economic zones has been due to the fast speed of increase in land used for building.The total area of the arable land in economic zone in the southeast of Shanxi decreased from10329.35km2in beginning period to10183.79km2in2010, the decreased arable land mainly became land used for building and woodland,93.87km2of arable area became the land used for building from2000to2010, accounting for0.91%of the initial arable area, and28.29km2of arable area became woodland. In addition, the total area of the arable land in the planned area of Xi’an decreased from6429.96km2at the beginning of research period to6347.80km2in2010. And the decreased arable land mainly became land used for building, grassland and woodland in turn. From 2000to2010,101.04km2of the arable land became the land used for building,82.18km2and39.64km2of the arable land became into the woodland and grassland respectively and accounting for1.28%and0.62%of the initial arable area respectively.(3) Quantitative analysis of land use and influencing factor shows that the natural and human factors associated with different land use types in southern area on loess plateau include landform slope and population density, indicating that the human factor in land use in the area is strong; human activity has been an important cause for the accelerating change rate of the land resource and ecological system, while the landform slope in natural factor is also an important factor related to land use; different land use type, especially the agricultural land and land used for building in urban area show the obvious negative correlation, indicating that landform slope is the restriction condition for land development and utilization. For different typical areas, the factor associated with the probability of occurrence of various land use types and spatial distribution is slope. It shows positive correlation with woodland and grassland in typical area by and large but presents negative correlation with other land types such as arable land and land used for building and water area. In addition, for the economic zone and Yanhe River valley in the southeast of Shanxi, population density is also a factor of influence with a larger contribution rate in land use change. The factors associated with probability of occurrence and spatial distribution in land use type in Qianhe River valley also include temperature condition besides slope, including average temperature and accumulated temperature more than or equal to10℃, especially for land for construction in research area; temperature condition is mainly related to climatic suitability for land for construction in the research area.(4) According to the evaluation results of time space change of land bearing capacity based on the typical area, from2000to2010, for the areas lacking of land bearing capacity in all districts and counties in different typical areas, some areas are stable with a slight decline, some areas are aggravating, while, for the areas with margin in land bearing capacity, some areas are on the decline in bearing capacity level, and some areas with margin in the bearing capacity are on the increase in bearing capacity. Generally, the land bearing capacity in all lands in four typical areas shows polarization differently, but displaying the same trend in general, indicating that the land bearing capacity in research area is on the decline, thus, the efficient and reasonable and scientific land optimization is needed to improve land bearing capacity. As to the polarization in the research area, the areas with margin in land bearing capacity should transfer their grain to the areas lacking of land bearing capacity on the premise of steady development of the economy in areas with margin in land bearing capacity so as to ensure the balance of the land bearing capacity in the research area.(5) Research of the ecology oriented scheme of optimum distribution of land based on CLUE-S shows that the Kappa index of the two simulated typical areas are on the high side, of which the economic zone in the southeast of Shanxi is82.16, while the lower reach of Yanhe River valley is79.37; the simulated effect of the economic zone in the southeast in Shanxi is better than that of the lower reach in Yanhe River valley in typical area. It is observed from the land use demand and structure based on GLP mode and the spatial distribution of land use from2011to2020in economic zone and Yanhe River valley in the southeast of Shanxi simulated by CLUE-S model that the land for construction in two typical areas are sharply increased while the arable lands therein are on the decrease. But the land for construction in economic zone in the southeast in Shanxi increased more. And all arable lands are on the decrease, indicating that the potential for regional economic development and urbanization is superior to that for the lower reach of Yanhe River valley. The water area in two typical areas is on the slight increase, with enlargement of woodland and grassland to varying degrees due largely to the policy on tree planting and returning farmland to forest land and grassland so as to achieve improvement in ecological environment.(6) Research of the optimum distribution scheme based on combination of land suitability with CA-Markov model shows that the main land sections in the current situation of land utilization in the research area of two typical areas need adjustment, of which the arable land in the planned area of Xi’an lies in the north foot of Qinling Mountain, including Zhouzhi County, Huxian County and the south of Chang’an District as well as part of Lantian County, in addition, there is a few of land in the north border between Sanyuan County and Fuping County is to be adjusted. It is observed from the land adjustment that the location unsuitable for arable land shall be adjusted into woodland and grassland, while the arable lands to be adjusted in Qianhe River valley are concentrated in Qianyang County in the southeast of the River Valley, including the middle northeast and southwest of Qianyang County near to Baojing. The lands in border between Fengxiang County and Qianyang County are also to be adjusted. In addition, the lands in Long County in upper reach of the river valley are to be adjusted.The research has made innovation and improvement in the following aspects:①The main meteorological elements on loess plateau about longitude and latitude and mathematical model as well as comprehensive index model are built based on prediction model as theoretical basis, based on which the natural area is subject to division and the division results have been tested. Based on the natural area division on loess plateau, the borderline on the southern area on loess plateau based on natural area division and administrative unit has been presented.②The land use change process for the past30years has been subjected to systemic and comprehensive analysis in the southern area on the loess plateau as research object, and the characteristic and time space discrepancy in land use change in different natural environment and social economic development within the research area have been analyzed by contrastive analysis in typical area.③Quantitative analysis of the character, scale and time space discrepancy of various land use types within research area have been conducted by Binary Logistic regression analysis in the research, in addition, in order to prevent the auto correlation influence of land use types in space, the spatial sampling is introduced to Binary Logistic regression analysis, the new scheme to perform quantitative analysis on the land use change and its influencing factor based on spatial restriction Logistic has been established.④The two schemes that are to combine the grey linear programming model (GLP) with CLUE-S model CA-Markov model respectively are put forward for different typical areas based on valley unit and based on urban economic zone so as to apply the optimum results to the spatial unit adopting the two schemes based on research and appraisal of land suitability based on the quantitative control and structure optimization. The ecology oriented land resource optimum distribution scheme to maximize the ecological service value and economic benefit in the land resource system has been put forward to offer the uniform research idea for development and utilization of land resource, protection of ecological environment and social and economic development.更多还原