【作者】 郭伟；

【导师】 赵春江； 冯仲科；

【作者基本信息】 北京林业大学 ， 林业装备工程， 2012， 博士

【摘要】 生态系统服务价值评估是生态系统可持续性研究的基础,根据历史时期的生态系统类型和地表覆盖状况,以生态服务价值最大化为目标,以空间适宜性为原则对生态系统类型数量和空间分布预测与优化具有十分重要的意义。北京地区地形地貌较复杂,生态景观类型丰富多样,城市密集、人口稠密、产业集聚,不合理的生态系统类型组合方式使生态系统面临较大压力,城市化进程过快使得生态环境资源利用不够集约、还存在牺牲生态环境效益获取经济效益的情况,使得研究区内社会经济发展与生态环境保护之间的矛盾较为突出。进行生态系统景观格局预测及优化研究,对于改变现行的不健全价值体系、增强人们的环境保护意识、生态环境资源集约合理利用,并实现可持续发展具有一定的意义。遥感技术具有空间宏观性、多分辨率(光谱和空间)、多时相、周期性、信息量丰富等特点,具有其他观测手段无可比拟的优势,即可以提供生态系统的宏观空间分布信息,又能提供局部的详细信息及随时间、空间变化的动态信息等,使其成为生态系统研究的有力工具。本研究选取北京市为研究区,以遥感技术为分析手段,在动态估算1978-2010年间生态系统类型的空间分布信息以及生态环境质量参数的基础上评估了研究区的生态系统服务价值,并以其最大化为目标,对研究区未来生态系统景观格局进行了预测及优化研究。研究内容及相关结论如下：(1)生态系统景观格局变化分析分析北京地区的地域特征,结合国家二级分类体系,得出适用于北京地区生态系统服务价值评估的分类体系为农田、森林、草地、水域、城市、荒漠六类；以16景Landsat系列中分辨率遥感影像为数据源,结合2010年166个野外调查数据,采用基于特征向量组合的神经网络算法提取1978年-2010年32年4个时期的生态系统类型空间分信息;利用转移矩阵与景观指数分析北京地区近32年来各生态系统类型面积内部结构变化规律与景观格局的时空变化规律：从1978-2010年间农田面积、荒漠大量减少,减少幅度分别为42.00%,58.15%,森林面积、城市面积大幅增加,增加幅度分别为35.87%,39.43%,草地变动幅度较大、水域基本保持稳定,农田与荒漠多转为森林和城市。北京地区从类型水平景观指数分析结果显示：森林、农田和城市三种生态系统类型景观面积百分比最高,景观破碎度与景观分维数低于草地、水域和荒漠,凝聚度则高于其他他三种生态系统类型,属于研究区的优势观类型；从景观水平看：斑块密度指数、平均分维数以及多样性指数的下降以及凝聚度指数的上升都说明研究区受到人类影响的程度在不断加大,使得景观破碎度越来越小,景观形状越来越规则,而各生态系统类型景观斑块在景观中趋于均衡。(2)生态环境质量参数估算与变化分析土壤侵蚀量和NPP是估算生态系统服务价值的物质量基础和必要前提,因此也是表征生态环境质量的重要参数。以前两章研究基础(空间化的气象、土壤、地形数据与历史各时期生态系统类型空间分布)为数据源,分别采用通用土壤流失方程(USLE)和光能利用率模型计算北京地区32年士壤侵蚀量以及NPP的时空分布,然后对这两个表征生态环境质量的重要参数的时空格局变化及影响因子进行分析,结果表明：32年来北京地区发生轻度、微度侵蚀的比例占总面积的70%-80%,其中轻度、微度侵蚀的区域面积占总侵蚀面积的90%,西部及东北部有少部分地区发生中度以上侵蚀。不同生态系统类型土壤侵蚀强度顺序为荒漠>森林>草地>农田。侵蚀量从1978年到2010年一直呈上升势态,2000年达到最高,然后开始回落。不同坡度下土壤侵蚀变化规律表现为坡度越高侵蚀强度越强,侵蚀总量表现为随坡度升高侵蚀总量先升高后降低；NPP均值和总量都是连年递减的,均值从1978年的1703.46gC/m2·a减少到2010年的1348.09gC/m2·a,总量从1978年的37.66TgC·a减少到2010年的22.25TgC-a。不同生态系统类型产生NPP的量为森林>农田>荒漠>草地>城市>水域。降水、太阳总辐射、温度都是影响NPP生产的重要因子,NPP与降水量、太阳辐射正相关,与温度负相关。(3)生态系统服务价值定量估算与变化分析结合北京地区生态环境特点,考虑数据的可取性和可靠性,确定了北京地区适用于遥感的生态系统服务价值评估指标体系,包括生产有机物、营养物质循环、涵养水、土壤保持、吸收和分解污染物质、气体调节6项服务功能；以木研究第四章研究基础(北京地区历史各时期土壤侵蚀量和NPP)为生态系统服务价值物质量,针对每项服务功能,计算北京地区1978-2010年32年4个时间期的单项单位面积服务价值的年际变化并分析不同生态系统类型下该项生态系统服务价值的变化特点;对北京地区生态系统服务价值的组成及其时空格局进行了深入的分析：在总价值中,生态系统涵养水的服务功能所产生的价值贡献最大,其次依次为营养物质循环服务价值、生产有机物价值、土壤保持价值、气体调节价值、吸收和分解污染物的价仇。32年间北京地区生态系统服务总价值一直呈现逐渐下降的趋势,1978年最高为1425.48x 108元/a,2010年最少为919.82×108元/a,降幅为35.47%,下降幅度较大。1978年-2010年单位面积服务价值变化趋势与服务总价值致,1978年最高,2010年最低,价值区间在8.70元/m2-5.61元／m2,变化幅度为35.45%。生态系统服务总值的空间分布呈现出北高南低、西北高东南低、山区高平原低的特点；不同土地利用／覆盖类型的生态系统服务价值大小顺序为森林>农田>城市>荒漠>水域>草地,而单位面积服务价值的大小顺序为农田>森林>水域>草地>荒漠>城市。(4)生态系统景观格局变化预测与灰色优化研究根据各历史时期生态系统类型景观格局空间分布,采用CA-Markov模型进行2020年景观格局预测,利用Markov预测各生态系统类型之间转移概率,利用CA实现空间转移模拟。CA-Markov模型可以较好地模拟北京地区生态系统景观格局变化趋势,但这种趋势从生态效益的角度来看是不合理的。为了使北京地区生态系统服务价值实现最大化,各生态系统类型之间结构布局合理,采用灰色系统与元胞自动机相结合,利用灰色理论实现空间模拟的优化。采用灰色线性规划算法,以生态系统服务价值最大化为目标函数,优化各生态系统类型的数量面积组合,利用灰色关联度筛选各生态类型空间适应性评价指标并确定其权重,然后采用灰色聚类进行空间适宜性评价,生成各生态类型空间适应性规则。最后,将数量优化方案与空间优化方案共同输入CA模型中,实现了2020年北京地区生态系统景观格局优化。本优化模型基本实现使生态系统景观格局既能在数量结构上达到优化的目的,又能在空间布局上达到优化,使各生态系统的空间布局优化和生态服务价值数量结构优化,最终实现了生态系统景观格局优化的既定目标。本研究的主要特色与创新之处：1.将以往利用单位面积价值当量对生态系统服务价值总量进行静态评估改为基于时间序列时空数据的动态评估。论文以时间序列的遥感、土壤、气象、社会经济为数据源,利用遥感技术、GIS空间分析技术、计算机技术,选取十壤侵蚀量与NPP为生态环境质量评价参数,动态估算了北京地区32年间(1978-2010)的单位面积单项生态系统服务价值,并分析了生态环境质量变化与生态服务价值构成及变化规律。2.耦合灰色理论与元胞自动机CA,对北京地区生态系统景观格局进行预测及数量与空间优化。首先,利用灰色线性规划算法,以生态系统服务价值最大化为目标函数,以生态环境、社会经济数据为约束条件,优化各类型生态系统的数量面积组合；然后,利用灰色关联度筛选各生态类型空间适应性评价指标并确定其权重,采用灰色聚类进行空间适宜性评价,估算各生态类型空间适应性规则；最后,在模型CA中将数量与空间优化方案作为演化规则,预测并优化了2020年北京地区的生态系统景观格局。更多还原

【Abstract】 Ecosystem services values assessment is the basis of the sustainability of ecosystems. It is of great significance to optimize the prediction of the ecosystem cover type’s number and space distribution. This paper/dissertation needs to be (according to) based on the historical ecosystem’s land cover conditions, in combination of the principle of the space suitability, to maximize the Ecosystem services values.In Beijing, the topography is relatively complicated. The ecological landscape is diverse, the population and the buildings are highly dense, and the industries are also quite concentrated. Unreasonable combinations of ecosystem types will bring the ecosystem enormous pressure; the social and economic development results in insufficient intensive use of the ecological environment resource, and particularly there are cases that people try to pursue economic benefits at the cost of eco-environmental benefits. All these problems make the conflict between the local people and eco-environment more pronounced.Hence, it is still of great significance to study the prediction and optimization of the ecosystem types and the landscape patterns, in order to change the current problematic value systems, to raise the people’s awareness of environmental protection, and help them to reasonably use the resources intensively.Remote sensing technology has quite unparalleled advantages when compared with other observation methods, in terms of its spatial macro, wide viewing angle, multi-resolution (spectral and spatial), multi-time phase, regular periods, and rich information. So, the remote sensing technology can not only provide the information of ecosystem macro-space distribution, but also provide detailed local information and the dynamic information over time and space, thereby making itself one powerful tool to study the ecosystem.This study took maximized the ecosystem value as a goal to predicted and optimized the ecosystem types and the landscape patterns in future used the remote sensing technology to carry out the assessment of the Ecosystem services values based on the ecosystem type spatial distribution and ecosystem environment quality parameter in Beijing during 1978-2010. The major contents and results in this dissertation are as follows: (1) Analysis of the landscape pattern changes in ecosystem typesAfter analysis of the geographical features of Beijing combined with Level-II national classification system, this paper defines the Beijing’s ecological value assessment in the following six types of ecosystem:farmland, forest, grassland, water area, city, and desert. This paper also employed the 16 views (scenes) Landsat remote sensing images with middle resolution as data source, in combination of the 166 field survey point data in 2010; the neural network based on characteristic vector method was used to extracted the ecosystem type’s space distribution which was defined in 4 stages (periods) for the 32 years from 1978 to 2010.In the meantime, Used transition matrix and landscape index to analyzes the change patterns of the ecosystem types and the internal structure, as well as the space change patterns of the landscape in the past 32 years in Beijing, On the basis of this analysis, this paper concluded the following:during the period of 1978-2010 the area of arable land, desert decreased considerably, by 42.00% and 58.15% respectively; the forest area and city have increased substantially, by 35.87%,39.43% respectively; the grassland changed enormously and the water area maintains stable, but the arable land and desert normally have been changed into forest and residential use land. Analyzing from types and landscape index, this paper drew the following conclusions:the three ecosystem types which were forest, farmland, and city, cover the highest percentage of the landscape. The degree of landscape fragmentation and fractal dimension were less than that of the grass, waters and desert, and the concentration was higher. So it is the dominant landscape type. From the point view of landscape level, the patch density index, the average fractal dimension, as well as the falling down of the diversity index and the going up of the condensation index, have indicated that the research area has experienced a greater impact from the human’s behaviors. This results in smaller and smaller landscape fragmentation, more and more regular landscape shapes, and also equilibrium of the landscape patches in the landscape as a whole.(2) Analysis of estimation and change in ecological environment quality parametersThe amount of soil erosion and NPP are the basis and necessary prerequisite for estimating the Ecosystem services values, also they are the important parameters to characterize the quality of the ecological environment. Using the study of the previous two chapters as the data source, which included its space meteorological, soil, terrain data and ecosystem land cover spatial pattern in different historical periods, this paper had calculated Beijing’s amount of soil erosion and NPP’s spatial and temporal distribution for a period of 32 years with the methods of Universal Soil Loss Equation (USLE) and light use efficiency model. Later, it carried out analysis of the spatial and temporal pattern change of these two ecologic environmental quality parameters and the influential factors, and concludes the following:in the past 32 years, the result showed that:the slightly soil eroded area covers 70% to 80% of all the area in Beijing. Specifically, mildly or slightly eroded area amounts to 90% of the whole erosion area, and in some areas of the western part and northeastern part, moderate and above erosion has occurred. Among all the ecosystem types, the soil loss ranks as follows:desert>forest>grassland>arable land. In addition, the erosion has shown an ascending trend from 1978 to 2010 peaking at 2000 but beginning to drop down subsequently. The soil loss in different slopes has shown such a pattern:the greater the slope, the more the soil loss, and in the curve of the overall erosion amount and the slope, it shows an uptrend trend first and then a downward trend. The NPP average and the total amount has decreased over these years, decreasing to 1348.09 gC/m2.a in 2010 from 1703.46 gC/m2.a in 1978, while the total amount down to 22.25 TgC/a in 2010 from 37.66 TgC/a in 1978. The NPP amount for different ecosystem types has shown such a pattern:forest> arable land>desert> grassland> city> water area. The affect NPP production important factors are precipitation, solar radiation, and temperature.The NPP has a positive correlation with precipitation and solar radiation, but negative correlation with the temperature.(3) The quantitative calculation and change analysis of Ecosystem services valuesConsidering the region ecological environment characteristics of Beijing and the availability and reliability of the data, this paper chose the Ecosystem services values evaluation index system suitable for remote sensing, which included six services functions as producing organics, nutrient recycling, reserved water, soil conservation, the pollution’s absorption and decomposition, and air conditioning.Based on the study of chapter 4 as the Ecosystem services values indicators, including the soil erosion and NPP in Beijing in different historical periods as research basic data, For each service functions, We calculates the annual change of the service value per unit area for each item in the four periods of the 32 years from 1978 to 2010, and analyzes the Ecosystem services values’s change patterns in different ecosystem types. From the in-depth study of the composition of the Ecosystem services values and their spatial and temporal patterns in Beijing:among the total values, the reserved water contributes the most to the service values, followed sequentially by the nutrient cycling, production of organic matter, soil conservation, air conditioning, and the pollution’s absorption and decomposition. In the 32 years, the total Ecosystem services values have shown a descending trend:the greatest is 1425.48×108 Yuan/a in 1978, the least is 919.82×108 Yuan/a in 2010, a decline of 35.47%. As regards to the service value per unit area for each item, it shows a similar trend as the service total amount between 1978-2010:the greatest in 1978 and the least in 2010, ranging between 8.70 Yuan/m2-5.61 Yuan/m2, with a change of 35.45%.For the space distribution of the Ecosystem services values, it showed such a pattern:high in northern part but low in southern part, high in northwestern part but low in southeastern part, and high in mountainous areas but low in plain areas. The Ecosystem services values for different land use/cover types showed such a pattern in the descending rank:forest>arable land> city> desert> water> grassland, but the service value per unit area ranks as:arable land> forest>water area> grassland> desert> city.(4) the study of the prediction of the ecosystem’s landscape pattern change and the gray optimizationAccording to the ecosystem type’s landscape space distribution in different historical periods, used the method of CA-Markov for the prediction of 2020’s landscape. Meanwhile, it used Markov to predict the transition probabilities, and the CA to realize the space transition simulation. The CA-Markov model can simulate the ecosystem types’ landscape pattern change trend preferably in Beijing, but this trend might look unreasonable from the point of view of ecological benefits. In order to maximize the ecological system serve value benefits of Beijing, and make a reasonable structural composition of the various ecosystem types, this paper used a combination of gray system and cellular automata to realize the space simulation with the gray optimization theory. Drawing upon the gray linear programming algorithm, this paper set the maximization of the Ecosystem services values as the objective function, to optimize the area combination of the various ecosystem types; then we used the gray correlations to filter the assessment indexes of the space adaptability for various ecosystem types, and determines its weight respectively; then, it adopts the gray concentration to assess the space suitability and generate the space suitability rules for different ecosystem types. Finally, we input the area optimization solution and the space optimization solution into the CA model, to realize the landscape pattern optimizations of the ecosystem types in 2010 for Beijing. This optimization model can basically achieve the purposes of optimizing the ecosystem types in terms of number and the space distribution, and finally the set goal of the optimization of the ecosystem type’s landscape patterns.The main features and innovations of this paper:1. It adopts dynamic assessment method of the spatial and temporal data based on the time sequence, instead of the past static assessment method of the per unit area amount to derive the total Ecosystem services values. The study used the time-series remote sensing images, soil, weather, and socio-economic benefits as the data sources, remote sensing technology, GIS space analysis, and computer programs were used. Chose the soil erosion and NPP as the parameters to assess the ecosystem environment quality, it dynamically estimated the per unit area Ecosystem services values in 32 years from 1978 to 2010 in Beijing, and analyzed the change of ecosystem environment quality, the composition and change patterns of the Ecosystem services values.2. This paper also coupled the gray theory and the cellular automata, to predict Beijing’s ecosystem landscape patterns and optimize in terms of number and space. First, we used the gray linear programming algorithm, set the maximization of the Ecosystem services values as the objective function, and the ecosystem environment and the social economic data as the independent parameters, to optimize the area combination of the various ecosystem types. Then used the gray correlations to filter the assessment indexes of the space adaptability for various ecosystem types, determines its weight respectively, the gray concentration was used to assess the space suitability and estimates the space suitability rules for different ecosystem types. Finally, it used the number and space optimization solutions as the basic rules in the CA model to predict and optimize the ecosystem system landscape patterns in 2010 for Beijing.更多还原