节点文献
基于元胞自动机的喀斯特石漠化模拟与预测
The Simulation and Prediction of Karst Rocky Desertification Based on Cellular Automata
【作者】 王晓学;
【导师】 吴秀芹;
【作者基本信息】 北京林业大学 , 地图学与地理信息系统, 2011, 硕士
【摘要】 西南喀斯特山区的石漠化问题是目前我国西部地区最为突出的地域环境问题之一。本研究以导致石漠化发生发展的自然、人文因子为切入点,尝试建立简单、有效的喀斯特石漠化模拟及预测的数学模型-KarstCA模型,该模型包含了影响喀斯特系统地表覆被变化的基本生态过程(如植物定居、植物死亡、水蚀风蚀引起的土地退化以及岩石成土过程等)。利用随机元胞自动机的特点,使喀斯特系统中植被-裸土-裸岩状态在一定概率下发生状态转移,建立基于生态过程的喀斯特石漠化元胞自动机模型,利用元胞自动机简单邻域规则产生复杂空间格局的优势,来模拟喀斯特系统石漠化现象在时空上的发展演化趋势。利用KarstCA模型模拟自然或/和人文因素影响下,喀斯特系统地表覆被特征的变化情况,定量化自然、人文因素对石漠化发生发展的贡献率;然后进一步探讨了不同人类活动强度情景下石漠化的时空特征、不同元胞大小对模型模拟结果的影响以及喀斯特石漠化元胞自动机模型的稳健性,并结合现有研究文献及渗透理论初步探讨石漠化发生的临界阈值。本研究得出的主要结论包括:1)在考虑自然、人文因素的情景下,KarstCA模型模拟结果有73.4%是正确的,而只考虑自然因素的情况下,有56.7%是正确的,说明模型较好的包含了石漠化发生发展的关键生态过程;进一步研究发现,当植被达到54%以上并继续增加时,裸岩发展趋势与之呈明显的负相关;2)1992年以来的人类活动对喀斯特系统植被的影响主要是正作用;不同人类活动强度显著影响斑块特征,都呈现出人类活动先使斑块异质性增加,随着人类活动的持续进行,斑块趋于同质化;3)元胞大小对KarstCA模型空间格局以及系统的稳定性都有影响,植被斑块-面积的关系的研究尺度大致在5-30m之间,而将其幂函数关系作为指示石漠化发生发展的有效性需要进一步深入研究。
【Abstract】 Rocky desertification of karst mountain region in Sourthwestern China is one of the most important regional environmental problems in Western China. Current researches on Rocky desertification are excessively concerned with factors inducing rocky desertification, methods evaluating the local karst environment, the measures recoverying ecosystem, but especially neglect the researches on mechanism and spatial structure of rocky desertification. This Dissertation focuse on the problems in the study of Karst rocky desertification, and try to develop a simple and effective mathematical model to simulate and predict the karst rocky desertification based on the natural and human factors leading to the development of desertification. The ecological process of surface cover that affects the karst system was also integrated into KarstCA models, such as colonization, plant death, land degradation and soil formation of rock. According to random characteristics of cellular automata, such transitons between the vegetation-bare soil-bare rocky were made in certain probabilities. Then a cellular automata model based on ecological processes of karst rocky desertification was established. KarstCA model can simulate the spatial and temporal development and evolution of desertification in karst systems according to a simple neighborhood rules to generate complex spatial pattern. This model considers configuration of neighborhood characteristics from three aspects from a variety of factors acting on different scales.KarstCA model can be used to simulate the natural and human factors,and only the scenarios of natural factors,to explore characteristics of land cover changes of karst system, which contributes to quantifying natural and human factors on the development of desertification. Then the author further explored spatial and temporal characteristics of rocky desertification under the scenarios of the different intensity of human activities and the different cell size, and tested the robustness of KarstCA model. Finally, the author discussed the experience of the critical threshold characteristics of rocky desertification in karst system combining with the existing research literature and penetration theory. The main conclusions of this study include:1)73.4 percent of the results of KarstCA model simulation were correct considering the natural and human factors.While 56.7 percent of the results were correct when only considering the case of natural factors. It was found in further study that the trend of bare rock was negatively related to the vegetation patch area while it amounted to more than 54 percent and continued to increase,2) The impac tof human activities on vegetation in the karst system was mainly positive effect since 1992. Different human activities significantly affected the vegetation patch characteristics. They all initially presented a heterogeneitious trend with the ongoing human activity, and then tended to become homogeneity,3) Cell size has some effects both on the spatial pattern of KarstCA model and system stability. The effectiveness of vegetation patches-area power function, as an indicator of the development of desertification, needs to be further studied.
【Key words】 Karst system; rocky desertification; cellular automata; linear spectral unmixing; human activity;