基于CA模型的城镇扩张模拟与分析

发布时间：2018-05-24 10:51

  本文选题：元胞自动机(CA) + 城镇化　； 参考：《中山大学学报(自然科学版)》2017年03期

【摘要】：引入逻辑回归CA模型挖掘城市发展的转换规则,模拟/预测了佛山市的城镇扩张,并采用转移矩阵和景观指数等方法分析城镇化过程。结果表明佛山的城镇化过程侵占了大量的农田、森林和其它(包括果园和池塘等)用地类型。运用CA进行模拟,其总精度和Kappa系数可达到93.7%和0.86。通过景观指数分析,佛山城市发展将会更为聚集,其最大班块指数增大。在城市优先发展情景下,到2020年将有191 km~2农田将转变为城市用地,占所有农田(706 km~2)的27%。在保护农田和保护绿地的情景下,农田和绿地可以到较好的保护,但是它将会占用其它土地类型以满足城市扩张的需要。通过分析空间变量和城市用地的关系发现,各个空间变量对城镇化的廊道作用显著。
[Abstract]:The transformation rules of urban development are excavated by introducing the logical regression CA model, and the urban expansion in Foshan is simulated / forecasted, and the urbanization process is analyzed by means of transfer matrix and landscape index. The results show that the urbanization process of Foshan occupies a large number of farmland, forests and other land types (including orchards and ponds). The total accuracy and Kappa coefficient of CA are 93. 7% and 0. 86% respectively. Through landscape index analysis, Foshan city development will be more concentrated, its largest block index increases. Under the priority urban development scenario, 191 km~2 farmland will be converted into urban land by 2020, accounting for 27% of all farmland (706 km2). In the situation of protecting farmland and protecting green land, farmland and green land can be protected better, but it will occupy other land types to meet the needs of urban expansion. By analyzing the relationship between spatial variables and urban land use, it is found that each spatial variable plays a significant role in urbanization corridor.
【作者单位】： 广东财经大学地理与旅游学院;中山大学地理科学与规划学院;
【基金】：国家自然科学基金(41531176,41371376) 国家社会科学基金(14BRK017)
【分类号】：F299.27;TP301.1
，

本文编号：1928788