南昌地区城市建成区景观结构特征对建设用地扩张的影响

Impacts of urban landscape structure on urban sprawl: A case researches in Nanchang

利用南昌地区 1988～ 2 0 0 0年期间的 4景城市建设用地分布图和叠图代数计算方法 ,分两个比较时段 (1988～ 1991年及1998～ 2 0 0 0年 )研究了快速城市化地区建成区景观格局特征对建设用地扩张的影响 ;重点分析了边缘非城镇像元的邻域特征、非城镇像元与最近建成区像元的距离以及建成区斑块的规模对边缘及全区非城镇像元向建设用地转化比率的影响。结果表明 ,非城镇边缘像元的邻域水平与其向建设用地转化比率之间的相互关系可以用一阶直线方程加以表述 ;非城镇像元与最近建成区像元之间的距离对转化率的影响呈幂函数关系 ,建成区斑块规模的对数与周围非城镇像元向建设用地转化比率之间的关系可以用二项式方程进行拟合。尽管两个比较时段之间的城市化程度差异明显 ,但拟合方程的类型保持一致 ,表明建成区格局特征对新增建设用地的影响是一种机制性的联系。最后探讨了将上述相关关系引入城市化景观动态模拟研究中的可能性以及在具体变量赋值过程中应当注意的有关问题。

Urban sprawl led by continuous expansion of urban built-up areas is one of the essential features of landscape change in regions experiencing rapid urbanization. Expansion of urban built-up areas is constrained by both the external factors ((e.g.,) topography, land use, and enforced protection measures) and the development needs within cities (e.g. growth of the socio-economic activities and population). Meanwhile, the existing urban structures, such as size and shape of the urbanized patches, and their distances to the non-urban patches around, will also exert significant constraints to the expansion of urban built-up areas. However, these constraints and their impacts have not been adequately addressed in previous research of urban landscape ecology. Though some case studies have already taken into account the morphological characteristics of the urbanized patches and the human settlement patterns of the peripheral non-urban pixels in landscape dynamics modeling, the mechanism of these constraints and their impacts remain unclear and constitutes a gap of research in this field. This gap has constituted a significant obstacle for further analysis of the urban sprawl and modeling of the dynamics of those areas experiencing urbanization. Existing models fail to take into account the influences of the current structural features of the urbanized area to the conversion of the peripheral non-urban pixels to the urban pixels, and as a result, the existing modeling research lacks effective analysis of built-up area expansion and associated mechanism. In this article, the author, using Nanchang City in Jiangxi Province as a case study area, used 4 scenes of TM satellite imageries (1988～1991 and 1998～2000) to extract built-up area distribution maps, followed by overlay and operator methods to examine the mechanism by which the existing structural features of the urbanized area exert influence to the expansion of new built-up areas. Three issues were addressed in this article: (1) influences on the ratio of conversion of the marginal non-urban pixels near the urbanized region into urban pixels caused by the neighborhood level of the marginal non-urban pixels; (2) influences on the ratio of conversion of non-urban pixels to urban pixels caused by the distance between the non-urban pixels and the nearest urban pixels; and (3) influences on the ratio of conversion of non-urban pixels to urban pixels caused by the size of the urban patches nearest to the non-urban pixels. Based upon the empirical results, the following conclusions are drawn: (1) There are striking differences in the urbanization process during the two study time periods. The increase of the built-up areas between 1988～1991 was achieved mainly by the expansion of the previously urbanized patches, whereas during 1998～2000, many new patches of built-up areas emerged. In addition, the increase of built-up areas in the first time period is much slower than that during the second period. (2) The conversion ratio of non-urban pixels to urban pixels has distinct correlations with neighborhood characteristics of the marginal pixels, distance between the non-urban pixels and the nearest urban pixels, and size of the built-up patches. The three correlations can be fitted by linear, power, and binomial functions. The forms of the best fitted formulas have no relations with urbanization rate, built-up area expansion mode, and initial status of urbanization. (3) The correlations among the data sets suggest that the existing urbanized areas have strong influences on the expansion of built-up areas. As the forms of the best fitted formulas are not related with the change of characteristics in the process of urbanization, it can be inferred that there exist further mechanisms behind these relations. Therefore, the results from this research provide both important clues for further exploration of the built-up area expansion in the process of urbanization, and strong scientific evidences for introducing these relations into research of urban landscape dynamics modeling.