Application of gray clustering method in the ecological classification of the cities in China

For the vast territory and the huge differences amongst cities in China, there is an urgent need to carry out research on the division of regional eco-types of cities, so that eco-city construction can be conducted to suit the local conditions. However, almost all the previous classification research is qualitative, such as experience summary and construction mode induction. In this paper, an ecological classification of the cities in China has been obtained by the comparing the effects of six methods: gray clustering, fuzzy clustering, the coefficient of mode ratio, comprehensive index, systematical clustering and k-means clustering. The research results show that: (1) compared with the other five clustering methods, the gray clustering is the best method for the eco-city classification. The classification based on gray clustering method performs better in the interpretation of regional ecological background and urban evolution, and the classification based on coefficient of mode ratio performs better in describing urban comprehensive conditions. (2) There are huge differences among cities in China in terms of regional ecological backgrounds, urban evolution and urban comprehensive conditions. Therefore, we should constitute the measures of development, management and control more locally, to enhance the operability of the eco-city ideology, and protect the special individual characters of different cities.