Abstract: The paper examines the urban growth patterns in Quanzhou City based on remote sensing images over the period 1993-2008. Using SLEUTH cellular automata model, we simulate two typical scenarios of urban development: the self-organizing expansion and planning oriented growth in Quanzhou. We also employ landscape metrics and spatial statistical analysis methods to analyze the spatial characteristics of urban spatial expansion in the study areas over the study period. The results show that: (1) The SLEUTH model is a useful tool to predict urban growth patterns under different scenarios. (2) Four landscape metrics, i.e., MPS, ED, AWMSI and MPI, for the planning-oriented growth scenario are significantly greater than those derived from self-organizing expansion scenario. The planning oriented scenario is also characterized by better connectivity and integrity of urban growth patches. This further demonstrates that the urban planning, if efficiently implemented, can help to achieve the goal of compact and sustainable development. (3) However, over the years, we find that urban expansion rate and spatial aggregation in the study area are weakening and the hotspots of urban growth also changes as follows. First, from 2008 to 2020, the distribution of urban growth hotspots is characterized by a "single-ring" pattern with an emphasis on the "cross-river development". Second, from 2020 to 2030, the layout of hotspots is more dispersedly distributed characterized by "bay-area-oriented development" and "isolated development". We argue that a combination of simulation, landscape metrics and spatial analysis can provide more reliable evidence to support policy decision making concerning city planning and management in China.

Key words: central city, urban growth, hot spots, SLEUTH model, spatial association, Quanzhou