Village classification in metropolitan suburbs from the perspective of urban-rural integration and improvement strategies: A case study of Wuhan, central China

Urban-rural integration is an ideal urban-rural relationship that balances and coordinates urban and rural development . Suburbs, as transitional zones between urban and rural areas, are usually treated as part of urban areas and are insufficiently considered in this integration. However, during rapid urbanization, the suburbs of metropolitan cities attract huge populations and industrial agglomeration . Mixed land use types, population structure, industrial construction, and underdeveloped infrastructure cause the suburbs of big cities to face significant spatial differentiation of socioeconomic development and public services, together with an ecological environmental pressure brought about by drastic land use change . The complexity, gradual change, and dynamics of suburbs render it difficult to map them effectively. In addition, due to different conditions such as the development basis, traffic conditions, spatial location, and resource endowments, the type and degree of integration of villages around metropolises are also differentiated, which determines their future development direction. Against this background, this study first maps the suburban boundaries at the village scale using a logistic-geographically weighted regression (GWR) model, which is a localized parameter estimation method. This study classifies the types of suburban villages in metropolitan areas using three dimensions: socioeconomic, spatial, and ecological integration. Finally, improvement suggestions have been proposed for villages with different types of urban-rural integration. Following findings are summarized: 1) The logistic-GWR model identifies 2626 suburban villages in Wuhan; 2) the integration differentiation identifies 370, 564, 216, and 365 suburban villages that exhibit comprehensive, ecological, socioeconomic, and spatial disintegration, respectively; and 3) solutions for promoting urban-rural integration of suburban villages are provided accordingly. • This study proposed a method of mapping suburban boundaries at the village scale based on multi-source data and a combined logistic-GWR model. • The verification based on the green space ratio, remote sensing data, and street view photos demonstrated the feasibility of the proposed method. • The suburban villages were classified by evaluating the dominant urban-rural integration type. • Improvement strategies of urban-rural integration for suburban villages were developed based on the evaluation results.