Identifying urban candidate brownfield sites using multi-source data: The case of Changchun City, China

Over the past 40 years, brownfield remediation has become an important issue of urban economic transition and revival, providing many social, economic, and environmental benefits to cities. Identifying urban candidate brownfield sites is an essential task of brownfield projects, but few attempts have been made to create a comprehensive dataset of their spatial distribution in China. This paper presents a new method for identifying urban candidate brownfield sites using multi-source data, including current urban land use maps, Baidu heat maps, business data, and Baidu street view. Our method for identifying urban candidate brownfield sites consists of four steps: (1) preparing and processing the basic data while screening potential contaminated sites using a current urban land use map, mapping the geographical coordinates of operating and non-operating businesses, and calculating the Baidu heat index; (2) integrating spatial multi-source data; (3) screening the intensity of use and the degree of abandonment for suspected sites in Baidu street view; and (4) conducting field surveys. The method was proposed and tested using a case study in the central region of Changchun, China, and 146 candidate brownfield sites were identified. The new data that we collected and our methodology can be used as quantitative and universally applicable tools within a discriminant analytical model framework, for urban geography, urban design, and landscape architecture. The resulting maps for urban candidate brownfields could also provide local government agencies with a technical reference for brownfield remediation.