The spatial decision support system (SDSS) is widely used in environmental problem management. In this paper, focusing on the air quality problem in the Pearl River Delta, China, we present a virtual geographic environment (VGE) system to integrate multiscale meteorological and air quality models for policy making. It is a comprehensive modeling tool to aid decision makers and various stakeholders to participate in air quality management by providing geographic visualizations and friendly interfaces. With nested multiscale models, a synthetic understanding of cross-boundary air quality processes can be captured to understand both regional and local effects. With the help of Linux-Apache-MySQL-Perl (LAMP) architecture, users can manage and retrieve modeling data and model parameterizations to reach a consensus on the simulation results and share modeling knowledge. Aided by a high-resolved emission inventory, such a multiscale system enables practical applications for various scenarios. As a case study, the system was applied to simulate and analyze the SO2 concentration process and local contribution in the Hong Kong Special Administrative Region (HKSAR) based on hourly simulation results with spatial resolutions of 0.5 and 3 km from multiscale models. The results from the multiscale modeling and the limited local contribution suggest that Hong Kong and the surrounding region should closely cooperate to develop a better environment.