Geographic knowledge, which is the higher level of geographic information, is acquired on the basis of geographic background, process, and results. The expression and sharing of geographic knowledge is important in many applications for urban and environmental planning and for any application in territorial intelligence. Expressing and sharing geographic knowledge is a significant issue in current geographic research. This paper is based on the evolution of geographic language from maps to geographic information system (GIS) and virtual geographic environments (VGEs). It aims to review the capability of VGEs to support geographic knowledge expression and sharing. To investigate the characteristics of different geographic languages and thusexpress and share geographic knowledge, we divided geographic knowledge into three levels: factual knowledge, rules and control knowledge, and decision-oriented knowledge. The evolution of geographic languages and their capability to support geographic knowledge expression and sharing were reviewed by discussing the relationship between geographic world, geographic language, and geographic knowledge. VGEs have more power to support geographic knowledge expression and sharing compared with maps and GIS. The outstanding features of such new generation of geographic language were reviewed considering their contributions to geographic knowledge expression and sharing. The features include interactive visualization, multi-channel perception, virtual reality technology, model base and model management, and spatially distributed collaboration. 1. Interactive visualization, multi-channel perception, and virtual reality technology enhance the expression of geographic world and geographic knowledge. The distance between geographic knowledge "writer" and "reader" is shortened. Thus, reaching a consensus on the expression and understanding of geographic knowledge becomes easy. Users in this virtual environment, which is the counterpart of the geographic world, can easily share geographic knowledge, unlike in traditional maps or dimension-reduced GIS. 2. Two cores of VGEs, data base and model base, are helpful in expressing and reusing geographic knowledge. Model base, which includes statistical models and dynamic geographic models, is the output from geographic facts and rules. Such geographic knowledge can be accumulated and reused by different users through an integrated model base. Users can analyze and simulate geographic phenomenon and process and acquire a high level of geographic knowledge to manage the geographic world. 3. Geo-spatial collaboration is developed in VGEs to support experts from multiple, spatially distributed disciplines and allow these experts to collaborate in conducting geographic simulation and analysis. In this way, they can share knowledge and obtain scientific cognition of the geographic world. Air quality problem is a crucial issue for the government of Hong Kong and the Pearl River Delta region. Thus, a VGE platform is developed for users from multiple disciplines to conduct air quality simulation and visual analysis by coupling meteorological and air quality models. In multiple perspectives, this case study analyzed the capability of VGEs in expressing and sharing geographic knowledge, with VGEs including virtual reality technology, geographic model base and management system, dynamic visualization, and geo-spatial collaboration technology. VGEs are under development and face key challenging issues. This paper is concluded with a discussion on future research concerning geographic knowledge creation, management, and application. © 2016, Science Press. All right reserved.
