Dynamic visualization of multi-dimensional urban environmental data : a case study of spatio-temporal air pollution dispersion in Hong Kong

Abstract

Urban landscapes have been vertically expanding and severely densified in recent years. Environmental issues, especially urban air quality’s deterioration, are being paid close attention to by both academia and general public. Increasing air pollutants pose great threats to citizens’ health. The congested and three-dimensionally (3D) expanding urban spaces have made air pollution and risks of exposure more severe. Street canyons, a semi-enclosed environment where buildings line up continuously along a street, form a key element of cities. Air pollutants in street canyons reveal higher concentration levels in comparison to background values due to poor natural ventilation and proximity to traffic emission. However, residents can hardly examine the air quality information of their surrounding environment via current media. To better observe, interpret and communicate an environmental phenomenon such as air pollution dispersion that varies in space and time, visualization of environmental data in urban areas requires transformation from the plain two-dimensional (2D) to more realistic 3D representation, and from static to dynamic displays.

This research adopts a semantic approach to examining multi-dimensional environmental data of dynamic changes in urban areas, aiming to propose an operable technical route in dealing with explosively large volumes of data about the urban environment which have been rapidly changing in their measurement quantities, dimensional attributes and temporal variations. A case study of dynamic 3D visualization of PM2.5 concentration was conducted in urban Hong Kong to illustrate the approach. Air pollution dispersions in space and time were investigated based on field measurement data from street canyons and open sites. These data were transformed from points to surfaces, and then to dynamic full 3D volumetric formats in a navigable virtual environment to explore the feasibilities of visualization implementation at different levels of completion. When discussing 3D dynamic visualization, many issues have to be considered; for example, the time/efforts needed to create a 3D model, the resolution and levels of detail of the model, as well as the availability of source data and software.

This study has attempted to propose a practical scheme for producing dynamic 3D visualizations to display urban environmental data. It hopes to formalize the stages and design processes in making visualization models, with particular reference to displaying air pollutant movements within an urban space. This work has presented a guided approach to creating understandable and well-balanced 3D models for enriching information communication based on different accessibilities in data, resource, and implementation techniques. The results of 3D dynamic visualization presented in this study can function as a useful tool to examine environmental risks and to draw attention to specific sections of an urban area needing improvement. Proper environmental visualizations can facilitate assessments of health risks or relative vulnerability with respect to urban morphology, thereby enabling better urban planning policy making and development initiatives to protect public health.