A GIS approach to analyzing microclimate variations and the urban heat island phenomenon in Hong Kong

Abstract

Urbanization is known to cause significant changes to the properties of local climate. The Urban Heat Island (UHI) has been shown to add an additional burden to the impacts of global warming, impose risks on human health, and indirectly worsen air and water qualities. Studies have shown that urban areas, compared to rural areas with less artificial buildup, experience higher local temperatures as a result of UHI. Hong Kong is one of the most densely populated cities in the world with a high proportion of her population residing in urbanized areas. The hot and humid sub-tropical monsoon climate echoed with densely packed high-rise buildings and a high traffic volume can give rise to severe local thermal discomfort.

To fully characterize the spatial and temporal aspects of UHI within Hong Kong’s urban areas, this study deployed 58 small, durable and low-cost logging sensors to measure road-side temperature/humidity measurements at 15-minute time intervals. The data collection was conducted over 17 consecutive days in the summer (September) and repeated in the winter (January) at two typical urban locations: Mong Kok and Causeway Bay. By employing the geographic information system (GIS) and global positioning system (GPS) software, the sampled data were mapped against urban structures and land uses to investigate the urban microclimate and the role of various environmental factors at the sampling locations. Official meteorological conditions for the duration of study were obtained from the Hong Kong Observatory to validate the sensors and to explore microclimate variations by comparing official (urban and rural monitoring stations) and sampled readings. Temporal variations of urban temperature were examined for daily, weekly and seasonal differences. Spatial and temporal variations were examined using spatial interpolation and hotspot analysis, as well as in a 3D representation with building models for better visualization.

This empirical study establishes the methodological feasibility and reliability of employing small and inexpensive logging sensors for widespread deployment in places with hot and humid subtropical climates. The validation outcome indicated that road-side and control measurements were strongly correlated (r>0.81) and their average difference was well within the ±1 C measurement accuracy claimed by the manufacturer. Significant microclimate variations within the urban area were observed and characterized by five environmental settings: major road, secondary road, public activity area, green park and tunnel. Temporal variations of UHI were evident for all time scales, with the daily highest UHI at around midnight and daily lowest at around noon to early afternoon which also exhibited urban cool island (UCI) effects. A UHI Threat Rating was introduced for better understanding of microclimate variations and easier appreciation of hotspots. A 3D-GIS building model enhanced spatial-temporal analysis of UHI over a near “real” and 3D environment. The study and its methodology set a sound foundation and provide essential framework for future studies on microclimate variation and UHI effects. Given that modern cities have mixed land uses and are increasingly vertical, this micro-level study helps address local issues on human comfort and brings in the broader picture of environmental health in an urban setting.