An evaluation of aster images for assessment of causative factors of Hong Kong's urban heat island

Abstract

The Urban Heat Island Effect appears when the cooling rate of a rural area is faster than that of an urban area due to urbanization, and it is particularly obvious during the nights and early mornings. Like other highly urbanized cities, Hong Kong experiences the urban heat island effect, whose formation may be affected by different urban and environmental parameters. Techniques adopted for previous research on the urban heat island in Hong Kong have adopted field campaigns and data from climatic stations. Remote sensing techniques combined with field data have only been applied recently, but only very few images are available for analysis due to difficulties of obtaining cloud free images over Hong Kong. Moreover, no research in Hong Kong has used satellite images to evaluate the influence of the natural and built environment on the urban heat island. Therefore, the main objective of this research is to evaluate the use of several ASTER thermal images at different times, combined with "in-situ" field data, for analysis of urban heat island effect in Hong Kong. The secondary objective is to evaluate the influence of different urban and environmental parameters on the heat island intensity in different regions, e.g. entire Hong Kong, Kowloon Peninsula, Sha Tin and Tai Wai, Tsuen Wan and rural Tuen Mun, at different scales of analysis by field campaigns, remote sensing and GIS techniques. Numerical models for predicting heat island intensity from some influential urban and environmental parameters are further derived based on the measured relationships between those parameters and the heat island intensity. In this research, numerical models for forecasting the urban heat island in several parts of Hong Kong are developed. The average optimal scales for predicting heat island intensity over the Kowloon Peninsula (50 m), Sha Tin and Tai Wai (150 m), Tsuen Wan (120 m), rural Tuen Mun (110 m) and the whole of Hong Kong (70 m), derived from these numerical models are demonstrated. As most parameters included in the models are important elements in urban planning, and significant relationships between heat island intensity and these parameters are observed in this research, the models are useful for predicting the probable climatic implications of future planning decisions, e.g. for urban renewal projects and the development of new towns.