Developing a high-resolution physiological equivalent temperature map to assess the spatial variation of thermal stress exposure for an elderly cohort of Chinese in Hong Kong

Abstract

Physiological Equivalent Temperature (PET) is an universal thermal stress measure which is calculated from air temperature (Ta), relative humidity (RH), wind velocity (WV) and mean radiant temperature (MRT) of an environment. Spatial analysis is essential to long term health effect study of thermal stress. To facilitate the spatial analysis on an elderly cohort living in urban areas of Hong Kong with over 10 years follow up, a high- resolution (100m x 100m) PET map which represents a typical pattern of thermal stress conditions during the baseline period (2000 – 2001) is suggested. Unfortunately and except for RH data, past meteorological data from the Hong Kong Observatory (HKO) are not of sufficient detail to take account of the complex terrain and dense urban morphology of Hong Kong. We propose to estimate meteorological surfaces of Ta, WV and MRT for a typical climate condition in Hong Kong at the baseline, and then calculate the PET at each grid cell. The geographic information systems is used to undertake the following processsing. The Ta surface is estimated from NASA Landsat thermal satellite images at the baseline and adjusted using the hourly Ta data from the HKO. The WV surface is estimated by Ground Coverage Ratio, which is highly correlated with WV ratio at pedestrian level, and adjusted with WV data from the HKO. The spatial difference of MRT, which is highly correlated to urban morphology, is estimated using Sky View Factor. The procedure is replicated using data in recent years to produce a PET map for 2008, which is compared against the Urban Climate Analysis Map of Hong Kong created by the Planning Department, to verify the proposed methodology. This methodology provides a feasible way to construct high-resolution map for thermal stress exposure in a city with complicated urban environment over longer time frame, and copes with the limitation of data availability in many long term environmental health researches. Acknowledgement: RGC grant 780512.