Background: Physiological equivalent temperature (PET), a thermophysiological indicator, has been developed to quantify thermal stress from the environment. However, studies focusing on spatial correlation of small areas within cities are limited. Aims: To construct PET in small geographical areas with a view to quantifying extreme thermal environments in Hong Kong. Methods: We constructed the PET from the Hong Kong Urban Climate Map (UCMap), which was created by the Planning Department in 2006 containingeight classes of thermal conditions at 100m×100m resolution (referred as grid hereafter). First, based on eight PET values computed from the thermal conditions including meteorological variables for each class,we assigned a PET value for each grid associated with a particular class. Second, we averaged the PET values of all the grids within each Tertiary Planning Unit (TPU), which is the smallest geographical demarcation for urban planning, to obtain the typical PET for all TPUof Hong Kong. We then calculated univariate Moran's I and performed Local Indicators of Spatial Association (LISA) analysis to detect global and local spatial variations of PET, respectively. Results: The PET in 145 TPU ranged from 26.3 °C to 31.3 °C. Univariate Moran's I was 0.6118, indicating positive autocorrelation of thermal conditions among adjacent TPU. The LISA analysis identified 39 hot spots (i.e. TPU with high PET are surrounded by TPU with high PET), 62 cold spots (i.e. TPU with low PET are surrounded by TPU with low PET), and 44 unclassified areas (i.e. TPU with high PET are surrounded by TPU with low PET or vice versa). The hot spots were mostly located in TPU with high population density where vehicular traffic flow and building density are high, and green areas are sparse. Cold spots were found in TPU of sub-urban areas where open spaces are prevalent. Conclusions: PET is a useful indicator for thermal environment, and can be used to identify hot and cold spotswithin the city.
展开▼
