Anthropogenic heat flux estimation over Hong Kong : a multi modelling approach

Abstract

The main objective of this study is to measure anthropogenic heat flux using satellite remote sensing, and evaluate the results using GIS-based modelling. In cities, anthropogenic heat emissions are sourced mainly for buildings, vehicles and human metabolic activities. The emissions vary significantly over time and space, and are not easily measured, so detailed models of anthropogenic heat emissions are not available for most cities. Previous studies have mainly used three approaches: inventory, energy balance enclosure, and building energy models. The present study is the first to model anthropogenic heat using three different approaches. Thermal infrared Images from the Advanced Spaceborne Thermal Emission and Reflection Radiometer, Level 1B (ASTER, L1B) satellite are used to estimated anthropogenic heat emissions using the energy balance approach. A new technique is developed to calculate the emissivity with a higher spatial resolution. This study measures urban morphometric parameters using the Geographic Information Systems (GIS) at a much finer scale, compared to previous studies, which only calculate approximate morphometric parameters of the city. Anthropogenic heat emissions are estimated from six different ASTER images of Hong Kong captured between 2007 and 2009. These images represent diurnal (day and night) periods and seasons (summer and winter). In this study, a global anthropogenic heat model called the Large-scale Urban Consumption of energY (LUCY) is adopted for application to Hong Kong, with the spatial resolution increased to 100 m from 2.5-arc minutes (~5-km spatial resolution). The anthropogenic heat emissions in LUCY greatly depend on population density data. Therefore, a new population scheme is introduced, in which daytime population data are included so as to consider the migration of population within grid cells at any given time. Other input parameters like energy consumption, air temperature, traffic information, and temperature schemes are also modified to match Hong Kong conditions. A GIS-based model using a bottom-up approach is utilized to account for anthropogenic heat emitted from each building block and every road in Hong Kong. Since these emission sources are captured at a finer resolution in the GIS-based model, the model data are used as the reference for evaluating the remote sensing and LUCY models. It may be noted that the model estimations are limited to the times of satellite image acquisition (~ 11am for the day and 11pm for the night) on the given date. In addition, the anthropogenic heat values estimated using the three different models is limited to 100 m spatial resolution. Anthropogenic heat emissions were found to be higher at day compared to nighttime emissions. Likewise, summer emissions were higher than winter ones. In all models, the highest emissions were observed in grid cells with tall commercial, business and residential buildings. Although the intensity levels differed across models, the spatial distributions remained essentially the same. Finally, a significant correlation between anthropogenic heat and UHI was noted when the results of the models were correlated with urban-morphometric parameters and ASTER surface temperature measurements. The anthropogenic heat values estimated using remote sensing agree well with those derived from the GIS-based model. However, there are certain differences in terms of advantages and disadvantages. Overall, the results from all the three models are usable as inputs to climate models. This should be useful to urban planners and other agencies in studies related to UHI.