Towards an Integrated Pedestrian Thermal Comfort Assessment System: An Agent-Based Approach

Pedestrians are important to sustainable cities. How to make outdoor spaces attractive to pedestrians has long been a primary concern in urban planning and design. The outdoor microclimate is found to be an important factor in affecting pedestrian's behaviors, i.e. space use. One key issue in this respect, is to evaluate the impact of the microclimate on pedestrian's thermal comfort condition. An effective tool that allows detailed investigation of pedestrian's thermal comfort and also provides implications of space use is in great need.;Effective assessment of pedestrian's thermal comfort is both a research challenge and a practical crux. Traditional bio-meteorological indicators such as PMV are only suitable for steady state subjects therefore will normally result in discrepancies when applied to transient conditions. Evaluation methodologies are also absent in the behavioral aspects: pedestrian traffic is always directed by certain activities, suggesting that their thermal comfort assessment should bear considerations of pedestrian's behaviors in order to provide pertinent planning implications.;The thesis establishes a conceptual framework for "context-sensitive" assessment on the meso-scale, and proposes to link the assessment of pedestrian's thermal comfort condition to the comprehensive urban context and therefore their activities . An integrated system is implemented with a comprehensive representation of the urban environment. A promising modeling technique, the agent-based modeling approach is taken to overcome the limitations imposed by traditional top-down models . The agent-based model simulates pedestrian's behaviors from the disaggregated scale, allowing detailed investigations of individual pedestrian's dynamic thermal comfort conditions . By examining system behavior emerged from the interactions of different individuals, implications for planning could also be revealed on the aggregated scale. With support of geographical information system, the system could be applied in real world context with complex urban environment. A detailed case study using a downtown area in Hong Kong as an example is explored to demonstrate the advantage of the agent-based modeling approach.;Through the case study, it is found that the presented agent-based simulation system can bring dynamic and behavioral understandings in investigating pedestrian's thermal comfort, which are more relevant in outdoor space planning but is not possible using traditional thermal comfort assessment methods. This suggests that the agent-based modeling approach is a good complement to traditional static and aggregated methods.;To sum up, the thesis work has significance from conceptual, technical and practical perspectives in terms of developing a spatial decision support tool. Conceptually, the thesis identifies the need for "context-sensitive" assessment in investigating pedestrian's thermal comfort, and adopts the agent-based modeling paradigm to account for pedestrian's dynamic and individualized features. Technically, an integrated GIS-based computational system is implemented to simulate pedestrian movement and assess pedestrian's thermal comfort from both static and dynamic aspects. Practically , the system can be applied in real world context and provide implications in urban design.