Electronic Thesis and Dissertation Repository

Thesis Format



Master of Science


Geography and Environment


Dr James A Voogt


This study aimed to better understand the spatiotemporal behaviors of "incomplete" surface temperatures, a subset of all the active surfaces in urban areas, that are relevant for outdoor thermal comfort assessments. This study answered the following questions: 1. What is the temperature of "incomplete surfaces" that are most relevant in urban thermal comfort assessments, and what is its spatiotemporal behavior in different urban neighborhoods? 2. How are the relevant incomplete temperatures related to nadir view remotely sensed surface temperatures (Tplan)? By combining distributions of wall temperatures from TUF-3D, an urban energy balance model, and horizontal surface temperatures from airborne remotely observed data with building polygons, the study defines and estimates the pedestrian temperature (Tped), an application-relevant incomplete surface temperatures for four Local Climate Zones (LCZ 5,6,7, and 8) in Phoenix, USA for both daytime and nighttime. The results show that Tped can be up to 8oC less than Tplan. The results also show that Tped varies within different sub areas of a large study area, and also it varies between different study areas. For daytime Tplan, LCZ 6 recorded the highest temperature value while LCZ 8 recorded the lowest temperature value, even though the difference between the two temperatures is very small.

Summary for Lay Audience

Urbanization significantly impacts the temperatures in urban areas. This transformation leads to an increase in temperatures, a phenomenon known as the "urban heat island" effect. This effect can present potential heat-related challenges for city inhabitants, such as health issues related to overheating and the increased energy consumption associated with cooling demands. Traditionally, remote sensing technologies like satellites have been employed by many researchers to study urban climates and monitor surface temperatures. However, viewing an urban area using remote sensing leads to a discrepancy between the surfaces observed by satellites, which preferentially view rooftops and ground areas, and the surfaces that are most relevant for specific urban heat applications. In the case of outdoor thermal comfort, these surfaces consist of the combination of wall and ground temperatures that emit heat directly to pedestrians. This combination of surfaces is “incomplete” with respect to the full three-dimensional urban surface as is the (different) combination of surfaces seen by remote sensing. The study was designed to understand the behavior of these relevant "incomplete" surface temperatures in urban areas, especially how they vary in time and space across different neighborhoods. To accomplish this, the study utilized a combination of airborne thermal remote sensing, which provided high-resolution thermal images, and LiDAR-derived building data, which offered detailed information about building characteristics. An energy balance model known as TUF-3D was used to generate wall temperatures of buildings in the study areas because these surfaces are not otherwise ‘seen’ by most remote sensing instruments. The focus of the study was to answer two main questions: What is the temperature of these incomplete surfaces that are most relevant to urban thermal comfort, and how do they vary across different parts of the city? Furthermore, how are these temperatures related to the temperatures seen in the straight-down view provided by remote sensing satellites? To answer these questions, the study focused on four Local Climate Zones in Phoenix, USA, using high-resolution thermal imagery captured between July 12-15, 2011. Through 4 a combination of wall temperatures from the TUF-3D model and horizontal surface temperatures from the airborne remote sensing data, the study defined and estimated the application-relevant incomplete surface temperatures. The results of the study reveal that researchers who depend solely on satellite view surface temperatures tend to overestimate the temperatures experienced by pedestrians by more than 8 degrees Celsius. This discrepancy has significant implications for understanding and mitigating urban heat challenges. Furthermore, the study developed first-of-their-kind multiple regression models that can be used by researchers to predict pedestrian-level temperatures. This model uses information about the plan temperatures, the cumulative wall areas in their research zone, and the ratio of plan to area, offering a promising predictive tool for future urban climate studies. Other results from the study show that heat load on pedestrian varies with the geometry of buildings in an area, with areas with buildings closely packed together being more thermally comfortable.