University of Twente Student Theses
Analyzing the impacts of urban morphology on land surface temperature in European cities
Dev Roy, S. (2024) Analyzing the impacts of urban morphology on land surface temperature in European cities.
|
PDF
9MB |
| Abstract: | With projections indicating that nearly 70% of the global population will reside in urban areas by 2050, rapid urbanization is profoundly modifying land cover with built surfaces, along with the transformation of the urban morphology (UM) and exacerbating the thermal environment of cities. UM constantly evolves in response to people’s needs and local contexts, leading to diverse building structures and materials. These changes significantly impact the thermal environment, making especially cities hotter than their surrounding areas. Given the growing risks, it is essential to study the relationship between UM and land surface temperature (LST). This study focuses on LST over air temperature because LST directly affects near-surface air temperatures and offers broader spatial coverage as it can be mapped using thermal infrared (TIR) data. It is more consistently available across diverse urban environments compared to sparse and unevenly distributed air temperature data measured by sensors on ground. Previous research has mainly focused on landscape metrics, spectral indices, and surface attributes’ impacts on LST. They have often neglected finer-scale building-level analyses which are the most important urban elements having influence on heat patterns. Although some studies have considered 3D aspects, comprehensive 2D analyses are scarce due to the underutilization of building data, particularly in European cities, despite better data availability. This research addresses this gap by investigating the impacts of UM on daytime LST during the summer period in Paris, Rotterdam, Milan, and Vienna, chosen as illustrative examples. Utilizing NASA’s high-resolution ECOSTRESS data, the research conducts a hotspot analysis to identify areas with significant temperature anomalies and understand their relationship with specific land uses. A comprehensive set of thirty urban morphometrics (UMMs) is measured using momepy, an open-source python toolkit, to study the UM patterns of these cities at the building level. Afterwards, a Random Forest Regression (RFR) model is applied at a grid level of 70 by 70 meters (the spatial resolution of ECOSTRESS) to explore the relationship between the UMMs and LST. The hotspot analysis reveals that most of the hotspots across all these cities are mainly business parks, industrial estates and manufacturing units. While the coldspots are predominantly low-density residential areas. The RFR model effectively captures the underlying patterns and relationships between UMMs and LST, explaining over 80% of the variability in LST across all cities. Key findings of the model highlight that mean height, orientation, alignment, building adjacency, and interbuilding-distance are the most influential UMMs across all cities. However, differences between cities also exist. For example, LST shows a positive correlation with mean building height in all cities except Paris. The patterns between orientation and LST varies, where Rotterdam and Milan exhibit a negative relationship, while Paris displays a contrasting pattern. Building adjacency and alignment demonstrated non-linear cooling effects on the urban environment across all cities. Lastly, the qualitative validation confirms that the variety of UMMs used in this study is both informative and crucial in this research field. The study outlines several implications for improving thermal comfort in existing as well as new urban developments across European cities. The limitation of the study encompasses the fact that the influence of UMMs do not demonstrate direct causality with LST and depend on various other factors such as building materials and wind patterns. Further research should aim to explore and deepen the understanding of the relatively novel UMMs used in this study, particularly regarding their influence on LST, as they have not been extensively explored in existing literature. |
| Item Type: | Essay (Master) |
| Faculty: | ITC: Faculty of Geo-information Science and Earth Observation |
| Subject: | 38 earth sciences, 43 environmental science, 74 (human) geography, cartography, town and country planning, demography |
| Programme: | Geoinformation Science and Earth Observation MSc (75014) |
| Link to this item: | https://purl.utwente.nl/essays/100706 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page