Dissertation Defense
Student: Kellyssa Loren de Lima Alves
Program: Meteorology
Title: “Numerical simulation with Local Climate Zones of urban heat island in the Metropolitan Region of Sao Paulo"
Advisor: Profa. Dra. Adalgiza Fornaro - IAG/USP
Judging comitee:
- Profa. Dra. Rita Yuri Ynoue – Presidente e Orientadora – IAG/USP
- Prof. Dr. Renato Luiz Sobral Anelli – FAU/Universidade Presbiteriana Mackenzie
- Profa. Dra. Flávia Noronha Dutra Ribeiro - EACH/USP
Abstract: The expansion of cities has intensified in recent decades, leading to urban impacts. The urban heat island (UHI), characterized by higher temperatures in cities compared to surrounding rural areas, is one of the most well-known consequences of urbanization on the local climate. In this context, the main objective of this study is to investigate the occurrence of UHI in the Metropolitan Area of São Paulo (MASP), highlighting the contribution of urban morphology and meteorological conditions to the phenomenon's evolution. To this end, the study used the Weather Research and Forecasting (WRF) numerical model and classified the urban space into Local Climate Zones (LCZ). In the MASP, the analysis identified 8 LCZ classes, ranging from high, compact buildings with minimal vegetation (LCZ 1, in the central part of São Paulo) to low, scattered buildings with vegetation (LCZ 9, in the peripheral areas of the MASP). Simulations were conducted from September 2014 to 2023 to evaluate the spatial variation of air temperature in the MASP and the behavior of the UHI relative to different rural reference points. Considering all urban locations, the highest average temperature during the period was observed in Jardim Paulista (LCZ 1) at 21.5 ± 5.1 °C, while Ribeirão Pires (LCZ 6) recorded the lowest, at 18.7 ± 4.5 °C. However, locations within the same LCZ type but in different regions of the MASP exhibited distinct results, indicating the influence of additional factors beyond local characteristics in shaping the observed patterns. A cluster analysis of the urban locations revealed a temperature gradient from southeast to northwest across the MASP. On days with meteorological conditions conducive to UHI formation (clear skies and calm winds), UHI intensity was higher. Conversely, the UHI effect was weaker on unfavorable days, with maximum intensities up to 4.8 °C lower than those recorded on favorable days. Additionally, the phenomenon was more pronounced at night, with peak intensity occurring predominantly at 18:00 local time (LT). Throughout the day, UHI intensities were lower. Regarding rural reference points, the importance of their location in assessing UHI magnitude was evident. Since the sea breeze mitigates air temperatures in areas closer to the coast, UHI intensity can vary depending on the choice of the rural reference point. Comparisons between urban locations and the southeastern rural point resulted in a more substantial UHI effect (with temperature differences of up to 9.7 ± 2.6 °C at 18:00 LT in the warmest areas) than when using the northwestern rural point (5.0 ± 2.2 °C at the same time for the exact locations). Observational data corroborated these findings, emphasizing the relevance and accuracy of atmospheric modeling in urban studies. Furthermore, areas with higher building density and lower sky view factors (LCZ 1 to 3) exhibited smaller reductions in UHI magnitude at the end of the nighttime period (06:00 LT) compared to less dense areas with more scattered buildings (LCZ 6 and 9). These results indicate the role of buildings in heat storage and delayed release, keeping urban areas warmer and generating intra-urban temperature differences. Therefore, UHI analysis should consider synoptic-scale meteorological conditions, regional circulation patterns, the choice of rural reference point, and urban morphology.
Keywords: urban heat island, Local Climate Zones, WRF