Impact of the Total Area and Spatial Distribution of Green Infrastructure on the Canopy Urban Heat Island at the Metropolitan Region of São Paulo - A Numerical StudyAbstract
A rapid verticalization to accommodate the citizens of the Metropolitan Region of São Paulo is altering the balance of radiation and atmospheric heat, highlighting the need to understand the impact that green and built infrastructure have on the canopy urban heat island phenomenon. This meteorological phenomenon occurs mainly due to the difference in landscape between urban and rural areas. Hypothetical scenarios with different green profiles were simulated using the WRF model coupled with SLUCM, and their results were compared to the current scenario using numerical data in order to observe the impact of green infrastructure. Comparison using output data showed that the total area of green infrastructure has great potential in reducing the intensity of the canopy urban heat island. The scenario with the largest total area and highest dispersion of green infrastructure recorded average urban temperatures 1.2 °C to 1.9 °C lower than the current scenario. Understanding the behavior of green infrastructure and its benefits is important for the development of municipal public policies that are in line with sustainable goals, and explicitly the relevance of urban parks and squares for local thermal regulation.
Keywords
green infrastructure; canopy layer urban heat island; urban climate; Metropolitan Region of São Paulo; urbanization effects; sustainable urban planning
Resumo
Uma rápida verticalização para acomodar os cidadãos da Região Metropolitana de São Paulo está alterando o balanço de radiação e calor atmosférico, destacando a necessidade de entender o impacto que a infraestrutura verde e construída têm no fenômeno da ilha de calor urbana de dossel. Esse fenômeno meteorológico ocorre principalmente devido à diferença na paisagem entre áreas urbanas e rurais. Cenários hipotéticos com diferentes perfis verdes foram simulados usando o modelo WRF acoplado ao SLUCM, e seus resultados foram comparados ao cenário atual usando dados numéricos para observar o impacto da infraestrutura verde. A comparação usando os dados de saída mostrou que a área total de infraestrutura verde tem grande potencial para reduzir a intensidade da ilha de calor urbana de dossel. O cenário com a maior área total e maior dispersão de infraestrutura verde registrou temperaturas urbanas médias entre 1,2 °C e 1,9 °C mais baixas do que o cenário atual. Compreender o comportamento da infraestrutura verde e seus benefícios é importante para o desenvolvimento de políticas públicas municipais alinhadas aos objetivos sustentáveis, e destaca explicitamente a relevância de parques e praças urbanas para a regulação térmica local.
Palavras-chave
infraestrutura verde; ilha de calor urbana de dossel; clima urbano; Região Metropolitana de São Paulo; efeitos da urbanização; planejamento urbano sustentável
1. Introduction
For centuries, the development of societies into urban configurations has been leading to deep changes in land use across geographical areas. This trend, chosen by humankind as the primary path to development, has driven consequences in the local environment that have been discussed for a long time. It is believed that the English scientist Luke Howard, in 1818, was the first to write about air temperature differences between a city and its surrounding undeveloped rural areas, being London as study area (Howard, 2007 [1833]). The topic only received a scientific term a century after Howard's study. The term, “urban heat island” (UHI), was first coined by another Englishman, Gordon Manley, and the definition is that urban heat island occurs when the urban area temperature is higher that its surrounding area (Manley, 1958).
Nowadays, the term ‘urban heat island’ is still used, but it has been segmented into three different kinds of UHI based on the surface where the temperature is being measured: surface urban heat island (UHIS), related to urban surfaces; canopy layer urban heat island (UHICL), related to the canopy layer; and boundary layer urban heat island (UHIBL), related to the planetary boundary layer (PBL). UHIS results from heat absorption by urban surfaces like asphalt and buildings, particularly noticeable during daytime (Arnfield, 2003; Oke, 1973). UHICL are influenced by vegetation density and surface characteristics, impacting heat and moisture exchange within the canopy layer (Grimmond and Oke, 1999). UHIBL alters atmospheric conditions in the PBL, affecting temperature and wind patterns, often exacerbating temperature differences between urban and rural areas (Sailor, 2011). The intensity of the UHICL is higher during the night, while the UHIS reaches its highest intensities during the day (Lombardo, 1985; Roth et al, 1989). However, in scenarios where a sea breeze occurs over the urban region, as demonstrated by Umezaki (2020), the peak of the urban heat island at the canopy layer can coincide with the occurrence of this event. In her study of the Metropolitan Region of São Paulo (MRSP), which experiences a sea breeze, the peak UHICL was observed mostly at 7 pm. This happens because the influx of relatively cooler air causes a temperature drop in the vegetated areas adjacent to the city, thereby establishing a greater temperature difference between the urban canopy layer and the surrounding rural areas (Umezaki, 2020).
Hosting a big population with dense urbanization, the Metropolitan Region of São Paulo has been a good study case for some insightful studies about urban climate due to its vast size, high population density, and significant economic activity. Lombardo (1985) was the first to scientifically describe the occurrence of this urban event in São Paulo and its characteristics. She pointed out the city's location and the daily climate conditions as contributing factors, but settled the urban land use in São Paulo as a major driving force, because of the high heat storage capacity in the city's buildings. Lombardo (1985) also discussed the role of green infrastructure in regulating urban temperatures, highlighting the lack of vegetation cover as another significant contributor to São Paulo's rising UHII. Barros and Lombardo (2016) shows leaf area index (LAI) has deep influences on urban surface temperature, greener neighborhoods with LAI greater than 0.67 were 8 °C colder than neighborhoods with LAI between 0 and 0.01. Their study concluded the deterministic role of trees in reducing surface urban temperatures in the MRSP. Ferreira et al. (2011) also registered significant surface layer temperature differences between urban and rural areas, with a 5 °C difference in the winter months and a 2 °C difference in the summer months. As seen, there is a lack of recent studies addressing the air temperature differences between urban and rural areas within the Metropolitan Region of São Paulo. Ferreira et al. (2011) study dates back to 2011, and even though he does briefly discuss canopy temperatures in his study, the MRSP has undergone significant urbanization changes since then.
This study aims to address the identified gap by investigating the influence of urban green infrastructure on the canopy layer urban heat island phenomenon within the Metropolitan Region of São Paulo (MRSP). Through valuable data and images, this research seeks to contribute to ongoing discussions about sustainable urban development. Furthermore, understanding UHI is crucial for several reasons. Firstly, UHI can significantly impact human health by exacerbating heat stress and related illnesses, particularly among vulnerable populations (Harlan et al., 2006; Harlan et al., 2012). Secondly, UHI can contribute to higher energy consumption as a result of increased demand for air conditioning to mitigate heat effects (Akbari et al., 2008; Sailor, 2011). Lastly, comprehending UHI dynamics is essential for effective urban planning aimed at creating cooler and more livable cities.
2. Methodology
The following subsections will describe the study area, the period selection, the model configuration and validation, the scenarios we chose for the simulations, and the canopy layer urban heat island intensity (UHIICL) calculation.
2.1. Study area
The study area was the Metropolitan Region of São Paulo (MRSP), an urban space composed of 39 municipalities with a significant concentration of buildings and more than 20 million inhabitants (Governo do Estado de São Paulo, n.d.; IBGE, 2022). We used three nested concentric domains. The parent domain had a horizontal spacing of 10 km by 10 km, the intermediate domain had a horizontal spacing of 2 km by 2 km, and the child domain had a horizontal spacing of 0.5 km by 0.5 km.
As it can be seen in Fig. 1, the innermost domain encompasses the vast majority of its urban space, aligning its center with São Paulo city's downtown. Certain areas, primarily located in the eastern region of the MRSP, were excluded from the domain due to limitations in computer processing power.
Representation of the outermost, intermediate and innermost nested domains (-23°44’23.64” S, -46°55’5.16” W; 23°44’23.64” S, -46°21’22.68” W; 23°24’21.96” S, -46°55’5.16” W; 23°24’21.96” S, -46°21’22.68” W). Source: Google Earth, 2022 (Adapted).
2.2. Period selection
The date chosen for conducting the simulations was 10th August 2021. This selection was made because it was a sunny, hot winter day without the presence of sea breezes. In Ferreira et al. (2011), the results showed that the intensity of the urban heat island was higher during the month of August, which guided the choice of this month for the simulations. Additionally, the MRSP experiences a dry winter, which facilitates the selection of a cloudless day with full or near-total solar incidence on the surface. Hourly averages of global solar radiation from the Environmental Company of the State of São Paulo (CETESB) Parque Dom Pedro II (P.D. PEDRO II) station (23°32’42.18” S; 46°37’39.59” W), obtained through the CETESB platform QUALAR, indicated little to no cloud cover, reaching 801 W.m-2at noon (Fig. 2c), whereas the monthly average was 577 W.m-2 (CETESB, 2021). The maximum hourly temperature of the day occurred at 15:00 local time (LT) and reached 28.8 °C (Fig. 2a), whereas the monthly average for 15:00 LT was 24.6 °C (CETESB, 2021). Lastly, the day was not influenced by the sea breeze, which helps cool the MRSP and changes the UHIICL hourly pattern (Umezaki, 2020). The nearest coastal area to the region is southeast, and on this day, during the hours when the sea breeze would occur (16 LT. to 18LT), the region experienced predominantly northwest winds (Fig. 2d), specifically from the 325° direction (CETESB, 2021). Figure 3 shows the wind speed and direction during the observation period.
Average air temperature, relative humidity, and global solar radiation from meteorological station CETESB P.D. PEDRO II. Left graphs show mean values for August 10th yearly and weekly averages from August 7th to 13th. The right graphs show hourly data for August 10th, 2021.
Wind speed and direction data on 10th August 2021 from meteorological station CETESB P. D. PEDRO II.
2.3. Model configuration
In this study, we used the Weather Research and Forecasting (WRF) mesoscale numerical weather prediction model, version 4.3.3 (Skamarock, 2021). The main parameterizations were: RRTMG scheme for longwave radiation, Goddard scheme for shortwave radiation, YSU scheme for boundary-layer, Unified Noah land-surface model with 4 soil layers. We coupled the WRF model with the Single-Layer Urban Canopy Model (SLUCM) (Chen et al., 2011) to improve the accuracy of urban energy balance prediction compared to the simplified default treatment in the model. We used the LCZ classification based on the WUDAPT project (Ching et al., 2018). Additionally, the NOAH Land Surface Model (He et al., 2023) was integrated with WRF and SLUCM to handle surface processes such as land use, vegetation, heat flux, surface energy balance, and provide information to the WRF model.
The model was configured to simulate the air temperature at a height of 2 m for a period of three days, from August 9th, 2021, to August 11th, 2021.
2.3.1. Model validation
To validate the model used in this research, we collected temperature data from three meteorological stations located in the city of São Paulo on August 10th, 2021. We ran a simulation round for the temperature at the locations of these three stations and compared the model's results with the observed data (Fig. 4).
The Mean Squared Error (MSE) for the National Institute of Meteorology (INMET) Interlagos was 4.1 °C, for the INMET Mirante was 1.9 °C, and for CETESB P. D. PEDRO II was 2.9 °C. Respectively, the Root Mean Squared Error (RMSE) was: 2 °C, 1.4 °C, and 1.7 °C. In general, a low MSE is usually under 1 °C, a moderate MSE is located between 1 °C and 4 °C, and a high MSE goes above 4 °C.
The model's response seems to be delayed. It could not capture the early afternoon temperature peak, nor the temperature drop in the sunset. It is interesting to note that in the early morning the model anticipated an unexpected temperature rise at INMET Mirante.
2.4. Scenarios
For urban land use, the Local Climate Zones classification (Stewart and Oke, 2012) was used, which includes ten types of urban land occupations. The configuration for the Metropolitan Region of São Paulo was extracted from the LCZ Generator portal of the World Urban Database and Access Portal Tools (WUDAPT, 2022) submitted on July 10, 2021, with a representativeness accuracy of 49%. The LCZs “2. Compact midrise” and “7. Lightweight low-rise” were not present and the predominant LCZ class was “3. Compact low-rise”. For rural land use, water bodies, and urban areas not assimilated by WUDAPT, we used the “MODIFIED_MODIS_NOAH” classification from the National Aeronautics and Space Administration (NASA, n.d.).
The study on the impact of improved green infrastructure in the MRSP worked with four scenarios, including one control scenario (called “current” in Fig. 5) and three idealized scenarios. In the “Greener” scenario, the LCZs “1. Compact high-rise” and “3. Compact low-rise” were replaced with their versions that have improved green infrastructure and increased spacing between buildings, namely “4. Open high-rise” and “6. Open low-rise”, respectively. In the “Intercalated” scenario there was an alternation between urban land occupations with and without green infrastructure in situations where grid cells were followed by urban land occupations without green infrastructure. In the “Boundary” scenario the MRSP was divided diagonally into two parts: the southwest and northeast regions. In the northeast region, all the LCZs “1. Compact high-rise” and “3. Compact low-rise” were transformed into “4. Open high-rise” and “6. Open low-rise”, respectively. We created the “Intercalated” and “Boundary” scenarios to investigate the influence of homogeneously increasing green infrastructure versus prioritizing an area. Table 1 shows the number of grid cells in urban classes for each scenario.
The simulated scenarios were not created to represent a realistic urban profile, or to set goals for urban planners, but rather to understand the local climate impact that can result from different urbanization settings. Some scenarios are exaggerated, with the focus on magnifying and isolating the effects of urbanization to better comprehend the urban heat island phenomenon. This study does not aim to recommend a specific green urbanization configuration, but rather to illustrate the potential outcomes in the city if such profiles were real.
2.5. Calculation of the canopy urban heat island intensity
The understanding of the impact of improved green infrastructure on the intensity of the UHICL in the Metropolitan Region of São Paulo will be achieved by comparing the UHIICL values across scenarios regarding the innermost domain. The calculation involves the difference between the average urban temperature (considering all urban land classes) and the average rural temperature (considering all non-urban land classes). The urban land classes considered include all LCZs and the “Urban and Built-Up” class from MODIS. The rural land classes encompass all classes present in the MODIS parameter except for “Urban and Built-Up”. Water bodies account for 0.9% of the land use and cover change area and are not included in the equation.
The equation is:
where UHIICL represents the urban heat island intensity in degrees Celsius; Tu is the average temperature of all urban land classes at 2 m of altitude in degrees Celsius; and Tr is the average temperature of all rural land classes at 2 m of altitude in degrees Celsius.
3. Results and discussion
Based on the data extracted from the simulations, it was observed that changing the green infrastructure profiles had a significant impact on the mesoscale characteristics of the Metropolitan Region of São Paulo (Fig. 6). This corroborates with Barros and Lombardo (2016) findings, which concluded the deterministic nature of trees in the urban temperature of MRSP. Increasing the quantity of trees and permeable layers by transitioning from clustered Land Cover Zones (LCZs) to spaced ones resulted in a significant decrease in urban temperature. Thus, as seen in Table 2, the greener scenario recorded the average lowest maximum urban temperature of 19.1 °C, and the average lowest minimum urban temperature of 12.8 °C. The current scenario recorded the average highest maximum urban temperature of 20.3 °C and the average highest minimum urban temperature of 14.7 °C. The intercalated scenario registered an average maximum urban temperature of 19.7 °C and an average minimum urban temperature of 13.8 °C. Finally, the boundary scenario recorded an average maximum urban temperature of 19.6 °C and an average minimum urban temperature of 13.7 °C.
Average air temperature difference (°C) comparison between the average maximum and the average minimum urban temperatures of the greener, intercalated, and boundary scenarios to the current scenario.
It was expected that the lowest temperatures would occur in the green scenario due to its larger total area and greater dispersion of green infrastructure. This is because trees and vegetated permeable surfaces affect the urban heat balance. One way this occurs is through the transpiration of aerial organs, such as leaves and stems, in the response to high temperatures. The incidence of light heats up these structures, and the liquid water contained in them undergoes vaporization. Stomata opens to release water vapor and the contained heat, causing the leaf or stem to cool (Gupta et al., 2018; Schulze et al., 1973). The lack of the greenness can be associated with a great portion of incident solar radiation being transformed into sensible heat due to the lack of water in the area exposed to sunlight. Sensible heat is a form of energy that fuels the movement of molecules, causing them to move around and collide with each other. The intensity of this molecular movement is measured as temperature. In contrast, incident energy can also be transformed into latent heat energy, which induces a change in the physical phase of the material, such as the transformation of liquid water into water vapor (FAO, 2005). Therefore, unlike the sensible heat, the conversion of solar incident energy into latent heat does not raise the temperature of the surrounding area as the air temperature. Therefore, greener areas cause a higher percentage of incident solar radiation to be converted into latent heat compared to highly built-up and densely populated areas.
Regarding the differences in rural temperatures, once again, the greener scenario exhibited the lowest temperatures (Fig. 7). However, in contrast to the urban situation, the temperature differences between rural scenarios were extremely low.
Air temperature difference (°C) between the maximum and minimum rural temperatures of the greener, intercalated, and boundary scenarios compared to the current scenario.
From Figs. 6 and 7, it can be observed that the temperature differences between the intercalated and boundary scenarios are not significant. The scenarios have similar proportions of total compact and spaced occupation areas, with the dispersion of green infrastructure being the main distinction between the two cases. Therefore, the collected data indicate that the rate of dispersion of urban green infrastructure in the MRSP may have little influence on the average urban temperature.
The daytime temperature maps show that intercalated and boundary scenarios exhibit similar patch patterns and tones (Fig. 8). It can be observed that differences in green infrastructure dispersion in scenarios with very similar total green infrastructure areas do not result in clearly distinct maps of daytime average temperature. Therefore, Fig. 8 supports the same observations made in Figs. 6 and 7 regarding the relationship between different dispersion rates within the same total vegetation area. It is interesting to note that the intercalated scenario land use pattern shown in Fig. 5 is reflected in the local temperature map on Fig. 8. This means that locally presence of trees can change neighborhood daytime temperatures, although it does not have a substantial impact on regional temperature as seen on Figs. 6 and 7.
Air temperature at 2 m of altitude maps during the daytime period considering the average between 06:00 and 17:00 LT for scenarios (a) current; (b) greener; (c) intercalated; and (d) boundary.
Significant differences are observed in the scenarios when nighttime average temperature maps are generated (Fig. 9). It can be noted that the patterns in the nighttime average temperature maps exhibit more similarities in arrangement with the land occupations shown in Fig. 5 than the daytime average temperature maps. Unlike Fig. 8, now the intercalated and boundary scenarios show substantial differences. In the boundary scenario, unlike the intercalated scenario, the eastern region lacks compact occupations and only has spaced occupations, resulting in lighter temperature tones. On the other hand, in the western region, it is the boundary scenario that exhibits darker tones since it has almost twice the number of compact-style LCZs when compared to the intercalated scenario. This dispersion antagonism between the two scenarios, which leads to the relationship described earlier between the western and eastern regions, explains the similarities in average urban temperature observed in Fig. 6 between the two scenarios: the intercalated scenario has an intermediate tone throughout the MRSP, while the boundary scenario has a more intense tone on one side and a lighter tone on the other. The southwest-northeast boundary line created in the boundary scenario is clearly visible in Fig. 9. Given that the observed current scenario is dominated by compact occupations and the greener scenario is almost the opposite, with all compact occupations transformed into spaced ones, it was expected that the nighttime temperature maps would exhibit antagonistic patterns.
Air temperature at 2 m of altitude maps during the nighttime period considering the average from 01:00 to 05:00 LT and from 18:00 to 24:00 LT for scenarios (a) current; (b) greener; (c) intercalated; and (d) boundary.
All these differences in nighttime temperature between occupations with higher and lower proportions of building materials are in accordance with Winbourn et al. (2020). The author states that reduction of heat storage in the urban surface is proportional to the quality and quantity of green infrastructure present, with heat storage being higher in built surfaces than vegetated ones. One of the reasons is anthropogenic pavements usually have lower albedo than vegetated surfaces, thus absorbing more incident radiation. Additionally, Asaeda et al. (1996) conclude in their study that pavements with high thermal conductivity, such as asphalt and concrete, will dissipate more heat during the nighttime than materials with low thermal conductivity, such as soil, wood, and leaves (as observed in Fig. 9).
Following Oke's (1973) definition of urban heat island, the urban climatic event occurs when there is a difference between urban and rural temperatures. Therefore, Fig. 10 shows that all four scenarios exhibit this issue as they have positive values for the UHIICL.
Graph showing the hourly difference between the average temperature of all urban classes and the average temperature of all rural classes in the four scenarios (UHIICL).
The collected data demonstrate that the intensity of the UHICL is higher in scenarios with a smaller absolute area of green infrastructure. Similar to the temperature graphs, the characteristics of vegetation dispersion, as explored in the intercalated and boundary scenarios, may not have great impact on UHIICL in situations with the same total area of green infrastructure. All four scenarios reached peak UHIICL at 01:00 LT. The dissipation of sensible heat radiation by urban structures occurred 7 h and 12 min after sunset (which took place at 17:48 LT on the previous simulated day). The lowest UHIICL value occurred at 17:00 LT in the current scenario, measuring 1.3 °C, at 15:00 LT in the greener scenario with a value of 0.5 °C, and at 16:00 LT both the intercalated and boundary scenarios registered a temperature of 0.9 °C. Thus, once again, the greener scenario emerges as the most promising for mitigating the urban heat island issue.
Unlike the values of UHIICL, the greener scenario exhibited the highest urban temperature amplitude (Table 2). Human health can be affected by temperature variations throughout the day, as well as by hourly values. The American College of Cardiology (2018) recognizes that rapid and extreme temperature fluctuations within a short period of time can increase stress on the body's response systems to temperature changes. The authors note that every five degrees of temperature range corresponds to a five percent increase in the likelihood of acute myocardial infarction (commonly known as a heart attack). Interestingly, the greener scenario recorded the highest temperature range. This is due to the scenario's low thermal storage capacity and overall low thermal conductivity, resulting in less resistance for the urban area to cool down at night. Therefore, solely considering this factor, residents in the greener scenario might be slightly more susceptible to acute myocardial infarction compared to residents in the current scenario. However, the hourly temperature value is another important factor in the onset of diseases and should be taken into consideration along with the temperature range. Sharovsky et al. (2004) observed an 11% increase in the risk of mortality from acute myocardial infarction in the temperature range of 23.8-27.3 °C compared to the range of 21.6-22.6 °C. Therefore, it should be considered whether the greater temperature amplitude in the greener scenario makes it more detrimental to the population given that the temperature range in the current scenario is not far off from the greener scenario and, in addition, exhibits a notable number of grid cells in the 23.8-27.3 °C range compared to the greener scenario. In the current scenario, the number of grid cells that exceeded 23.8 °C throughout the day represented 100,436 km2. In the greener scenario, only 80,969 km2. In the intercalated scenario, it was 89,009 km2 and in the boundary scenario it was 88,185 km2. Figure 11 shows the area with temperatures higher than 23.8 °C for each scenario and the main difference occurs at night.
In Fig. 12, some differences in the heights of the planetary boundary layer are observed during daytime. The current scenario encompasses various areas with intense red tones, predominantly in the eastern zone of MRSP. This region of the study area has the highest concentrations of compact land use that tend to present higher sensible heat flux. Boundary line is clearly visible in the image's lower half (Fig. 12d). As there is less sensible heat in the greener scenario, the planetary boundary layer recedes, resulting in the lighter color intensities among the four scenarios. Greener and intercalated scenarios have very similar color and wind profiles among all MRSP. This may be linked with the great green infrastructure dispersal rate both scenarios have.
Daytime average planetary boundary layer height for (a) current, (b) greener, (c) intercalated, and (d) boundary. Arrows represent average wind at 10 m of altitude for the current scenario in (a); (b), (c), and (d) represent the difference in average daytime wind between current and greener, intercalated, and boundary, respectively. Colorbar is planetary boundary layer height in meters. Note that the wind scales are different between panels to enhance the differences.
The height of the PBL is highly dependent on the turbulent fluxes of momentum and heat. At night, due to anthropogenic surfaces gradually releasing the stored heat after sunset and generating greater thermal turbulence compared to vegetated surfaces, it is expected that the patterns depicted in Fig. 13 will exhibit significant similarity with the land use maps shown in Fig. 4, with higher PBL height values over more urbanized areas. Greener scenario wind profile is similar to intercalated scenario, as previously presented at daytime PBL figure. Boundary scenario wind profile is similar to its daytime wind profile.
Nighttime average planetary boundary layer height for (a) current, (b) greener, (c) intercalated, and (d) boundary. Arrows represent average wind at 10 m of altitude for the current scenario in (a); (b), (c), and (d) represent the difference in average nighttime wind between current and greener, intercalated, and boundary, respectively. Colorbar is planetary boundary layer height in meters. Note that the wind scales are different between panels to enhance the differences.
4. Conclusion
Based on the data obtained from the simulations and their comparisons with the current scenario, it can be stated that the absolute area of green infrastructure has a significant influence on the intensity of the canopy urban heat island and the average temperature values in the Metropolitan Region of São Paulo, while also altering the profiles of winds and the planetary boundary layer. The greener scenario has proven more effective in mitigating the UHIICL in the MRSP, indicating that the density of buildings and the presence of trees, shrubs, and permeable soil are directly related to mesoscale meteorological characteristics. Therefore, increasing greenery can contribute to improved thermal comfort, particularly in situations of intense heat.
The data in this study showed that the dispersion of green infrastructure may not substantially impact the average UHIICL and urban temperature values at the city, or metropolitan region, scale. However, it significantly affects the spatial distribution of temperature, the local dynamics of the urban heat island, and the thermal comfort of residents in specific areas, as the map figures in this paper demonstrate. Therefore, this study highlights the crucial importance of considering local reforestation public policies and the maintenance of urban parks and squares to achieve the sustainable urban development goals locally. Because trees have a significant local cooling impact, urban reforestation programs should prioritize low-income neighborhoods to advance environmental justice and ensure equitable access to affordable cooling by reducing the need for air conditioning and fans, ultimately improving the affordability of housing and quality of life for all residents.
While this study emphasizes the need for local green interventions, its purpose was not to promote a major shift in the urban development patterns of the already consolidated neighborhoods in the Metropolitan Region of São Paulo to achieve the greener scenario, since it would lead to mass displacement and enormous financial cost for the municipalities involved. Instead, this study can provide guidance for new neighborhoods being built from the ground up. The data presented here demonstrate that to maintain a pleasant environment for residents, urban planners can not consider deforestation as a land use transition option. Instead, they should seek to mix residential buildings with healthy green infrastructure with a good leaf area index.
Acknowledgements
We want to thank the São Paulo Research Foundation (FAPESP) for providing the financial support (Grant #2021/11483-9). We want to thank the National Council for Scientific and Technological Development, grant #306889/2022-6, for funding this work.
References
-
AKBARI, H.; POMERANTZ, M.; TAHA, H. Cool surfaces and shade trees to reduce energy use and improve air quality in urban areas. Solar Energy, v. 70, n. 3, p. 295-310, 2001. doi
» https://doi.org/10.1016/S0038-092X(00)00089-X -
AMERICAN COLLEGE OF CARDIOLOGY (ACC). Heart attacks often follow dramatic changes in outdoor temperature Available from https://www.acc.org/about-acc/press-releases/2018/02/27/11/19/heart-attacks-often-follow-dramatic-changes-in-outdoor-temperature, accessed in May. 2022.
» https://www.acc.org/about-acc/press-releases/2018/02/27/11/19/heart-attacks-often-follow-dramatic-changes-in-outdoor-temperature -
ARNFIELD, A. Two decades of urban climate research: a review of turbulence, exchanges of energy and water, and the urban heat island. International Journal of Climatology, v. 23, n. 1, p. 1-26, 2003. doi
» https://doi.org/10.1002/joc.859 -
ASAEDA, T.; CA, V.; WAKE, A. Heat storage of pavement and its effect on the lower atmosphere. Atmospheric Environment, v. 30, n. 3, p. 413-427, 1996. doi
» https://doi.org/10.1016/1352-2310(94)00140-5 -
BARROS, H.; LOMBARDO, M. A ilha de calor urbana e o uso e cobertura do solo em São Paulo-SP. GEOUSP Espaço e Tempo (Online), v. 20, n. 1, p. 160-177, 2016. doi
» https://doi.org/10.11606/issn.2179-0892.geousp.2016.97783 -
CHEN, F.; KUSAKA, H.; BORNSTEIN, R.; CHINGM J.; GRIMMOND, C.S.B.; et al. The integrated WRF/urban modelling system: development, evaluation, and applications to urban environmental problems. International Journal of Climatology, v. 31, n. 2, p. 273-288, 2010. doi
» https://doi.org/10.1002/joc.2158 -
CHING, J.; MILLS, G.; BECHTEL, B.; SEE, L.; FEDDEMA, J.; et al. WUDAPT: An urban weather, climate, and environmental modeling infrastructure for the anthropocene. American Meteorological Society, v. 99, n. 9, p. 1907-1924, 2018. doi
» https://doi.org/10.1175/BAMS-D-16-0236.1 -
COMPANHIA AMBIENTAL DO ESTADO DE SãO PAULO (CETESB). Sistema de Informações da Qualidade do Ar Available from https://qualar.cetesb.sp.gov.br/qualar/home.do, accessed in Sep. 2022.
» https://qualar.cetesb.sp.gov.br/qualar/home.do -
FERREIRA, M.J; OLIVEIRA, A.P.; SOARES, J.; CODATO, G.; BáRBARO, E.W.; et al. Radiation balance at the surface in the city of São Paulo, Brazil: Diurnal and seasonal variations. Theoretical and Applied Climatology, v. 107, n. 1, p. 229-246, 2011. doi
» https://doi.org/10.1007/s00704-011-0480-2 -
FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS (FAO). Frost Protection: fundamentals, practice and economics Available from https://www.fao.org/4/y7231e/y7231e00.pdf, accessed in Oct. 2022.
» https://www.fao.org/4/y7231e/y7231e00.pdf -
GOVERNO DO ESTADO DE SãO PAULO. Região Metropolitana de São Paulo (RMSP) Available from https://rmsp.pdui.sp.gov.br/?page_id=127, accessed in Apr. 2022.
» https://rmsp.pdui.sp.gov.br/?page_id=127 -
GRIMMOND, C.S.B.; OKE, T.R. Aerodynamic properties of urban areas derived from analysis of surface form. Journal of Applied Meteorology, v. 38, n. 9, p. 1262-1292, 1999. doi
» https://doi.org/10.1175/1520-0450(1999)038<1262:APOUAD>2.0.CO;2 -
GUPTA, S.K.; RAM, J.; SINGH, H. Comparative study of transpiration in cooling effect of tree species in the atmosphere. Journal of Geoscience and Environment Protection, v. 6, n. 8, p. 151-166, 2018. doi
» https://doi.org/10.4236/gep.2018.68011 -
HARLAN, S.L.; BRAZEL, A.J.; PRASHAD, L.; STEFANOV, W.L.; LARSEN, L. Neighborhood microclimates and vulnerability to heat stress. Social Science & Medicine, v. 63, n. 11, p. 117-124, 2006. doi
» https://doi.org/10.1016/j.socscimed.2006.07.030 -
HARLAN, S.L.; DECLET-BARRETO, J.H.; STEFANOV, W.L.; PETITTI, D.B. Neighborhood effects on heat deaths: social and environmental predictors of vulnerability in Maricopa County, Arizona. Environmental Health Perspectives, v. 121, n. 2, p. 197-204, 2012. doi
» https://doi.org/10.1289/ehp.1104625 -
HE, C.; VALAYAMKUNNATH, P.; BARLAGE, M.; CHEN, F.; GOCHIS, D.; et al. The community Noah-MP Land Surface Modeling System Technical Description Version 5.0 Technical Note, 285 p., 2023. doi
» https://doi.org/10.5065/ew8g-yr95 - HOWARD, L. The Climate of London by Luke Howard (1833). London: IAUC, 286 p., 2006.
-
INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATíSTICA (IBGE). Panorama do Censo 2022 Available from https://censo2022.ibge.gov.br/panorama/, accessed in Mar. 2022.
» https://censo2022.ibge.gov.br/panorama/ - LOMBARDO, M.A. Ilha de calor nas metropoles o exemplo de sao paulo (1985). São Paulo: Hucitec, 244 p., 1985.
-
MANLEY, G. The mean temperature of central England, 1698-1952. Quarterly Journal of the Royal Meteorological Society, London, v. 79, n. 340, p. 242-261, Apr. 1953. doi
» https://doi.org/10.1002/qj.49707934006 -
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION (NASA). MODIS Land Cover Type/Dynamics Available from https://modis.gsfc.nasa.gov/data/dataprod/mod12.php, accessed in Jun. 2022.
» https://modis.gsfc.nasa.gov/data/dataprod/mod12.php -
OKE, T.R. City size and the urban heat island. Atmospheric Environment, v. 7, n. 8, p. 769-779, 1973. doi
» https://doi.org/10.1016/0004-6981(73)90140-6 -
ROTH, M.; OKE, T.R.; EMERY, W.J. Satellite-derived urban heat island from three coastal cities and the utilization of such data in urban climatology. International Journal of Remote Sensing, v. 10, n. 11, p 1699-1720, 1989. doi
» https://doi.org/10.1080/01431168908904002 -
SAILOR, D.J. A review of methods for estimating anthropogenic heat and moisture emissions in the urban environment. International Journal of Climatology, v. 31, n. 2, p. 189-199, 2011. doi
» https://doi.org/10.1002/joc.2106 -
SCHULZE, E.D.; LANGE, O.L.; KAPPEN, L.; BUSCHBOM, U.; EVENARI, M. Stomatal responses to changes in temperature at increasing water stress. Planta, v. 110, p. 29-42, 1973. doi
» https://doi.org/10.1007/BF00386920 -
SHAROVSKY, R.; CESAR, L.A.M.; RAMIRES, J.A.F. Temperature, air pollution, and mortality from myocardial infarction in São Paulo, Brazil. Brazilian Journal of Medical and Biological Research, v. 37, n. 11, p. 1651-1657, 2004. doi
» https://doi.org/10.1590/S0100-879X2004001100009 -
SKAMAROCK, W.C.; KLEMP, J.B.; DUDHIA, J.; GILL, D.O.; LIU, Z.; et al. A description of the advanced research WRF Model version 4.3 Technical Note, 162 p., 2019. doi
» https://doi.org/10.5065/1dfh-6p97 -
STEWART, I.D.; OKE, T.R. Local climate zones for urban temperature studies. Bulletin of the American Meteorological Society, v. 93, n. 12, p. 1879-1900, 2012. doi
» https://doi.org/10.1175/BAMS-D-11-00019.1 - UMEZAKI, A.S. Estudo numérico das características da ilha de calor urbana da Região Metropolitana de São Paulo: o caso de 2014. Master's Dissertation, USP, São Paulo, 2022.
-
WINBOURNE, J.B.; JONES, T.S.; GARVEY, S.M.; HARRISON, J.L.; WANG, L.; et al. Tree transpiration and urban temperatures: current understanding, implications, and future research directions. BioScience, v. 70, n. 7, p. 576-588, 2020. doi
» https://doi.org/10.1093/biosci/biaa055 -
WUDAPT. LCZ Generator Maps Available from https://www.wudapt.org/lcz-maps/, accessed in Sep. 2022.
» https://www.wudapt.org/lcz-maps/
