Exploring the Meteorological Impacts of Surface and Rooftop Heat Mitigation Strategies Over a Tropical City

Different heat mitigation technologies have been developed to improve the thermal environment in cities. However, the regional impacts of such technologies, especially in the context of a tropical city, remain unclear. The deployment of heat mitigation technologies at city‐scale can change the radiation balance, advective flow, and energy balance between urban areas and the overlying atmosphere. We used the mesoscale Weather Research and Forecasting model coupled with a physically based single‐layer urban canopy model to assess the impacts of five different heat mitigation technologies on surface energy balance, standard surface meteorological fields, and planetary boundary layer (PBL) dynamics for premonsoon typical hot summer days over a tropical coastal city in the month of April in 2018, 2019, and 2020. Results indicate that the regional impacts of cool materials (CMs), super‐cool broadband radiative coolers, green roofs (GRs), vegetation fraction change, and a combination of CMs and GRs (i.e., “Cool city (CC)”) on the lower atmosphere are different at diurnal scale. Results showed that super‐cool materials have the maximum potential of ambient temperature reduction of 1.6°C during peak hour (14:00 LT) compared to other technologies in the study. During the daytime hours, the PBL height was considerably lower than the reference scenario with no implementation of strategies by 700 m for super‐cool materials and 500 m for both CMs and CC cases; however, the green roofing system underwent nominal changes over the urban area. During the nighttime hours, the PBL height increased by CMs and the CC strategies compared to the reference scenario, but minimal changes were evident for super‐cool materials. The changes of temperature on the vertical profile of the heat mitigation implemented city reveal a stable PBL over the urban domain and a reduction of the vertical mixing associated with a pollution dome. This would lead to crossover phenomena above the PBL due to the decrease in vertical wind speed. Therefore, assessing the coupled regional impact of urban heat mitigation over the lower atmosphere at city‐scale is urgent for sustainable urban planning.


Introduction
The planetary boundary layer (PBL) is the lowest portion of the atmosphere and responds to surface forces, including frictional drag, evapotranspiration, heat and momentum transfer, and terrain-induced flow modifications, with a time scale of about an hour or less (Masson et al., 2020;Stull, 1988). The depth of the PBL varies spatially and temporally, ranging from hundreds of meters to a few kilometers. PBL depth can be measured as the height of potential temperature and mixing ratio inversions within the mixed layer or as the height where surface-driven turbulence production ceases (Brandi et al., 2021;Stull, 1988). PBL characteristics over urban areas are modified due to turbulent processes generated by the morphological structure of the urban fabric and anthropogenic heat (AH) release (Baklanov et al., 2011;Ching et al., 2019;Chow et al., 2012;Masson et al., 2020;Miao & Chen, 2008;Nadimpalli et al., 2022;Niyogi et al., 1999;Pelliccioni et al., 2016;Routray et al., 2010). During extreme heat events, the PBL structure is further altered via regulation of energy, and momentum, transport between the surface and overlying atmosphere, resulting in severe impacts on the health and quality of life of urban citizens (Hamdi et al., 2016(Hamdi et al., , 2020Patel et al., 2022;Tewari et al., 2023).
A range of heat mitigation technologies has been proposed to combat environmental consequences associated with extreme urban heat. These strategies fall under the broad umbrella of green infrastructure (e.g., urban lawns, trees, and green roofs (GRs)) and biophysical, or engineered, materials (e.g., cool materials (CMs)), including pavements and roofs, and super-cool broadband radiative coolers (Akbari et al., 2001;Baniassadi et al., 2019;Carlosena et al., 2020;Chow et al., 2012;Georgescu et al., 2013Georgescu et al., , 2014Krayenhoff et al., 2018;Santamouris, 2014aSantamouris, , 2014bSantamouris & Yun, 2020;Takebayashi & Moriyama, 2007). Broadband radiative coolers are emerging as a promising technology , in large part because most heat mitigation technologies reduce daytime urban temperatures but have minimal impact during the evening and nighttime hours (Georgescu, 2015). Achieving daytime sub-ambient cooling has been a breakthrough in the field since buildings require cooling during the peak day period (Santamouris & Feng, 2018). To cool under the sun, materials need high reflectivity in the solar and high emissivity in the atmospheric window (8-13 μm). As a result, these materials have achieved more significant temperature reduction than CMs. A recent, low-cost scalable and sprayable polymeric materials reduced the baer substrates' temperature, under non-ideal climatic conditions with relative humidity up to 63%, up to 12°C temperature drops compared to the substrate . By maximizing the rejection of solar radiation and energy dissipation during all hours of the diurnal cycle, radiative cooling devices reach sub-ambient cooling under the direct sun without consuming additional energy while also reducing temperatures during nighttime hours (Z. Chen et al., 2016;Santamouris & Feng, 2018). A meta material film achieved a cooling power of 93 Wm −2 under direct sunshine at noon (Zhai et al., 2017). These types of materials have been studied recently on a city scale weather simulation models both during the summer and winter period. Researchers simulated the effect of three different super cool materials (SCMs) scenarios during the summer time in Kolkata metropolitan area (KMA) and found that the maximum decreases of net radiation varied between 468.3 and 236.3 Wm −2 and the maximum decreases of latent heat were between 41.3 and 13.1 Wm −2 . The SCMs are excellent urban heat mitigators, nevertheless they can lead to an overcooling penalty during the cold seasons. A study carried out in Kolkata concluded that under the most reflective and emissive scenario, the maximum decrease of net radiation was 354.9 Wm −2 .
Another study proposed different mitigation technologies on a street canyon and showed that radiative cooling materials were among the most efficient strategies in the study, decreasing up to 1.6°C ambient temperature and 24.2°C surface skin temperature (Bartesaghi-Koc et al., 2021). As mentioned before, CMs and SCMs are very reflective and can lead to an overcooling penalty at times. Recent research addressed this knowledge gap, calculating different solutions with optically modulated SCMs (Khan, Carlosena, et al., 2022). Optically modulated SCMs can be an efficient urban heat mitigation technology for use during the summer period without overcooling potential. Khan, Carlosena, et al. (2022) concluded that the modulation of both solar reflectivity and infrared emissivity could increase the ambient temperature up to 0.4°C and 1.4°C during the day and night, respectively. Decreasing the burden of heat for urban environments has considerable benefits across society, including reduction of energy consumption, improvement in thermal comfort and health (Akbari et al., 2004;Ali-Toudert et al., 2005; H. Chen et al., 2009;Santamouris, 2020;Santamouris et al., 2015). technologies in the urban environment for extreme urban heat management, such as the newly developed supercool materials. However, careful attention should be paid to unintended consequences.
Implementing urban heat mitigation technologies at the city-scale will modify surface heat and moisture transfer processes which will, in turn, modulate PBL dynamics. In general, higher sensible heat fluxes lead to more substantial vertical mixing over urban areas than rural surroundings. However, sensible heat flux is reduced due to the deployment of traditional heat mitigation strategies such as green and cool roofs, generating a less vigorous PBL with reduced vertical wind speeds and a lower PBL height (Brandi et al., 2021). For example, using the Weather Research and Forecasting (WRF) model, Sharma et al. (2016) reported a decrease in the PBL height of 0.5 km upon deployment of green/cool roofs across the Chicago Metropolitan Area. These results are consistent with Georgescu (2015), who demonstrated a reduction in turbulent kinetic energy and PBL height more significant for CMs than GRs under a scenario characterized by future urban expansion in California. Recent numerical modeling work has provided additional support indicating a decrease in primarily daytime PBL height with reduced impacts during non-sunlit portions of the diurnal cycle (Brandi et al., 2021;Song et al., 2018). However, much of the previous work has been conducted over cities located in developed nations (Yang et al., 2019). There is a need of similar research over conurbations in developing nations where meteorology, PBL dynamics, and human activities can lead to severe degradation of air quality. For example, Raman et al. (1990) reported that the PBL height of New Delhi (India) could reach 4.7 km during the pre-monsoon months. An important question is whether and to what extent would the deployment of heat mitigation strategies lower the height of the PBL, with important consequences for air quality (Charusombat et al., 2010) and health.
Additional modifications to urban environments (e.g., absorbed solar radiation via enhancement of reflective materials, relative humidity/water vapor via incorporation of green infrastructure, or wind speed and circulation) are likely to have a feedback that could compensate for the decrease in the PBL depth. For example, reduced ozone production can be expected if less solar radiation is absorbed because of cool roof deployment. Radiative coolers could increase the dryness of cities due to the offset of incoming solar radiation (i.e., less radiant energy available for partitioning to latent heat flux) and enhance the emission of longwave radiation. Incorporating green infrastructure may have unintended consequences on urban pollution, creating additional health problems (Santamouris & Osmond, 2020). Evidence is mounting that the built environment can modify convective precipitation (Hamdi et al., 2020;C. L. Zhang et al., 2009). Similar alterations in the surface energy balance, due to incorporating heat mitigation strategies, may further modulate precipitation magnitude or patterns over or downwind of urban areas (J. Liu & Niyogi, 2019).
We build upon previous work by focusing on the regional scale meteorological impacts of heat mitigation technologies on a tropical urban environment (Khan, Akbari, et al., 2022). A majority of urban heat mitigation studies for the tropics have focused on the application of different strategies to improve the thermal balance of large and medium-sized cities such as Singapore (Li & Norford, 2016;Mughal et al., 2019;Tan et al., 2015), Hong Kong (L. Chen & Ng, 2013;Cheng et al., 2010;Jim, 2015;Ng et al., 2012;Wong et al., 2013), Colombo (Emmanuel et al., 2007), Bhubaneswar (Swain et al., 2023), Mumbai (Dwivedi & Mohan, 2018), and Kolkata (Chatterjee et al., 2019). Despite the growing importance of urban climate mitigation research (Gonzalez et al., 2021;Khan, Akbari, et al., 2022;Krayenhoff et al., 2021), there exists a dearth of investigations focused on the impact of heat mitigation technologies on PBL dynamics in general and over tropical urban environments in particular. To our knowledge, there is no published assessment of the regional impact, in tropical urban areas, of the different presented heat mitigation strategies on lower atmospheric dynamics. The nexus between air pollution, green infrastructure, reflective materials, and human health remains unclear and requires examination prior to large-scale deployment.
This study aims to improve our understanding of the impacts of heat mitigation technologies on lower boundary layer dynamics at the city-scale using the WRF model. A historical and organically developed tropical coastal city characterized by a high-density residential area has been selected as our testbed: the (KMA). During the past decades, the KMA has undergone significant urban growth and has experienced considerable heat stress with continuously increasing urban heating intensity (Chatterjee et al., 2019;Khan, Akbari, et al., 2022). We build on prior work to investigate the potential impacts associated with the explicit representation of five different heat mitigation technologies. These include (a) cool materials -CMs, (b) super-cool broadband radiative coolers, (c) green roofs-GRs, (d) vegetation fraction change, and (e) a combination of these technologies. Our work addresses important environmental concerns associated with the city-scale deployment of urban heat mitigation 1. How effective are heat mitigation measures in the tropical urban environments? 2. How does a city-scale heat mitigation effort alter the city-scale thermal environment? 3. What is the spatio-temporal variability of the PBL dynamics caused by the heat mitigation measures?

Methods and Materials
To investigate the impact of different urban heat mitigation technologies on surface meteorological fields and PBL dynamics at city-scale, we employed the community Weather Research and Forecasting (WRF v4.3;Skamarock et al., 2021) model coupled with a single-layer urban canopy model (SLUCM, Kusaka et al., 2001). WRF is a state-of-the-art, non-hydrostatic, fully comprehensive mesoscale numerical weather prediction framework used here to investigate the effectiveness of implementing cool roof technologies on urban building's roof surfaces to mitigate the urban overheating and its influence on regional climate.

WRF Model Configuration
To address urban meteorological and environmental issues, SLUCM is dynamically coupled with the WRF Noah-Land Surface Model scheme (F. Chen et al., 2011). Our modeling domain utilizes three nested domains centered over KMA (22.57°N,88.36°E) with a grid spacing of 18, 6, and 2 km ( Figure 1). The details of the model physics parameterizations used in this study are listed in Table 1. The Kain-Fritsch (Kain, 2004;Zheng et al., 2016) convection scheme was used for the outer two domains but was turned off for the innermost domain. We used the default AH input in SLUCM with a fixed diurnal pattern, assuming a peak value of 50 and 90 Wm −2 for high-intensity residential and commercial areas.
We used the National Centers for Environmental Prediction (NCEP) Global Forecasting System operational global analysis and forecast data set (NCEP-GFS 0.25 Degree Global Forecast Grids, 2015) as initial and lateral boundary conditions during model integration. NCEP-GFS data are provided on a 0.25 × 0.25-° grid and are available at a three-hour frequency (https://rda.ucar.edu/datasets/ds084.1/). We focus our simulations on hot, pre-monsoon days characterized by considerable urban heating and associated health risk. Extreme heatwaves aggravate conditions in urban areas by boosting temperatures within the built environment and increasing urban temperatures more intensely than rural temperatures (Khan & Chatterjee, 2016;Khan, Akbari, et al., 2022). Under such adverse weather conditions, mitigation strategies are critically needed to reduce peak ambient temperatures and minimize any heat-related health risks. Therefore, we objectively selected sweltering days with peak temperatures exceeding 39.2°C (Express News Service, 2021) for two consecutive very hot summer days during the pre-monsoon month of April to assess how heat mitigation strategies could impact the lower atmosphere of a tropical coastal city. We conducted 2-day simulations where the initial 24-hr is considered as model spin-up. This simulation is the control experiment compared to available observations to ensure model skill for appropriately reproducing the region's meteorology. In addition, we conducted sensitivity experiments (using an identical model configuration) to examine the impacts resulting from the implementation of a variety of heat mitigation strategies, including radiative coolers, cool roofs, GRs, increased vegetation fraction, and a final scenario that integrates the strategies above as a Cool city (CC) simulation ( Table 2). The simulations were from 6 April through 8 April 2020, coinciding with India's critical COVID-19 confinement period . In addition, we have added two consecutive very hot summer days during the pre-monsoon month of April as additional case studies, each of 48-hr duration in the year 2019 and 2018. Further analysis were conducted with group of simulation that have different meteorological condition for each April month in the year 2019 and 2018 (Table 1). All WRF simulation output was analyzed for the concluding 24-h results. Our focus on simulation periods of a typical hot summer day, instead of the execution of the model in climate mode, permits the execution and analysis of a range of sensitivity tests to quantify the response of different heat mitigation technologies.

Numerical Design for Heat Mitigation Technologies
One base case (control) and five additional numerical simulations were carried out for the selected period for the present study. Initially, the control simulation with the original urban landscape (as represented by the moderate resolution imaging spectroradiometer data set) and urban parameter table were executed to assess the reliability of WRF by comparing simulation results with local observations. To better understand the regional influence of the different heat mitigation strategies on the urban environment, the Control case is considered the urban reference condition and is compared with simulations using different heat mitigation technologies. We simulated five types of heat mitigation technologies: 1. CMs, which increase the albedo value of all urban surfaces from 0.2 (control case) to a maximum of 0.65 (Morini et al., 2017); Short-wave radiation Dudhia (1989) Longwave radiation Rapid radiative transfer model scheme Mlawer et al. (1997) Planetary boundary layer Asymmetrical Convective Model version 2 (ACM2) Pleim (2007) Cumulus parameterization Kain-Fritsch (KF)  . 3. Vegetation fraction (VEG) increase, which reduced the urban fraction by 50%. We note that this is a realistic scenario that maximizes vacant spaces already available in the city (Chatterjee et al., 2019); 4. CC experiment, a combination of CMs and GRs (on building facets and surfaces, a higher increase in reflectivity and emissivity has been taken into consideration for these CMs) were simulated over the urban domain; 5. Super-cool materials (SCMs), which are more reflective than standard CMs, wherein the albedo and emissivity of all urban surfaces increased from 0.20 to 0.96 and 0.90 to 0.97, respectively .
Recent studies concern these newly developed SCMs. For example, a material based on titanium dioxide particles on top of silicon dioxide or silicon carbide nano particles with 90.7% solar reflectivity and a 90.11% sky window emissivity could theoretically achieve 5°C below ambient under direct solar radiation. However, the experiment conducted in Shanghai did not achieve sub-ambient cooling due to high humidity (Bao et al., 2017). On the other hand, a nano porous structure was tested in four different locations and climates. The material showed a distinct behavior in Phoenix, where it reached a sub-ambient temperature decrease of 3°C which is half of the cooling expected (Mandal et al., 2018). Moreover, a radiative cooler with 98% solar reflectivity showed a temperature decrease of up to 2.7°C at noontime under direct sunlight in humid Hong Kong and provided the proof of concept for the stringent solar reflectivity demand under high humidity (Zhong et al., 2020). Researchers have also calculated the cooling performance curve of an ideal emitter under different total water vapor columns. They found that the cooling power decreases from 115.9 to 32.6 Wm −2 with the increase of total water vapor from 0 to 6,000 atm-cm. However, subsequent increases in water vapor had a negligible effect, increasing from 6,000 to 8,000 atm-cm merely decreases by 3.3 Wm −2 (C. . SCMs are extremely reflective when compared to third generation cool roofs or white paints, and are particularly effective at emitting radiation with wavelengths ranging from 8 to 13 μm (Santamouris & Yun, 2020). The field measurements have been carried out based on data from small samples tested over very short periods of time. These findings demonstrated that SCM cooling ability is influenced by the geography and climate of the city or location (Baniassadi et al., 2019;Feng et al., 2021). The cooling effect is more noticeable in dry atmospheres with a clear sky. However, when the atmosphere is cloudy or humid, available water vapor in the lower atmosphere may reduce the significant amount of infrared radiation emission (Lim, 2019). Our prior work for Kolkata  using SCM showed a good cooling potential despite the fact that Kolkata is a tropical wet and dry climatic city. As highlighted above, humidity plays a significant role in the ability of the materials to radiate heat. As the amount of water vapor in the atmosphere increases, the atmospheric transmittance decreases, and the atmospheric irradiance rises, causing the radiative cooler to absorb more atmospheric radiation and as a result, limit cooling performance. Although some experiments have not achieved sub-ambient temperatures under humid weather, others have shown considerable temperature reduction. Moreover, a few experiments have found that SCMs may not last in a variety of weather conditions or fit easily into all urban building environments (Gentle & Smith, 2015). Further details on the properties of each simulation are shown in Table 2.

Observational Data
We used six regional meteorological stations available from Weather Kolkata ( Figure 1. It is important to note that even though the present simulations represent a model configuration that accounts for 2 km grid spacing within the innermost grid, individual station readings are obtained at a particular spatial location. Therefore, we considered each station separately and compared the values of model-derived 2 m temperature and dew point temperature at the grid point nearest to the observation stations. In addition, the availability of observational data with a temporal frequency of 1-hr and equivalent model output generation allows for the assessment of the diurnal cycles. Besides this, we have also validated the PBL height with upper air Radiosonde data for variable energy cyclotron centre Calcutta Observations (Kolkata) (http://weather.uwyo.edu/) for city of Kolkata. The PBL height is not a direct observation and is estimated from sounding (only 1-2 times in a day) and ground-based data using Richardson approach (Lee & De Wekker, 2016;Schmid & Niyogi, 2012).

Model Evaluation and Validation
To evaluate the performance of the WRF-SLUCM system, we compared hourly simulated 2 m air temperature, dew point temperature, and upper air sounding against local measurements for the control case simulation over urban grid cells in the innermost domain. A statistical comparison of the mean bias error (MBE), mean absolute error, root mean square error (RMSE), correlation coefficient (r), and the index of agreement (IOA) for hourly 2 m air temperature for the 24-hr duration are listed in Table 3. The model evaluation is based on the correlation between the WRF simulations and observations for 2 m-temperature across the diurnal cycle ( Figure 2). The coupled WRF-SLUCM model accurately captures the temperature observed at different stations (mean R = 0.98; mean bias = 0.72°C) for ALP, RBG, SHP, PLB, JRB, and JKA. The base case simulation produced urban meteorological conditions and statistically agreed with local observations (p < 0.05). The simulated average urban heat island (UHI) intensity varied from 1.8°C to 2.1°C for daytime and nighttime, respectively, in high-density residential areas relative to surrounding rural landscapes. The range of MBE and RMSE of air temperature was 0.4°C-1°C and 1.2°C-1.9°C, respectively. The range of IOA was 0.98 to 0.95, with average values of 0.96 when considering all observation stations. The model slightly underestimated the nighttime 2 m air temperature, potentially resulting from underestimating anthropogenic heating over the urban domain . We also assessed the impacts on local meteorological stations as they are most likely influenced when using the urban canopy scheme (UCM) scheme. The well-simulated daytime warming is balanced by the nighttime cooling, resulting in a diurnal range similar to observations. The threshold "comfort level" of dew points is greater than 20°C for the city stations. The difference is identical when quantifying impacts on local meteorological stations. Although WRF does not display considerable warm (comfort) bias over urban locales, the representation of the 24-hr averaged diurnal range of dew point temperature is well captured (Figure 3 and Table 4). In addition, model biases are most likely caused by: (a) lack of proper urban morphological representation and (b) uncertainties in model physical schemes, input data used, and locally meaningful urban biophysical parameters (Fung et al., 2022). Nevertheless, our initial evaluation highlights that the model can realistically replicate the urban environment, including a well-simulated evolution of the diurnal cycle of near-surface temperature and dew point. Moreover, the model framework can be used to examine regional meteorology and investigate the regional influence of different heat mitigation technologies.
The accurate representation of the PBL height is one of the important aspects for any sensitivity analysis, especially in simulation with turbulence, wind, and lower atmospheric impacts analysis. This is particularly important given our focus on lower atmospheric impacts. Our study uses a new combined local and nonlocal closure PBL scheme, the Asymmetrical Convective Model version 2 (ACM2) (Alapaty et al., 1997(Alapaty et al., , 2001Pleim, 2007). Overall, our results indicate a small overestimate in PBL height relative to observations (see Figure S1 in Supporting Information S1). The correlation coefficient of >0.85 for each year also indicates good agreement between simulated and observed PBL height. In particular, the model simulates the diurnal evolution of the PBL height well: the sharp ramp-up that occurs shortly after sunrise is reproduced accurately, as is the late afternoon/early evening collapse. Because PBL height is not a real observation, the performance of the WRF simulations is reasonable in terms of its ability to simulate high PBL height in tropical or sub-tropical atmosphere. A case study with ACM2 PBL scheme shows that the WRF model has performed reasonably well over the subtropical region of Delhi , and, (f) Joka (JKA). The red and black circles are for the specific data in the high-density residential areas. (Mohan & Bhati, 2011). The ACM2 PBL and Pleim surface layer scheme have been found to be well suited for short and extended simulations (i.e., few hour to more than a week) at the city-scale (Gilliam & Pleim, 2010).

Meteorological Impacts of Surface and Rooftop Heat Mitigation Strategies
In this section, we quantify the effect on near-surface meteorological impacts of the implementation of several heat mitigation technologies. The conventional roof is taken as the reference simulation for each urban configuration tested and summarized here for the year 2020 experiment.

Regional Impacts on the Urban Energy Budget
The difference in the spatial variability of net radiation between the control case and sensitivity experiments incorporating different mitigation technologies (CMs, GRs, VEG, CC, and SCMs) is shown in Figure 4. The results, presented for daytime and nighttime conditions (14:00 LT and 02:00 LT), revealed that the maximum net radiative reduction was for SCMs and CMs due to their enhanced reflection of solar radiation. However, due to evaporative cooling over the urban domain, net radiation increased for GRs and VEG.    (Table 6). We note that our analysis of mitigation technologies' effect on surface energy fluxes was performed without first evaluating whether the model could successfully replicate the surface energy balance. Models that simulate the 2-m air temperature well, do not necessarily correctly simulate the surface energy balance (Shaffer et al., 2015). In the absence of available surface energy balance data, as they do not exist for the locations and period of investigation, we compared our control simula tion results with values from the literature. Therefore, we used the reported heat flux results from Singapore, a tropical city that has been the focus of considerable research, as a reference. Results revealed a maximum decrease of sensible heat flux at 14:00 LT of 180 Wm −2 when highly reflective materials were applied on building rooftops and a maximum decrease of sensible heat flux of about 85 Wm −2 when the green vegetation fraction in the urban landscape was increased.
The surface energy balance for urban areas resulting from the deployment of various heat mitigation technologies in different tropical cities is also presented. For example, Li and Norford (2016) reported that the maximum decrease of latent heat flux at 14:00 LT was about 20 Wm −2 for CM technologies, but a latent heat flux increase of 195 Wm −2 was simulated by increasing green vegetation cover for Singapore. These reported values are similar to the present results for the tropical context.
The maximum decrease of heat storage occurred at 14:00 LT, and the values were 84.8, 90.8, 90.6, 83.2, and 58.6 Wm −2 for CMs, GRs, VEG, CC, and SCMs, respectively. The maximum reduction of net inflow radiation during peak hour (14:00 LT) was 256.2, 162.2, and 236.3 Wm −2 for CMs, CC, and SCMs respectively, but an increase of Q net during the peak hour (14:00 LT) was found in GRs and VEG of 35.6 and 41.8 Wm −2 , respectively. These result from the high temperature at mid-afternoon, low surface albedo (conventional reflective surfaces), and warmer lower atmosphere than the ground surface. The average decrease of Q net over the urban surface during daytime was 143.2, 88.1, and 132 Wm −2 for CMs, CC, and SCMs, respectively, but an increase in Q net of 21.9 and 26.9 Wm −2 for GRs and VEG was found. On the other hand, during nighttime, the average increase of Q net was 6.9, 5.1, 6.3, 8.1, and 5.6 Wm −2 for CMs, GRs, VEG, CC, and SCMs, respectively. The changes in the surface energy budget directly impact near-surface variables and the overlying PBL, which are the focus of the subsequent sections.

Regional Impact on Surface Meteorological Fields
In this study, we used 2 m ambient air temperature (ΔT ambient ), surface skin temperature (ΔT surface ), 2 m dew point temperature (ΔT dew ), relative humidity at 2 m (ΔH relative ), and 10 m-wind speed (ΔW speed ) to examine the regional impact of different heat mitigation technologies. The difference of 2 m ambient air temperature (ΔT ambient ) between the control and experimental cases showed that the SCM simulation exerted the largest regional influence for most experiments. The GRs experiment led to a decrease in ΔT ambient of 0.88°C, as a result of rooftop gardening transpiration. Results showed that the maximum reduction of ΔT ambient , over urban areas, at 14:00 LT was1.6°C for SCMs, 1.5°C for both CMs and VEG, and 1.4°C for the CC experiment (Table 7). The diurnal profile of ΔT ambient reduction is also presented as average over daytime and nighttime for each experiment ( Figure 5). During the daytime, the average ΔT ambient decreased by 1.1°C for SCMs, 0.7°C for GRs, 0.5°C for VEG, and 1.1°C for CMs and CC. The average ΔT ambient for CMs and CC were spatially distributed over the urban surface.
On the other hand, during nighttime, the thermal reduction was less compared to daytime: 0.6°C, 0.59°C, 0.67°C and 0.4°C for SCMs, CMs, CC, and GRs, respectively. The nighttime reduction for SCMs occurred because of the high emissivity value of 0.97. When the vegetation fraction over urban areas increased, the thermal reduction of ΔT ambient during nighttime was greater (by 0.8°C) than daytime due to the regular supply of sufficient moisture (at 20:00 local time) for GRs with irrigated surface and less storage during the day. Different heat mitigation technologies resulted in a decrease of urban ΔT surface during daytime and nighttime. This decrease resulted from the offset of incoming solar radiation and the effects of green infrastructures whose residual impacts lasted into the evening and nighttime hours ( Figure 6). Results indicate that the SCM experiment resulted in the greatest maximum reduction of ΔT surface, 4.2°C, primarily due to the high reflectivity value used in walls and ground; as a result, it had the highest regional impact among the strategies discussed. The average ambient temperature decrease caused by heat convention was substantially higher in the low density than in the high-density parts of the city. The CMs and CC experiment revealed a decrease of ΔT surface by 4°C for both experiments. The average  reduction of ΔT surface during the daytime was 2.9°C for SCMs, 2.8°C for CMs, 2.7°C for CC, 1.7°C for GRs, and 1.3°C for VEG. The increase of moisture content resulting from additional evapotranspiration with the use of GRs and increasing vegetation fraction (VEG) led to an increase of ΔT dew of 2.5°C for GRs, VEG, and combined cases; the changes in near-surface ΔT dew were negligible in the CMs and SCMs experiment but during nighttime ΔT dew underwent a slight increase over urban areas (Figure 7). In the case of the SCMs experiment, near-surface ΔT dew endured maximum increase of 2.1°C due to the presence of high humidity. By impacting ΔT ambient , a modification of near-surface relative humidity was simulated for all thermal mitigation strategies (Figure 8).
The SCMs experiment showed a robust regional effect on ΔH relative (a maximum decrease of more than 3% compared to the control case). Compared to the control case, the increase of ΔH relative during daytime was slightly less than 2.5% for both CMs and GRs. Nevertheless, the VEG and CC experiment increases by 2.5%-3.5% over the urban core area because of additional evapotranspiration. Wind speed also changed by implementing different heat mitigation technologies over the urban area. The SCMs experiment showed a maximum change of wind speed (an increase of 2 ms −1 ) over the urban domain during the daytime due to a modification of surface absorbed solar radiation, which affects the local surface pressure variability that gives rise to microscale circulation. The GRs experiment did not considerably affect the wind speed, further influencing the ambient temperature reduction. Nevertheless, the CMs and CC slightly increased wind speed during the daytime due to the thermal gradient between core-urban to suburban environments. In tropical warm climates elevated roof surface skin temperatures caused by the absorption of solar radiation are undesirable. Reflective surface coatings aid achievement of high solar reflectance. This might notably cool the urban environment by reflecting infrared rays. The implementation of heat mitigation technologies certainly changes the urban radiative and non radiative properties. The uses of different heat mitigation strategies reduce the sensible heat available for transmission to the air or to building envelopes. The CCs (combined of CM and vegetation) increases the evapotranspiration in urban areas through soil and plants on surface (redirecting available energy to latent heat), while a reflective material increases the reflection of incoming solar radiation in urban areas by increasing the albedo of roof surfaces. These mechanisms can be further illustrated by considering the one-dimensional energy balance for an infinitesimally thin layer of roof materials at the roof-air interface by simple urban energy balance. The SCMs, CMs, and CCs impact on the wind speed differ due to higher reflectivity of the materials and vegetation effects. Results show that the maximum urban heating is reduced during the peak hour for all experiments. SCMs are more efficient than the CMs and CCs in reducing the urban heating with maximum differences due to "regional high" effects. The reductions of the urban heating vary linearly with the increasing materials fractions, but slightly nonlinearly with the increasing albedo for few hours. The maximum reduction in wind speed is simulated with SCMs, and CMs. While previous studies report that the advection of moist air from rural areas is a key mechanism, this study shows that this is not the case for the extreme urban event due to the very dry and warm conditions, and instead, convective rolls could play a more important role (Niyogi et al., 2006).

Regional Impacts on Atmospheric Boundary Layer Dynamics
The different mitigation technologies of urban heating can change the near-surface meteorological condition and modify the atmospheric boundary layer structure (  Our simulation results showed that all the thermal mitigation experiments would modify the atmospheric boundary layer structure over the urban area, although this impact was apparent mainly during the daytime (Figure 9). Some comparatively small changes also occurred outside the urban areas. These results might be numerical artifacts resulting from numerical instabilities of the model rather than the surface-induced forcing that was illustrated for KMA. In general, the atmospheric boundary layer structure dynamics are strongly regulated by the heating at the surface, which manifests as changes in the PBL state (Song & Wang, 2015;D. L. Zhang et al., 2011).
The simulated vertical profile of temperature, relative humidity, and wind speed over the urban environment was used to investigate the impact of the atmospheric boundary layer structure. The maximum reduction of the air temperature of the cool roof and CC experiment extended from the urban surface to approximately 0.3 km at night, and a gradual decrease of air temperature up to 2.5 km occurred at 14:00 LT (Figure 10). Heat mitigation strategies are expected to augment the static stability of PBL at a diurnal scale and reduce the magnitude of vertical winds, enhancing subsidence in the lower atmosphere (Brandi et al., 2021). The maximum decrease of air temperature occurred during the daytime, extending from the urban surface to approximately 2 km in the PBL. For the GRs experiment, the maximum reduction of air temperature extended from the urban surface to approximately 0.5 km above, at noon, of the simulated period. The SCMs simulation led to a maximum decrease of vertical air temperature over urban surfaces, with an impact extending to 1.5 km during daytime. Above the atmospheric boundary layer, the rate of reduction of air temperature gradually decreased compared to the atmospheric boundary layer below. During the daytime, an increase of relative humidity (greater than 9%) was evident within the atmospheric boundary layer, but a decrease of similar magnitude was found above the boundary layer. In the GR experiment, relative humidity changes were less significant than the other daytime scenarios. The maximum increase of relative humidity occurred during daytime hours and extended up to 1 km from the urban surface in the SCMs experiment. The vertical wind speeds decreased during daytime and slightly increased during nighttime, extending from the urban surface to approximately 3.5 km for all experiments. The maximum decrease of vertical wind speed that occurred during daytime was 0.5 ms −1 in the SCMs experiment due to a deficit in input solar radiation. We found that daytime PBL height significantly decreased over the urban area by 600 m for SCMs and 500 m for CMs and CC experiments, but the GRs and VEG experiment underwent reduced modification of the PBL over the urban area. On the other hand, the PBL height underwent an increase for both CMs and GRs during nighttime.

Regional Impact on Sea Breeze Circulations
The impact of land surface heterogeneity with large-scale implementation of heat mitigation strategies on sea breeze circulation of an urban area in a tropical environment has become an important subject of scientific interest (Bechtel et al., 2017;Chatterjee et al., 2019;Patel et al., 2020). The intensification of sea breeze circulation is dependent on the large-scale synoptic background, which plays an essential role in modulating the prevailing wind near the surface (Freitas et al., 2007). This local circulation occurs due to a thermal incongruity between land and sea (Lin et al., 2008;Wang et al., 2018). There is mounting evidence that these circulations can extend more than 100 km horizontally inland (Hamdi et al., 2012), and vertically may reach up to 3 km during mid-afternoon. In the vertical dimension, the height of the PBL in tropical coastal cities is associated closely with  the advection of the sea breeze. The circulation can be modified when a heat mitigation strategy is implemented. The heat mitigation strategies could alter the PBL height and potentially trigger localized circulation over the urban domain.
Our results indicated that the onset of the sea breeze was delayed to mid-afternoon due to the "regional high" effect within the lower PBL and offshore synoptic wind above the PBL . The denser cool air over the urban domain flowed toward the suburban area to replenish the buoyant warm air due to urban cooling by city-wide implemented heat mitigations technology or strategies. The CMs and SCMs could suppress the vertical lifting of urban thermals, modifying low-level motions in hot summer days resulting in a deceleration of the sea breeze front compared to other strategies. In some cases, the cooling effects of sea breezes appeared to be more discreetly in the urban center due to surrounding high-rise buildings and their roughness parameters ( Figure 11). The sea breeze-based cooling was not directly transferred to the surface due to surface roughness heterogeneity and internal boundary layers (Gilliam et al., 2004). The horizontal wind shear and frontal lifting owing to surface roughness and heterogeneity could setback the onset of sea breeze front in the urban core. The potency of the sea breeze advection was subjected to the dimension of the city, which moderates the urban heating effect. Thus, our results indicate that urban heat mitigation strategies have modified the thermal and dynamic profile in the urban boundary layer and sea breeze circulation (Aflaki et al., 2017;Niyogi et al., 2020;Wang et al., 2018). This synoptic flow prevailed in the opposite direction of sea breeze, and a weakened sea breeze development can lead to the accumulation of secondary pollutants in the back of the front. As shown in Patel et al. (2021), there is also likelihood that a convergence band can be formed due to green roofing, which can create localized heavy rains. The larger drop in PBL is due to heat mitigation effects in densely packed urban areas of the city. Due to super-cool broadband materials, the PBL (black thick line) tends to be much lower due to offset of direct solar energy and reduced the vertical wind speeds during daylight. This effect, in turn, led to lower vertical convective mixing. The PBL also rapidly contracts due to a reduction of thermals from the urban surface after sunset and low temperature. The thermal characteristics of the surface, influence the diurnal range of surface skin temperatures, and net inflow radiation at the surface and its variation with height, which determine the radiative warming or cooling of the surface and the PBL associated with heat mitigation options.
We also found changes in the moisture fields for our thermal case, which could potentially trigger afternoon thunderstorms.

Sensitivity Analysis for Different Meteorological Conditions
We conducted one control, in addition to a five-sensitivity experiment for consecutive 3 days (of 2018, 2019, and 2020) during pre-monsoon seasons. In 2020, we conducted a 2-day simulation where an initial 24-hr is considered as model spin-up time and remaining 1-day simulation was analyzed in this study. Additional two sets of experiments were conducted in April 2019 and April 2018 for a better understanding of the physical mechanism of said sensitivity experiment.
The average modifications or decreases in the surface standard meteorological field for 2018 and 2019 are fairly close to those for the experiments conducted in 2020. The average correlation values for each standard meteorological field range from 0.89 to 0.92 during the phase of the experiment, and they are highly significant at the 0.05 level of significance. It may be deduced that a single day simulation is sufficient to accurately replicate each surface standard field for heat mitigation measures. In the supplementary information, average changes (experiment minus control) representing the 3-case study average (2018, 2019, and 2020) in net inflow radiation, 2-m ambient temperature, surface skin temperature, dew point temperature, relative humidity, PBL height are shown in Supporting Information S1 for daytime and nighttime ( Figures S2-S7).
The difference in spatial and diurnal scale variability of the partition of urban surface energy balance and thermal balance (mainly 2-m ambient air temperature and surface skin temperature) was calculated for a three-member ensemble study by WRF-SLUCM. This ensemble study shows that the hourly average of April month in the 2019 and 2018 simulations reproduced almost similar results (±4 Wm −2 ) for urban surface energy balance as well as urban thermal balance (±0.3°C) compared to 2020. The maximum reduction of sensible heat flux during peak hour average (14:00 LT) was 171. In the case of latent heat flux and heat storage, the same result was reproduced in a 2-day average simulation with slightly fluctuating between and within the year due to unforeseen atmospheric changes with the synoptic weather condition in pre-monsoon seasons. The ensemble study showed that the maximum decrease in the net inflow radiation from the urban surface was 242.2 Wm −2 for CMs, 147.3 Wm −2 for CC, and 223.4 Wm −2 for SCMs, but an increase of 39.3 Wm −2 for GRs and 47.6 Wm −2 for VEG compare to control case for 2019 simulation. For 2018 simulation, the rate of reduction of net inflow radiation from the urban surface was 249.5 Wm −2 for CMs, 152.3 Wm −2 for CC, and 227.6 Wm −2 for SCMs, but an increase of 38.8 Wm −2 for GRs and 42.3 Wm −2 for the VEG experiment. So, the energy balances of urban surface from different meteorological conditions simulation consistently produces nearly the same pattern of diurnal as well as spatial 2-day simulation as present in Section 4.1. The urban surface energy balance in 2019 and 2018 simulation results are given in Supporting Information S1 (Tables S1 and S2).
In this study, the impact of heat mitigation technologies on the surface meteorological field, 2-m ambient air temperature, and surface skin temperature were compared with the 2-day simulation. for consecutive year simulation (April for both years) are also presented in Supporting Information S1 (Tables S3  and S4). Different meteorological conditions simulation for the 2 years in the WRF-SLUCM model produced an almost consistent pattern of urban energy as well as thermal balance as compared with a 1-day simulation with the same physical parameterization process within an urban environment are presented in Section 4.2.

Discussions: Limitations, Impacts and Future Research
The primary drawback of the current study is the use of single-day simulations. Lower atmospheric processes cannot be well captured by a single-day simulation. However, because to a lack of long-term observed data over Kolkata, we conducted it for a single day-run with various meteorological conditions for two selected days in the month of April in 2018, 2019, and 2020. Further research can include longer time scales. Compared to meteorological and other climatological studies, there are very few studies on the impact of cool roofs or GRs and their combinations on the urban climate, especially in the tropics. As the air-quality effects of urban surface modification by heat mitigation measures are complex and, nonlinear, comprehensive emissions processing, meteorological and chemical transport models are needed to accurately determine potential impacts on air quality for policy-making purposes. The ambient pollutant concentrations will be affected due to the mixing height and ventilation (A. Sharma et al., 2016). Cool roofs can reduce temperature-dependent emissions of precursors to ozone (O 3 ) and particulate matter in urban areas by lowering ambient temperatures, resulting in a slower rate of volatile organic content evaporation and NOx emissions (Millstein & Menon, 2011). Cool materials might lead to meteorological modifications that may impact ozone and PM 2.5 concentrations besides having enhanced ultraviolet reflection (Epstein et al., 2017). Whether ozone ultimately increases or decreases in the most populated areas will depend on the relative importance of multiple physicochemical pathways, including ozone decreases from temperature reductions, ozone increases from reduced ventilation and mixing, and ozone increases from possible ultra-violet radiation increases.
When assessing the impacts of cool roofs, it is important to consider all the environmental and economic consequences, as benefits from a reduction in heat-related mortality may outweigh the increase in mortality from enhanced PM 2.5 pollution. Although there are other techniques to control ambient air pollution in urban areas through emission reduction policies, the availability of urban heat mitigation strategies is limited in tropical cities. Without a comprehensive analysis of all the benefits of cool roofs it would be a mistake to discourage this technology solely based on air quality alone. Although there are many studies on air quality changes due to mitigation strategies, there is limited research focused on tropical coastal cities. Future research addressing the effects of mitigation strategies on urban air quality parameters are needed, and are beyond the scope of the present research.
CMs and SCMs, as they reflect solar radiation, are likely to reduce vertical mixing and lower PBL heights, negatively impacting urban pollutant dispersion. The lower atmosphere increases pollution and creates additional health problems due to possible urban air quality deterioration in green fraction and reflective materials. The modification in the PBL as a result of the deployment of different mitigation strategies also impacts the temperature and moisture profiles over the urban areas and has potential impacts on cloud and precipitation processes. However, the latter topic of research is the subject of future work. Most of the recent studies have focused only on heatwave episodes or synoptic hot summer days, while the potential long-term (year or decade scale) effects of various mitigation strategies also need to be investigated, especially to understand the effects on precipitation processes (J. Liu & Niyogi, 2019;Patel et al., 2021).
Furthermore, highly reflective materials can reduce the moisture transported to the PBL, and subsequently, cities could be drier (Huang et al., 2023). In other words, GRs can append additional moisture to the PBL with extra latent heat flux and remove air pollutants through dry and wet depositions but entail a significant volume of water for regular irrigation to maintain cooling efficacy and plant activity, which is an added challenge for urban planners. Further, increased coagulation of pollutants with water vapor may create more health problems. Hence, it is challenging to answer the question "which heat mitigation technologies or strategies are suitable for improving thermal fields of the tropical coastal city?" given the multifaceted trade-offs in the energy-water-ecosystem-service nexus. For example, SCM will substantially decrease the cooling needs and thus will decrease the emission of pollutants and AH. On the other hand, GRs will increase the latent cooling load of buildings and can potentially increase pollutants and AH.
Additional consideration must also be made for the vertical thermodynamic profiles. For example, heat mitigation technologies will decrease the maximum convective available potential energy (CAPE) in the urban core and increase the spatio-temporal variability of CAPE (Houston & Niyogi, 2007), which entails that the local landscape modification can affect the regional convective precipitation pattern through land-atmosphere feedbacks (Oh & Sushama, 2021;Sati & Mohan, 2021). With the implementation of heat mitigation technologies, atmospheric moisture is reduced; initially, convective precipitation could decrease over the core urban but enhanced downwind (J. Liu & Niyogi, 2019). This work is the subject of ongoing and future research.
Furthermore, we present a detailed spatio-temporal investigation of several simulated scenarios with a horizontal and vertical extent of PBL variables. The effect of heat mitigation technologies on convective urban rainfall can be measured with cloud microphysics retrieved from satellites and ground-based radars. Determining the real-time interaction between the relatively unexplored behavior of the PBL and air pollution is recommended for future model development, and is not within the scope of this study. No prior study presents holistic impacts of different available mitigation technologies on the surface, energy flux, and PBL dynamics. Most of the previous work has focused only on surface meteorological fields in other climates, with few exceptions (e.g., Brandi et al., 2021). The results presented here provide insight for atmospheric scientists and urban planners alike, enabling prioritization of the spectrum of heat mitigation strategies within a local context.

Conclusion
We studied the impact of the implementation of these scenarios on the regional urban energy budget, on the surface meteorological fields, atmospheric boundary layer dynamics and on sea breeze circulations. This study researched the regional impacts of different heat mitigation technologies on a tropical coastal city using the regional climate model (WRF-SLUCM). The study focused on a hot summer day when heatwaves prevailed.
To allay the single-day simulations concerns, we have added two consecutive very hot summer days during the pre-monsoon month of April to assess how heat mitigation strategies as additional case studies, each of 48-hr duration in the year 2019 and 2018. These additional simulations demonstrate that our initial results are completely consistent with 2020 scenarios. These additional simulations indicate that our initially conclusions are indeed robust. With the physically based SLUCM, the WRF model efficiently captured the meteorological conditions prevalent over the urban area. One base case (control) and five additional numerical simulations (GRs, CMs, SCMs, VEG, i.e., increase in the vegetation fraction and "cool city-CC") were studied. The modification of albedo, emissivity, and rooftop vegetation (roof only), to roof, wall, and ground alter the urban radiation balance.
The impacts of CMs, GRs, "cool cities," and super-cool materials led to lowering the sensible heat, surface skin temperature, and ambient temperature. The increased roof albedo to 0.65 was as effective as the expansion to 100% GRs and CC for the present experiments under the tropical urban environment considered. However, increasing the localized "global" albedo from 0.20 to 0.96 caused a cool island effect and a decrease of the ambient temperature up to 1.6°C compared to the control case. On the other hand, GRs led to higher near-surface relative humidity due to enhanced evapotranspiration. The near-surface relative humidity increased in the CC experiments due to the combined effects of reflective materials and rooftop greening on the surface energy balance.
We showed that the impact of conventional CMs, GRs, vegetative surfaces, and CC on PBL is less notable during the daytime, possibly due to low reflectivity, solar radiation emissivity, and less evapotranspiration rooftop vegetation. However, super-cool materials could significantly lower the height of the PBL due to the considerable reduction of surface absorbed solar energy. In tropical Kolkata, cool and super-cool materials may be promising technologies to reduce the sensible heat flux by about 60%-80% (by ∼150-300 Wm −2 ) compared to the control simulation. The difference of 2 m ambient air temperature between the control and experimental cases showed that the super-cool material simulation showed the largest regional influence for most experiments. The GRs experiment led to a decrease in ΔT ambient of 0.88°C, as a result of rooftop gardening and transpiration.
The regional impacts of heat mitigation technologies have been evaluated using prognostic and diagnostic model-induced meteorological variables. The regional effect is greater for some heat mitigation scenarios than others. The study assessed ambient temperature, surface-specific and relative humidity impacts when the roofs, walls, and ground were covered with high albedo materials or the GRs reached a higher fraction. The changes in the vertical profiles of potential temperature indicate a more stable PBL over the urban areas for cool and super-cool materials compared to the GRs and CC scenario. These highly reflective cool and super-cool materials decreased the PBL height; the diurnal averaged decrease of the model diagnosed PBL height could reach more than 500-600 m over the urban area, although it depends on the physical processes of PBL during day and nighttime. The decrease in ambient temperature and increase in specific/relative humidity due to the urban heat mitigation strategies extended from the urban surface to the top of the PBL. Simultaneously, the intersect phenomena occurred above the PBL due to significant decreases in the vertical wind speed.
The results revealed that cool and super-cool materials are efficient technologies to modify the lower atmosphere of the urban domain. However, the urban heat mitigation technologies can also reduce the vertical mixing over urban areas, may potentially degrading air quality in cities, in alignment with prior work in non-tropical cities (Brandi et al., 2021;Georgescu, 2015). Future research is ongoing to expand these results to more locations in different climates. To conclude, extensive but holistic approach has been undertaken on the cooling potential and local meteorological implications of surface and rooftop heat mitigation methods in tropical urban settings. We evaluated newly developed SCMs alongside other mitigating measures. Practioners may use the findings of this study to acquire a complete understanding of the predicted performance of surface and rooftop heat mitigation measures when applied city-wide to avert extreme urban heat and counteract its impacts on indoor thermal comfort and local thermodynamic efficiency (Ye & Niyogi, 2022).

Data Availability Statement
The software and data that support the findings of this study are openly available.