Quantitative analysis of passive cooling measures in achieving a thermally comfortable urban environment

Abstract

The advancement in technology and rapid globalization have resulted in the emergence of metropolitan cities and agglomerations that are constantly expanding to meet the ever-growing demand of migrating populations. With climate change being an inevitable reality, rampant urbanization has exacerbated impacts such as extreme weather events and the development of urban heat islands. Several researchers and the scientific community have proposed mitigation strategies based on the spatio-temporal changes in Land Use/Land Cover (LU/LC) and Land Surface Temperatures(LST). The studies, however, have not yet established the definite effect of using proposed mitigation strategies on land surface temperature reduction for different urban landscapes. Henceforth, this research aims at quantifying the effect of several mitigation strategies on land surface temperatures, and the analysis was performed over Kolkata city (area under Kolkata municipal corporation). The study focused on developing scenario-based analysis to capture the effectiveness of the interventions required with changing urban density, location and built morphology. The land use and land surface temperature maps developed in the course of the investigation indicated that the impervious urban surfaces correspond to the highest surface temperatures across the study area. In conjunction with the normalized urban heat island (NUHI) index and LST results, the land surface temperature map for Kolkata city was divided into five heat zones. The five priority areas for quantifying the impact of mitigation strategies on LST reduction were chosen from heat zones 4 and 5 (depicting high and very high LST categories as per the global standards). The mitigation scenarios comprising surface greening and changes in building material were applied to the selected areas with varying urban morphologies. The research analysis performed denoted the effect of individual and combined mitigation scenarios based on the efficiency of strategy to bring an optimum cooling effect. It was observed that the cool strategy resulted in the highest temperature difference for both neighbourhood and building levels, whereas the combination scenario corresponded to the highest decline in surface temperatures across the neighbourhood level

Keywords:

• Land surface temperature
• land use/land cover
• thermal effect
• heat zoning and mitigation
• Kolkata city
• Urban Heat Island
• Surface greening
• Building material

Introduction

Today, 55% of the total world population resides in urban areas, which is predicted to increase to 68% by 2050. Urbanization can be understood as a form of paved surface growth to accommodate the increasing human activities that have economic, social, political and geographical implications (Ramachandra et al. 2014). With the advancement in technology coupled with the rising global economy, a more significant proportion of this population shifts towards cities every year for better opportunities and quality of life. Consequently, this has resulted in the emergence of metropolitan cities and agglomerations constantly expanding to fulfil the rising demand for accommodating the migrating population. This rampant urbanization coupled with unchecked population growth has put a strain on the limited natural resources and has also had serious ecological and environmental effects (Nimish et al. 2020). As per projections, the gradually shifting population from rural to urban areas coupled with the overall growth of the world population would contribute to the urban areas by another 2.5 billion people by 2050, 90% of which is expected to take place in Asia and Africa (United Nations, Department of Economic and Social Affairs, Population Division, 2019). Today when climate change has become an unchanging reality of the world, these effects are only exacerbated by the same (Lal 2017). Extreme weather events such as heat waves, floods etc., have become more frequent and global temperatures are rising at an alarming rate.

Urban areas are seen to experience higher temperatures compared to the countryside, resulting in a phenomenon called the Urban Heat Island (UHI) effect, which leads to increased temperature extremes and health risks in the cities (Liang and Shi 2009; Karimi et al. 2021). The conversion of green spaces into impervious paved surfaces, roads, and buildings results in a decrease of the albedo, thus causing the surface temperature to rise (Vujovic et al. 2021). Several researchers across the globe have analysed spatio-temporal changes using land use, land cover and land surface temperatures and proposed several mitigation strategies to alleviate the UHI effect (O’Malley et al. 2014; Nuruzzaman 2015; Salman and Saleem 2021). These mitigation strategies, however, pose different impacts concerning their mitigation potential and the area where they are applied. However, there exists a huge gap in understanding how effectively these strategies can improve the micro-climatic conditions in a neighbourhood. It is thus of utmost importance to quantitatively investigate the effect of each mitigation scenario on LST to help examine the efficiency of each strategy and find a scenario that brings about an optimum cooling effect if implemented. This research aims to develop scenario-based analyses to capture the potency of integrated and individual passive cooling interventions with changing urban density, location and built morphology. The study quantitatively analyses various mitigation strategies on a neighbourhood level to understand and quantify their effect on LST. The analysis was carried out by fulfilling various objectives including (i) Understanding the impact of each land use category on Land Surface Temperature; (ii) Identification of various urban heat centres/pockets; (iii) Quantifying and comparing the efficiency of several mitigation strategies for reducing the overall LST on neighbourhood and building level.

Review of literature

An aggregate increase in the global urban area is estimated to be equivalent to the total land area of Germany, France, Spain and Italy combined by 2050. The onset of urbanization has brought about considerable positive impacts on society such as improvement of the socio-economic status, rise in Gross Domestic Product, better lifestyle, etc. On the other hand, unplanned and rampant urbanization has resulted in a tremendous impact on the environment on a global scale (Weerakoon 2017; Patra et al. 2018). The most direct consequence of unplanned urbanization is the dominance of low-density urban settlements and the change of previously mono-centric compacted cities into dispersed, discontinuous, fragmented, and polycentric urban patterns (Taubenböck et al. 2009). Urban sprawl has been studied in various regions around the world and conclusively impacts climate, agriculture and the environment (Bharath et al. 2013) (Prakash et al. 2021). simulated the future urbanization patterns of Bangalore city and inferred that inadvertent urbanization would severely erode the sustenance of natural resources in upcoming years (Wang et al. 2017). analysed the megacities of China from 1990 to 2010 and concluded that economic growth, urbanization, and industrialization result in increased carbon emissions (Seto and Shepherd 2009). stated that urbanization impacts climate change systems – surface temperatures, clouds and precipitation, surface hydrology, carbon and nitrogen cycle. Alterations in the landscape can be quantitatively evaluated by developing land use/land cover maps for the area of interest.

Land use dynamics have seen a steep change due to rapid urbanization over the years. Various studies indicate that it has led to unprecedented growth in cities, especially in the form of built-up areas that have resulted in a considerable decline in vegetative cover showing that the cities are growing at the cost of green cover (Reis 2008; Bharath et al., 2018; Parveen et al. 2018; Shahfahad et al. 2020; Chandan et al. 2021). Urban Heat Island, which is defined as warmer urban areas than the surrounding rural areas (Luber and McGeehin 2008; Hajat et al. 2010), are the best illustration of how changes in LULC impact the local and regional climate (Arnfield 2003; Yow 2007; Pielke et al. 2011; Chun and Guldmann 2014; Halder et al. 2022; Shahfahad et al. 2022). UHI is directly associated with the rising LST and is the main causal factor of the heatwaves. UHI formation can be directly connected to the LULC of a region since it is ascribed to the difference in the albedo and heat retention capacity of artificial and natural fabrics over an urban area. Heat waves and severe temperatures are observable repercussions of climate change on human health, infrastructure, and ecology all across the world. However, these impacts are significantly detrimental in densely populated and urbanized cities and metropolitan areas (Aram et al. 2020). Research analysis shows the rising issues of urban heat islands in several Indian cities, owing to large variations in the region’s landscape (Talukdar et al. 2021; Mohammad and Goswami 2022; Gohain et al. 2023). It acknowledged that diminished green cover in addition to increased impervious surfaces leads to a reduction in the amount of evapotranspiration and latent heat flux in cities, attributing more energy to sensible heat (Oke 1982; Grimmond and Oke 1991). Scientific studies have shown that UHI severely affects building energy usage, particularly during the summer, and the safety of several susceptible population segments (Santamouris 2014; Battista et al. 2023). Rising temperatures have different effects depending on age, geography, gender, social and economic status. These alterations largely impact the local meteorology (local wind patterns, development of clouds and fog, humidity levels and amount of precipitation). Additionally, it leads to power outages, increased mortality rates, deteriorated air quality, rise in extreme events and disasters, increased thermal discomfort, food and water crisis and poor health. Subsequently, UHI can also be attributed to the increase in urban pollutants concentration (SO_X, NO_X, ozone) (Stathopoulou et al. 2008), and the city’s carbon footprint (Santamouris 2007).

According to the Intergovernmental Panel on Climate Change (IPCC), the average global temperature increased by 0.85 °C from 1880 to 2012, the global average sea level has risen by 19 centimetres from 1901 to 2010, the arctic sea shrunk by 1.07 million sq. km of ice every decade post-1979 and 2100 the increase in global temperature is expected to increase 1.5 °C. Global warming is thus one of the greatest concerns of today’s climate crisis which coupled with the urban heat island effect has exposed city residents to higher heat stress risk than rural dwellers (Chapman et al. 2017). Landscape temperatures are directly linked with the sensible and latent heat fluxes, therefore, holding a pivotal role in defining local, regional as well as global processes over the surface of the earth (Mannstein 1987) and are an important indicator of global warming and the greenhouse effect (Jia et al. 2007).

Thus, various mitigation measures are urgently required on building and neighbourhood level to solve the issue of thermal discomfort at the micro-scale and source itself. Studies have been conducted on urban heat islands and control strategies to handle and mitigate steeply rising land surface temperatures in urban areas. It can be concluded that mitigation strategies can be categorized into two major categories (Sailor 2006). First, is increasing the albedo of urban surface which is accomplished through high albedo roofing and paving technologies (Zinzi and Fasano 2009; Li et al. 2014; Costanzo et al. 2016). Second is increasing evapotranspiration, which can be achieved through a combination of decreasing the fraction of impervious surfaces and planting vegetation in urban areas such as trees, vertical gardening, roof greening, etc (Aflaki et al. 2017; Chatzinikolaou et al., 2018; He 2019). Additionally, a combination of two or more measures can be applied to estimate the effectiveness of integrated technologies for reducing the overall thermal-related issues for the residents. This analysis can serve as a piece of literature that would be useful in determining neighbourhood-level region-specific mitigation measures for small urban pockets with varying building typologies.

Methodology and study area

Study area

One of the major cities of India was considered for analysis based on infrastructure (planned and unplanned), level of urbanization, socio-economic factors, population, etc. The city of Kolkata in the eastern part of India (Figure 1) was selected for conducting this analysis. It is the seventh largest city in India and is rapidly growing in population density and urbanization. According to Census (2011), Kolkata has witnessed a continual rise in population, however, the rate of rise has been reducing, owing to saturation in the available land for construction. It consists of 144 wards that lie under the jurisdiction of the Kolkata Municipal Corporation (KMC). Historically, mass migration has boosted Kolkata, creating a demand for more residential areas. LULC modifications are changing rapidly and intruding on the natural landscape at the same time. Kolkata is located in the Ganges delta and its climate is greatly influenced by river Hooghly. The hot and humid climate is characterized by three distinct seasons, summers, monsoons and winters. The maximum temperatures during the summers cross 40 °C in April followed by the monsoons extending up to October with maximum rainfall experienced in July (350 mm). As a result of increased anthropogenic interventions, the city is facing an issue of deteriorated air quality (increased pollution levels) due to construction activities, alteration in land use and traffic congestion. Additionally, the city is facing micro-climatic changes, which are directly impacting the thermal comfortability and well-being of the residents.

Figure 1. Study area – kolkata city.

Datasets

The study employs the Landsat satellite dataset at different wavelength ranges to develop land use maps and estimate land surface temperature changes. The selected years’ (as mentioned in Table 1) images were acquired from the United States Geological Survey (USGS) earth explorer portal. Secondary data was collected via Google Earth and Bhuvan. The georeferenced images for specific dates from the multi-spectral optical range bands were utilized for preparing LULC maps and calculating normalized indices maps, while the images from the band associated with thermal range of the Electromagnetic (EM) spectrum were employed for the preparation of LST maps.

Table 1. Dataset used for analysis.

Data	Source	Purpose
Landsat 5 (TM) − 1999, 2009	USGS Earth Explorer (https://earthexplorer.usgs.gov)	Land use and land cover analysis, quantification of land surface temperature
Landsat 8 (OLI/TIRS) − 2017, 2021
Google Earth	https://earth.google.com/web	Validation and classification
Survey of India toposheets and Bhuvan	www.surveyofindia.gov.in/bhuvan. nrsc.gov.in/	Derivation of the base layer (Administrative boundary)

Land use analysis

The retrieved maps and remotely sensed data were cropped to the selected study area. The study area was classified into four land use classes: built-up, vegetation, water and others using Gaussian Maximum Likelihood classifier, which is a supervised classification algorithm that classifies data based on probability density function (Bharath et al., 2018). The algorithm considers the mean and covariance matrices for categorizing each pixel into various classes. Furthermore, the classification was evaluated by performing an accuracy assessment using ground truth data and Google Earth for 2009, 2017 and 2021, while for 1999, land use was evaluated for accuracy via 70% and 30% of signatures used for developing classified map and reference map respectively. Accuracy assessment was derived by calculating overall accuracy and kappa coefficient, via a confusion/error matrix. An elaborated explanation of the entire procedure is mentioned in (Nimish et al. 2020).

Land surface temperature retrieval

The range 10.4−12.5 μm in the electromagnetic spectrum is the thermal infrared region. The bands that sense in these regions were used to extract land surface temperature for the selected years. The quantification was carried out using a single-channel algorithm, which is one of the most common, easy-to-apply and widely accepted methods to estimate LST. The algorithm can be split into a three-step procedure: generation of thermal thematic maps, estimation of land surface emissivity and quantification of land surface temperature. The equations used to calculate LST are shown in Table 2. The entire method is detailed in (Nimish et al. 2018).

Table 2. Equations - quantification of LST.

Quantification of LST	Equation used
Conversion of DN into spectral radiance	$L_{\lambda }=(\mathit{\text{Gain X DN}})+\mathit{\text{Offset}}$
Conversion of radiance into at-satellite brightness temperature	$T_B=\frac{K_2}{\mathrm{\text{ln}}\hspace{0.17em}\left(\frac{K_1}{L_{\lambda }}+1\right)}-273.15$
Conversion of at-satellite brightness temperature into LST	$T_s=\frac{T_B}{1+\left(\frac{\lambda T_B}{\rho } X \mathrm{\text{ln}}(\epsilon )\right)}$

Here, λ is the wavelength of the thermal band at which the relative response is maximum with respect to wavelength (μm); L_λ is spectral/at-satellite radiance for wavelength ‘λ’; DN is digital number; Gain and Offset are constants obtained from Metadata; T_B is at-satellite brightness temperature; K₁ and K₂ are calibration constants; T_s is Land Surface Temperature; ε is the emissivity of the element; $\rho =\frac{\mathit{\text{hc}}}{\sigma }=1.438\mathrm{ }X\mathrm{ }{10}^{-2}\mathit{\text{mK}}$ (h is plank’s constant, c is speed of light, σ is Stefan Boltzmann constant)

Identification of urban heat centres

The priority areas that require immediate intervention in terms of alleviating the effect of urban heat islands were identified based on three parameters:

1. Normalized Urban Heat Island Index (NUHII):

   It is a measure that is used to define the intensity of urban heat islands throughout the area under consideration (Nimish et al. 2020). Figure 2 elaborates on the steps involved in calculating NUHII for the study area.

2. Categorizing based on LST values:

   The study area was divided into 5 zones for the year 2021 (Table 3), based on land surface temperature ranges. The classification was done using the mean LST value and standard deviation as demonstrated in (Xu et al. 2011).

3. Based on LST classes and Universal Thermal Climate Index (UTCI):

   Land surface temperature (LST) serves as a micro-indicator of UHI effect. It represents the temperature of the ground which can influence the thermal conditions experienced by humans in outdoor environments. Higher LST values often correlate with warmer ambient temperatures, which can impact human comfort.

   UTCI is the measure of the human physiological reaction to the thermal environment expressed in degrees Celsius. The universal thermal climate index (UTCI) denotes the synergistic heat exchanges that occur between the thermal environment and the human body, including its energy budget, physiology, and clothing. UTCI takes into account population clothing adaption in response to actual ambient temperature. The UTCI is calculated using four variables: 2 m air temperature, 2 m dew point temperature (or relative humidity), 10 m wind speed, and mean radiant temperature (MRT) and the range is described in Table 4.

Figure 2. Flowchart showing steps for identification of urban heat centres.

Table 3. Table showing the division and identification of urban heat centres.

Zone	Temperature Class	Class Range	As Calculated from LST map of 2021 (°C)
1	Very low temperature	T ≤ Tmean − 1.5STD	≤ 32.82
2	Low temperature	Tmean − 1.5STD < T ≤ Tmean - STD	32.82 < T ≤ 33.92
3	Medium temperature	Tmean – STD < T ≤ Tmean + STD	33.92 < T ≤ 38.32
4	High temperature	Tmean + STD < T ≤ Tmean + 1.5STD	38.32 T ≤ 39.42
5	Very High Temperature	> Tmean + 1.5STD	> 39.42

T_mean = Mean temperature.

STD = Standard deviation.

Table 4. Universal thermal comfort index and physiological stress categories. Source: (Copernicus 2018).

Physiological Stress Categories	UTCI Range (°C)
Very Strong Cold Stress	< − 40
Strong Cold Stress	− 40 to − 27
Moderate Cold Stress	− 3 to 0
Slight Cold Stress	0 to 9
No Thermal Stress	9 to 26
Moderate Heat Stress	26 to 32
Strong Heat Stress	32 to 38
Very Strong Heat Stress	38 to 46
Extreme Heat Stress	> 46

Several studies (Huang et al. 2016, 2023) have used the UTCI stress categories (Table 4) to evaluate outdoor thermal comfort. There are numerous indices that have been developed and used to assess thermal comfort over the last few years (Khaire et al. 2024). However, UTCI has been preferred in this study as it has proven to surpass other indices in terms of physiological relevance as it is more closely aligned with human perception. In addition, The World Meteorological Organization (WMO) has endorsed UTCI as an international standard for assessing thermal comfort.

It is well established that the land surface temperature is greater than the air temperature. The difference between LST and air temperature ranges from 2 – 7 °C depending upon the cloudy or clear sky conditions (Gallo et al. 2011). Thus, an average of 4 °C was taken to calculate the corresponding air temperatures for the obtained land surface temperatures for Kolkata City. Thus, the heat zones from the LST and NUHII maps were obtained and compared with UTCI temperature ranges and their corresponding physiological stress categories. The overlapping areas indicate heat zones, where mitigation strategies should be applied to minimize the rising LSTs. Further, Urban Heat Centres (UHCs) were identified from heat zones that exhibited the highest land surface temperatures and UHI intensity. They were divided into areas of 200$\times$200-meter grids and further analysis was performed on the derived area.

Development of mitigation scenarios and materials used

The key urban elements that must be addressed for urban planning, and which have a significant influence on the variance of land surface temperature on a local scale are the areas with buildings, green spaces and pavements. Mitigation strategies, therefore, must be built around them (Gago et al. 2013). The scenarios thus, to be underlined for the study area would revolve around (as described in Figure 3):

• Base Scenario: To compare the impact of mitigation strategies on the selected areas the first scenario considered was the business-as-usual scenario, where the building material chosen is a conventional burnt clay brick and the roof material is concrete. The building envelope was modelled with the same amount of vegetation/trees present in the selected study area with no additional greening on the roof or façade. The material selected for roads is dark asphalt.

  • Material used: The material used for modelling the built environment, including the walls, roofs and pavements, along with their characteristic thermal properties, are listed in Table 5.

• Green Scenario: Based on the urban morphology and the hot and humid climate of Kolkata city, the different forms of urban greening were applied to the selected study areas:

  • Roadside Greening: Trees in parks and around commercial areas cause a significant reduction in temperature due to the combined effect of shading and evapotranspiration. They may create cool islands in the city (Shashua-Bar and Hoffman 2000).

  • Façade Greening: It aids in protecting against the solar radiant heat and decreases glare and sound absorption. Green walls also facilitate filtering the air, stabilising the microclimate, and providing a humidity regulation effect (Aflaki et al. 2017).

  • Roof Greening: Green roofs aid in keeping lower air temperatures during the day and higher at night in the summers. Energy consumption in buildings with green roofs was found to be lower than those without such roofs, and could even be improved by natural ventilation during the summer (McPherson et al. 1988; Aflaki et al. 2017).

• Material used: Plants that are suitable for the hot and humid climate of Kolkata were selected for the different forms of urban greening. The material listing is as tabulated in Table 6

  • Roof Greening: It is essential to consider the qualities such as availability, ease of maintenance, growth period, and ability to withstand the climate while choosing grass for the roof. Therefore, for the selected study area Zoysia Grass was chosen. It is easily available and can withstand all seasons. It is a fast-growing grass and is easy to maintain (Gardening 2020).

  • Façade Greening: the plants selected for vertical greening must also satisfy the conditions of being light and weight, easy to maintain, readily available and resilient to harsh summer heat. The plant thus selected for façade greening is Ficus Panda (Gardening 2020), which is readily available in and suitable to Kolkata city.

  • The soil used as the substrate for the selected grass is Sandy Loam Soil. The prerequisite for the substrate is that it must be light in weight, nutritious and provide good drainage. Loamy soil is readily available in Kolkata and can easily support grass (Gardening 2020).

• Cool Scenario: The temperature on summer days can be reduced by altering the materials used in buildings and pavements due to each material’s albedo. High albedo materials when used, result in lesser absorption of solar radiation by the building envelopes and urban structures. Subsequently leading to the lowering of land surface temperatures (Taha 1997). The strategies in which cooling can be achieved through changing surface albedo for the selected study areas were selected:

  • Cool Roofs: The roof is an important part of the building envelope since it directly impacts the building’s energy consumption and the thermal comfort of the inhabitants. Cool roofs work by collecting less heat and reflecting more of the sunlight that falls on the roof back into the atmosphere than a conventional roof surface (NRDC 2018).

  • Reflective Pavements: In a traditional grid plan, sidewalks take up around 16 per cent of the ground surface. In rectangular parts typical of communal housing complexes, this percentage can reach as high as 23% (Gago et al. 2013). Rough and dark-coloured surfaces such as the typical asphalt road absorb more heat than lighter pavements, thus, exhibiting high temperatures in warm climates (Doulos et al. 2004).

  • Building Material: Using energy-efficient materials for construction instead of the conventional clay brick that exhibits high thermal conductivity may reduce the energy spent for climate control in building environments. Insulating construction materials may conduct a lesser amount of heat than clay bricks. The alternative material that was thus used for the study is fly ash bricks. These are unconventional bricks manufactured from the wastes generated by industries, namely, fly ash mixed with cement and sand. Fly ash bricks absorb a lesser amount of heat, thus keeping the buildings cooler in the summer.

    Materials used:

  • Cool Roof-. The material selected for the cool roof scenario is Acrylic base high reflectance paint coated on the roof to reflect sunlight to the extent that the roof does not get heated up during harsh summers (Lu et al. 2019). It is commonly available and easy to apply with minimum maintenance. The materials are tabulated in Table 7.

  • Reflective Pavement- Unlike the dark asphalt road having low reflective properties, Thin Yellow Asphalt is selected as road material for this scenario. The yellow thin layer of asphalt mixes elastomeric asphalt binder with infrared reflective pigments and aggregates (Kyriakodis and Santamouris 2018).

  • Material Scenario- The alternative material selected in place of energy-intensive clay brick is ACC fly ash brick. It is light in weight, load-bearing and possesses good insulation properties. It can maintain optimal ambient temperature by keeping the inside air cool during the summers and warm during the winters, thus saving energy on climate control which indirectly contributes to UHI (Makaka 2014).

• Combination Scenario: A scenario was developed including all the mitigation strategies, namely, greening, cool roofs, reflective pavement and altering of building material to understand the variation in surface temperatures. The aim is to reduce the number of impervious surfaces by adding greenery and increasing the reflectivity of sealed surfaces. Thus 50% of the roofs were greened, whereas the rest were coated with high reflectance paint, in addition to replacing all dark asphalt roads with reflective pavements and finally façade and roadside greening according to the urban canyon layout of each selected study area.

Figure 3. Scenarios considered for the study.

Table 5. Listing material and their properties used in base scenario (source: https://material-properties.org/).

Parameter	Material			Inference
	Burnt Brick	Concrete Roof	Asphalt Road
Absorption Coefficient	0.60	0.70	0.85	The materials used exhibit a low value for reflection and therefore result in a high amount of heat release in the form of long-wave radiation. Very high emissivity suggests the materials emit almost all the heat incident.
Transmission	–	–	–
Reflection Coefficient	0.40	0.30	0.15
Emissivity	0.90	0.90	0.90
Specific Heat (J/Kg*K)	650	840	900
Thermal Conductivity (W/m*K)	0.60	1.90	2.00

Table 6. Listing materials and their properties for green scenario (source: (Ntoulas et al. 2013; Wei et al. 2020)).

Parameter	Material			Inference
	Zoysia Grass	Ficus Panda	Sandy Loam Soil
Albedo	0.2	0.2	0.2	Greening materials are seen to have a low albedo which signifies a lower reflection of heat. However, they contribute to cooling due to evapotranspiration and shading.
Volumetric Heat capacity (J/m^3 K)	–	–	1.32
Transmittance	0.30	0.30	–

Table 7. Listing material and their properties for cool scenario.

Parameter	Material			Inference
	Acrylic base high reflectance paint	Thin yellow asphalt	AAC Fly ash brick
Absorption Coefficient	0.04	0.35	0.52	The materials that exhibit considerably higher values for reflection remain cool and therefore prevent surface temperatures from rising.
Transmission Coefficient	0.01	–	–
Reflection Coefficient	0.95	0.65	0.48
Emissivity	0.94	0.90	0.75
Specific Heat (J/Kg*K)	1350	–	950
Thermal Conductivity (W/m*K)	0.20	–	0.25

(Source: https://material-properties.org/; [Kyriakodis et al. 2018; Gourav et al. 2017]).

Materials used

The overall summarized structure of the analysis is described in Figure 4.

Figure 4. Workflow for modelling mitigation scenarios for identified urban heat centres.

Modelling and simulation

Mitigation scenarios identified were simulated and modelled on Envi-met for each of the selected study areas to understand and quantify their effect on the surface temperatures as shown in Figure 5.

Figure 5. Method for simulation on Envi-met software (Chatzinikolaou et al., 2018).

Volume Construction: The building elements were constructed using building footprints and heights that were obtained from Google Earth and Digital Elevation Model (DEM). The heights of the buildings were calculated with the help of Digital Surface Model (DSM) and Digital Terrain Model (DTM) at 30 m resolution, which was then combined with the building boundaries to model the study areas.

Microclimate preparation: Microclimatic data was collected from the nearest weather station to examine the effect of mitigation strategies on the same. The meteorological parameters that were considered and the temporal variation of temperature, relative humidity, wind speed and wind direction were analyzed.

• Model Parameters –

Table

Simulation Date: 25-05-2021	Duration: 24 h	Wind Speed: 7.78 mph
Wind direction: 22°	Min Temperature: 21 °C	Max Temperature: 38 °C
Min Humidity: 39 %	Max Humidity: 75%

The simulation of the temporal variation of several thermodynamic parameters on a micro-scale range was carried out on the summer day of May 25th, 2021, by creating a three-dimensional model of resolution 2 m $\times$2m $\times$2m for all five of the selected study areas. The software Envi-met was used for simulation which is based on the Reynold’s Averaged Navier Stokes equation and the Energy Balance equation. These equations govern the heat transfer processes occurring within the simulated environment. It allowed three-dimensional non-hydrostatic modelling of building-air-vegetation interactions inside the desired urban environment. A typical time-space resolution of the model of 10 s/0.5–10 m was chosen (Ambrosini et al. 2014). The key components and models used for simulating the surface temperature for this study are:

1. Thermal Properties: thermal properties of for each material used, such as reflectance, absorptance, emissivity and thermal conductivity etc. was included, to accurately simulate the exchange of heat energy between surfaces and the surrounding environment.

2. Vegetation Model: a detailed vegetation model accounting for the influence of vegetation on surface temperatures through processes such as transpiration, shading, and interception of solar radiation.

3. Building Surfaces: the thermal properties as well as the orientation of building surfaces, including walls and roofs were considered to simulate their impact on surface temperatures.

Results and analysis

Land use Analysis

The study area was classified into four land use classes namely- built-up, vegetation, water and others using a supervised classification algorithm (Gaussian Maximum Likelihood classification) (Bharath et al., 2018). This classifies the data based on the probability density function while considering the mean and covariance matrices. The temporal land-use analysis shows that there is an increase in the percentage cover of urban area from 47.6% to 75% during the period of 1999 to 2021 (Table 8). The stark decrease in the percentage cover of vegetation shows that urban vegetated land has been converted to built-up areas for residential and/or commercial purposes.

Table 8. Percentage covered by each land use class for the years 1999, 2009, 2017 and 2021.

Land Use	1999	2009	2017	2021
Urban	47.26 %	59.48 %	68.24 %	75 %
Vegetation	37.52 %	35.51 %	27.48 %	19.61 %
Water	3.75 %	3.53 %	3.95 %	3.44 %
Others	11.47 %	1.48 %	0.33 %	1.68 %

Land surface temperature analysis

The land surface temperature map was retrieved for selected years using single channel algorithm and it was observed that the higher LSTs were associated with built environments. Land surface temperature distributions across the kolkata city area for the years 1999, 2009, 2017 and 2021 shows maximum values for the Northern region (Figure 6), which is characterised by dense commercial areas. The values of LST range are tabulated in Table 9. The surface temperatures were observed to decline from North to South kolkata, owing to lower urban density. A considerable temperature drop is seen on the Southern periphery of the city, which comprises sparsely built areas and is dominated by vegetation. The least temperatures were observed in the easternmost region of the city, which comprises irrigated agricultural fields. The low temperatures can be signified by the presence of water bodies in the form of aquaculture.

Figure 6. Land surface temperature distribution across kolkata city for 1999, 2009, 2017 and 2021 (LST legend is in °C).

Table 9. Land surface temperatures for 1999, 2009, 2017 and 2021.

Land use	Year	Min. Temp. (°C)	Max. Temp. (°C)	Mean Temp. (°C)	Year	Min. Temp. (°C)	Max. Temp. (°C)	Mean Temp. (°C)
Urban	1999	27.72	46.34	38.67	2009	25.64	43.29	36.03
Vegetation		27.97	44.82	34.69		26.40	41.74	33.39
Water		25.55	44.06	30.86		25.63	38.59	29.99
Others		24.68	44.06	35.23		26.40	37.79	30.12
Urban	2017	23.20	35.30	29.37	2021	28.42	43.05	36.82
Vegetation		22.37	32.95	26.99		28.04	40.61	34.25
Water		22.12	32.67	25.23		28.07	40.90	31.56
Others		22.38	29.09	24.28		30.98	42.47	35.80

Urban Heat Centres

1. Normalized Urban Heat Island Index (NUHII)

   The NUHII map obtained from the LST of rural (Figure 7) areas as well as maximum and minimum LST (2021), helped in identifying the areas exhibiting the highest intensity of heat islands (0.41-0.65). It could be observed that higher values of NUHII coincide with the regions having the highest LSTs. The Northern region in kolkata city comprises high-density urban areas (particularly commercial developments), which corresponds to the highest intensity of heat islands. Almost the entire study area shows positive values for NUHII except for water bodies and vegetated spaces.

2. Classification of Land Surface Temperature of urban area based on mean temperature values and standard deviation

   After classifying the LST map (2021) into the five heat zones (Figure 8) according to the method demonstrated by (Xu et al. 2011), the corresponding land surface temperatures are calculated for the defined UTCI stress categories as shown in Table 10.

Figure 7. NUHII map for kolkata city for the year 2021.

Figure 8. Heat zone map for kolkata city for the year 2021.

Table 10. Land surface temperature and UTCI stress categories.

Zone	Temperature Class	Class Range	As Calculated (°C) from LST map of 2021
1.	Very low temperature	< Tmean − 1.5STD	≤ 32.82
2.	Low temperature	Tmean − 1.5STD < T ≤ Tmean - STD	32.82 < T ≤ 33.92
3.	Medium temperature	Tmean – STD < T ≤ Tmean + STD	33.92 < T ≤ 38.32
4.	High temperature	Tmean + STD < T ≤ Tmean + 1.5STD	38.32 < T ≤ 39.42
5.	Very High	> Tmean + 1.5STD	>39.42

The high (zone 4) and very high (zone 5) land surface temperature zones correspond to strong and very strong heat stress categories on the UTCI index. Furthermore, the LST map was compared with the NUHII map and it was found that the high and very high LST zones coincide with those exhibiting high NUHII. The areas, therefore, lying in zone 4 and 5 exhibit high LST and urban heat island intensity and require immediate intervention to alleviate the same as shown in Figure 8.

Selection of sites based on urban density, heat zones and NUHII

The Urban heat centres were to be identified from heat zones 4 and 5 (Figure 9) which exhibited the highest land surface temperatures and UHI intensity. They were further divided into grids of 200 × 200 metres and the following areas were chosen as urban heat centres:

• Shobha Bazaar (high-density urban area)

• Rabindra Sadan (high-density urban area)

Figure 9. Heat zone maps divided into grids of 200 × 200 m.

The remaining three Urban Heat Centres were selected from heat zone 3 and were moderate-density areas, taking into account the increasing trend in surface temperatures and urban land use class (Figure 10).

• Rabindra Sadan (moderate-density urban area)

• Rabindra Sarovar (moderate-density urban area)

• Phoolbagan (moderate-density urban area)

Figure 10. Satellite image of selected sites- UHC 1, UHC 2, UHC 3, UHC 4 and UHC 5 respectively.

The sites selected for study sprawled throughout the city of Kolkata and multiple visits were made to each. Observations related to the location, density, height, vegetation, etc. were made and are shown in Table 11.

Table 11. Details of selected sites.

Location	Shobha Bazaar, North Kolkata	Rabindra Sadan, West Kolkata	Watgunj, West Kolkata	Rabindra Sarovar Lake, South Kolkata	Phoolbagan, Central Kolkata
Density	High-density urban area ∼105 buildings	High-density urban area ∼57 buildings	Low density, high rise residential apartments ∼4 buildings	Low density, residential apartments ∼8 buildings	Moderate-density (Mixed use) ∼23 buildings
Roof type	Flat roof structures	Flat roof structures	Flat roof structures	Flat roof structures	Flat roof structures
Height	Ranging from G to G + 1, mostly G	Ranging from G to G + 2, mostly G + 1	Ranging from G + 3 to G + 12	Ranging from G + 2 to G + 3	Mostly G + 2
Vegetation	Minimum Vegetation	Minimum Vegetation	Some Trees	A good number of trees	Trees along roads and in some cases buildings
Area	100m x 100 m	100m x 100 m	120m x 120 m	120m x 120 m	100m x 100 m

Mitigation scenarios

The mitigation scenarios (Table 12) created were simulated for each selected study area, which was modelled with existing conditions pertaining to the built environment and climate. The surface temperature was evaluated for each area considering the meteorological parameters for the 25th day of May 2021. Analysis was carried out on two levels namely, the neighbourhood level and the building level for 2:00 pm which is the hottest time of the day. The following are the specifications for each of the simulated scenarios for the chosen areas of study.

Table 12. Specification of the scenarios.

Scenario	Material	UHC 1	UHC 2	UHC 3	UHC 4	UHC 5
Base Scenario	Burnt Brick Walls	✓	✓	✓	✓	✓
	Concrete Roof	✓	✓	✓	✓	✓
	Asphalt pavement	✓	✓	✓	✓	✓
Green Scenario	Roof Greening: Zoysia Grass	✓	✓	✓	✓	✓
	Façade Greening: Ficus Panda	✗	✗	✓	✓	✓
	Roadside vegetation	✗	✗	✓	✓	✓
Cool Scenario	Cool roof: Acrylic-based high-reflectance paint	✓	✓	✓	✓	✓
	Reflecting pavement: thin yellow asphalt	✗	✗	✓	✓	✓
Material Scenario	AAC fly ash brick walls	✓	✓	✓	✓	✓
	Acrylic-based high-reflectance paint on the roof	✓	✓	✓	✓	✓
Combination Scenario	All of the above	✓	✓	✓	✓	✓

Simulation results

1. Neighbourhood Level - The variation in temperatures was observed on the ground surface level around the buildings. The following are the resulting LST ranges for each of the simulated scenarios over the selected sites as described for all scenarios below in Figure 11.

   Figure 11. Surface temperature variation across UHC-1, UHC-2, UHC-3, UHC-4 and UHC-5 for (a) base, (b) green, (c) cool, (d) material, (e) combination scenario at the neighbourhood level.

   
   

   1. Base Scenario:

      The base scenario constituted the study areas in their existing condition in terms of building material as well as built morphology. The surface temperature ranged from 30.20 °C – 54.69 °C, 33.74 °C – 54.57 °C, 23.09 °C – 52.25 °C, 33.49 °C – 53.95 °C and 29.99 °C – 53.65 °C for UHC 1, UHC 2, UHC 3, UHC 4 and UHC 5 respectively. The maximum temperatures were associated with the roads made of dark asphalt for all the selected study areas. For UHC 1 and 2 in Shobha bazaar and Rabindra Sadan area, characterized by highly dense urban layout with minimal vegetation, high temperatures (44.00 °C – 54.69 °C and 44.15 °C – 54.57 °C) were distributed throughout except for the narrow roads between the buildings due to mutual shading. UHC 3 and 4 in Watganj and Rabindra Sarovar Lake areas, which are planned residential societies show a prominent variation in temperatures. The highest temperatures (36.05 °C – 52.25 °C and 40.31 °C – 53.95 °C) were exhibited by the paved pathways around the buildings that are made of dark concrete, while the shaded parts were observed to correspond to mild temperatures (32.81 °C – 39.29 °C and 35.28 °C – 38.37 °C). Both areas consist of vegetation in the form of grass patches and trees where the lowest temperatures were observed (23.09 °C – 36.05 °C). Lastly, UHC 5 in Phoolbagan, which is characterized by a dense urban layout show similar temperature distribution to UHC 1 and 2 except for lower temperatures confined to the areas covered with trees.

   2. Green Scenario:

      Strategies namely, roof, façade and roadside greening were modelled for each study area as per their built morphology. UHC 1 and 2 allowed only roof greening owing to the dense building layout. The surface temperatures ranged between 26.75 °C – 53.92 °C and 30.44 °C – 54.16 °C respectively, showing a slight decrease (0.44 °C – 1.34 °C) when compared to the base scenario. UHC 3, 4 and 5 allowed all three greening strategies to be simulated, however, only roadside greening, that is the addition of trees shows a considerable effect on the surface temperatures. The temperature ranged from 23.09 °C – 49.01 °C, 23.12 °C – 52.58 °C and 26.65 °C – 53.84 °C for UHC 3, UHC 4 and UHC 5 respectively. Increasing the number of trees throughout the area resulted in surface temperature reduction and cooler temperatures (23.09 °C – 29.57 °C, 26.40 °C – 36.22 °C and 26.61 °C – 36.82 °C) were found to be associated with a larger percentage of the area when compared to the base scenario.

   3. Cool Scenario:

      Strategies such as painting the roof and pavement with high-reflectance paint were adopted for the cool scenario. For high-density urban layouts of UHC 1 and UHC 2, only a cool roof strategy was allowed due to the narrow roads. The surface temperatures ranged between 30.31 °C – 54.70 °C and 30.46 °C – 55.20 °C for UHC 1 and UHC 2 respectively. In the case of UHC 3, 4 and 5 both the roof and pavements were coated with high-reflectance materials. The surface temperatures range between 23.10 °C – 52.31 °C, 23.10 °C – 52.31 °C and 27.68 °C – 45.93 °C respectively. Apart from the surfaces shaded from the building and trees, the pavements directly exposed to the sun exhibit temperatures in the range 36.08 °C – 42.57 °C, 38.37 °C – 47.63 °C and 34.43 °C – 42.26 °C respectively, showing a decline in LST compared to those seen in the base scenario.

   4. Material Scenario:

      A sub-scenario in addition to reflective materials, involved the replacing of the burnt clay bricks as building material with fly ash bricks, for all the selected study areas. It was observed that a very slight reduction in surface temperature was observed as compared to the cool scenario for all the areas. The surface temperatures ranged between 30.34 °C – 54.82 °C, 30.32 °C – 54.74 °C, 23.06 °C – 51.97 °C, 28.92 °C – 50.10 °C and 27.33 °C – 44.77 °C for UHC 1, UHC 2, UHC 3, UHC 4 and UHC 5 respectively.

   5. Combination Scenario:

      The final scenario constituted an integration of all the mitigation strategies for each study area to understand their effectiveness in combination. For UHC 1 and 2 all pavements were replaced with high reflectance material and the burnt clay bricks with fly ash bricks. 50% of the rooftops were greened while the rest were coated with high-reflectance paint. Both UHCs 1 and 2 show considerable LST differences in the case of combination scenario and the temperatures range from 29.88 °C – 49.93 °C and 27.82 °C – 46.42 °C for UHC 1 and UHC 2 respectively. For UHC 3, 4 and 5 in addition to all the strategies mentioned above roadside greening in the form of trees was considered wherever possible. The ground surface temperatures ranged from 22.60 °C – 47.36 °C, 25.08 °C – 46.01 °C and 27.55 °C – 45.52 °C for UHC 3, UHC 4 and UHC 5 respectively. A significant drop in LST was observed, and milder temperatures were identified to cover a considerably larger area than all the other scenarios.

      Further, a point was selected within each study area to analyse the temperature difference for each scenario. A graph was plotted to compare the impact of each scenario on the surface temperature for that point (Table 13 and Figure 12). The following are maps showing the surface temperature distribution across each study area for every mitigation scenario on the neighbourhood level.

      Figure 12. Graphs denoting variation in mean surface temperature at selected point for each scenario over UHC-1, UHC-2, UHC-3, UHC-4 and UHC-5 on 25.05.2021 at neighbourhood level.

      

      Table 13. Maximum, minimum & mean surface temperature at selected point for each scenario over UHC-1, UHC-2, UHC-3, UHC-4 and UHC-5 on 25.05.2021 at neighbourhood level.

      Scenario​	UHC - 1			UHC - 2
      	Surface Temperature			Surface Temperature
      	Max (C°)​	Min (C°)​	Mean (C°)​	Max (C°)​	Min (C°)​	Mean (C°)​
      Base	51.92​	23.09​	32.61	48.60	21.86	30.30
      Green​	50.53​	21.51​	31.09	47.11	22.09	29.97
      Cool​	51.92​	21.53​	32.52	47.20	22.18	30.07
      Material​	45.89​	20.17​	30.64	46.64	19.92	27.84
      Combination​	40.03​	19.86​	27.53	39.72	20.08	26.13
      Scenario​	UHC − 3			UHC − 4
      	Surface Temperature			Surface Temperature
      	Max (C°)​	Min (C°)​	Mean (C°)​	Max (C°)​	Min (C°)​	Mean (C°)​
      Base	51.36	21.91	31.79	52.53	21.47	35.79
      Green​	49.40	21.02	31.22	47.86	21.17	32.38
      Cool​	43.25	21.33	29.70	44.91	20.98	32.80
      Material​	42.21	20.13	27.74	44.52	20.71	32.27
      Combination​	41.58	20.49	28.09	40.39	20.81	30.20
      Scenario​	UHC − 5
      	Surface Temperature			Surface Temperature
      	Max (C°)​		Min (C°)​		Mean (C°)​
      Base	46.47		21.01		29.32
      Green​	45.71		21.36		29.30
      Cool​	38.81		21.09		28.17
      Material​	37.85		19.87		26.89
      Combination​	38.78		20.57		27.23

2. Building Level

   The variation in LST was observed on the building level that is the façade and roof surfaces. The following is the resulting temperature ranges for each of the simulated scenarios for the chosen areas of study and is as shown in Figure 13.

   Figure 13. Surface temperature variation across UHC-1, UHC-2, UHC-3, UHC-4 and UHC-5 for (a) base, (b) green, (c) cool, (d) material, (e) combination scenario at the building level.

   
   

   1. Base scenario:

      The base scenario constituted the study areas in their existing condition in terms of building material as well as built morphology. The surface temperature ranged from 34.78 °C – 57.99 °C, 37.69 °C – 63.54 °C, 33.19 °C – 56.72 °C, 39.23 °C – 61.39 °C and 34.87 °C – 59.40 °C for UHC 1, UHC 2, UHC 3, UHC 4 and UHC 5 respectively. The roof LST were considerably high for all the areas as the material used is concrete. The roof surface temperatures range from 47.68 °C – 55.41 °C, 49.18 °C – 57.80 °C, 48.88 °C –56.72 °C, 50.31 °C – 53.08 °C and 45.77 °C – 48.50 °C respectively. The façade surface temperature for UHC 1, 2 and 5 lies in a considerably low range (37.36 °C – 45.10 °C, 37.69 °C – 40.56 °C and 37.60 °C – 40.32 °C respectively), owing to the dense building layout that shades majority of them from direct sun. For UHC 3 and 4 the façades in the south and west, which were exposed to the direct sun exhibit the highest LSTs ranging between 51.49 °C – 56.72 °C and 55.85 °C – 61.39 °C respectively.

   2. Green Scenario:

      The greening scenario in UHC 1 and 2 was confined only to rooftop greening owing to the dense urban morphology, which limited the scope of greening on streets and facades. The surface temperature over the roof showed a decrease in temperature and ranged from 33.11 °C – 38.27 °C and 32.60 °C – 41.90 °C for UHC 1 and UHC 2 respectively. For UHC 3, 4 and 5, the low building density layout allowed façade and roadside greening in addition to rooftop greening. The surface temperature over the roof thus showed a decrease in LST and ranged from 28.13 °C – 38.22 °C, 29.20 °C –33.94 °C and 31.31 °C – 34.35 °C respectively. The temperature across the facades unlike the highly dense urban areas show a significant drop, as they were covered with vegetation (plants). The trees not only contributed to reducing the ground surface temperature, but shading also reduced the façade temperatures to some extent.

   3. Cool Scenario:

      The cool scenario comprised coating the concrete roof with acrylic-based high-reflectance paint for the building level. The roof temperatures ranged from 25.37 °C – 30.53 °C, 26.67 °C – 29.55 °C, 26.67 °C – 29.55 °C, 27.15 °C – 31.04 °C and 28.03 °C – 31.44 °C for UHC 1, UHC 2, UHC 3, UHC 4 and UHC 5 respectively. The facades exhibit the same LSTs as the base scenario since the same building material was used.

   4. Material Scenario:

      The sub-scenario involved replacing the building material with fly ash bricks which have insulation properties in addition to the cool roof strategy. The surface temperature showed a decrease in LST, compared to the base scenario, however, a very slight difference was observed in comparison to the cool scenario. The surface temperatures ranged from 26.80 °C – 30.70 °C, 26.01 °C – 33.54 °C, 26.09 °C – 29.16 °C, 27.11 °C – 31.21 °C and 28.02 °C – 35.22 °C for UHC 1, UHC 2, UHC 3, UHC 4 and UHC 5 respectively.

   5. Combination Scenario:

      The final scenario constituted an integration of all the mitigation strategies for each study area to understand their effect in combination as explained in the previous section. The surface temperatures across rooftops (25.37 °C – 27.95 °C and 26.88 °C – 30.84 °C) shows a decrease in the areas coated with reflective paint, whereas those covered with greenery (33.11 °C – 38.27 °C and 34.38 °C – 39.69 °C) show a slight rise compared to the green scenario. For UHC 3, 4 and 5 in addition to all the strategies mentioned, 50% of the facade surfaces were greened, while fly ash bricks were used for the remaining portion. The surface temperature across rooftops for UHC 3, 4 and 5 (23.36 °C – 29.40 °C, 23.93 °C – 27.99 °C and 24.22 °C – 32.33 °C respectively) show a decrease in areas coated with reflective paint whereas those covered with greenery (33.65 °C – 36.92 °C, 33.60 °C – 38.48 °C and 33.60 °C – 38.48 °C) also show a slight decrease in LSTs when compared to the green scenario.

      A point was selected within each study area to analyse the temperature difference for each scenario. A graph was plotted to compare the impact of each scenario on the surface temperature for that point. The following are maps showing the surface temperature distribution across each study area for every mitigation scenario on the building level and are illustrated in Table 14 and Figure 14.

      Figure 14. Graphs denoting variation in mean surface temperature at selected point for each scenario over UHC-1, UHC-2, UHC-3, UHC-4 and UHC-5 on 25.05.2021 at building level.

      

      Table 14. Maximum, minimum & mean surface temperature at selected point for each scenario over UHC-1, UHC-2, UHC-3, UHC-4 and UHC-5 on 25.05.2021 at building level.

      Scenario​	UHC - 1			UHC - 2
      	Surface Temperature			Surface Temperature
      	Max (C°)​	Min (C°)​	Mean (C°)​	Max (C°)​	Min (C°)​	Mean (C°)​
      Base	51.26	22.13	35.47	54.27	22.29	36.48
      Green​	33.78	17.94	26.93	35.22	17.92	27.51
      Cool​	31.63	20.65	26.91	29.72	18.76	24.54
      Material​	30.72	20.13	25.96	30.46	20.50	25.91
      Combination​	29.73	17.82	24.24	30.28	20.48	25.83
      	UHC − 3			UHC − 4
      	Surface Temperature			Surface Temperature
      Scenario​	Max (C°)​	Min (C°)​	Mean (C°)​	Max (C°)​	Min (C°)​	Mean (C°)​
      Base	52.30	22.42	35.50	54.96	22.39	36.79
      Green​	31.48	21.65	27.55	32.81	18.94	27.36
      Cool​	28.81	19.37	24.30	29.06	18.39	24.07
      Material​	28.76	19.11	24.18	29.13	18.02	24.03
      Combination​	28.66	19.12	24.15	28.30	20.26	24.97
      ​	UHC − 5
      	Surface Temperature			Surface Temperature
      Scenario​	Max (C°)​		Min (C°)​		Mean (C°)​
      Base	52.30		22.42		34.04
      Green​	31.48		21.65		27.26
      Cool​	28.81		19.37		25.06
      Material​	28.76		19.11		24.96
      Combination​	28.66		19.12		24.53

Discussion

Kolkata is one of the most largely agglomerated cities in India and encompassed a population of 1.5 million during the beginning of the nineteenth century, with approximately 2/3^rd residing in the core city and the rest in the sub-urban. Industrialization in the late 30s led to a significant rise in the population, owing to a large number of migrants settling in the city. At present, the projected population of Kolkata in 2023 is 6.2 million. To accommodate the need of this mammoth population, substantial alterations in the landscape have led to a rise in the surface temperatures, which further instigated the issue of heat waves. The study focused on developing and comprehending the notion of urban heat centres and subsequently simulating various mitigation scenarios to quantify their effectiveness at the neighbourhood and building scale.

• Base Scenario: The existing conditions were modelled in terms of the built environment and surroundings. The building materials were burned clay bricks and concrete that constituted the walls and roof, respectively whereas the roads were modelled considering dark asphalt. The base scenario exhibited the maximum surface temperatures for all study areas for both neighbourhood and building levels. This is attributed to the presence of low albedo materials like concrete and dark asphalt pavements that absorb most of the heat incident on it and emit it in the form of long-wave radiation resulting in higher surface temperatures (O’Malley et al. 2015). This is due to the building materials that exhibit high thermal masses and have high heat retention properties 76[]. The temperature distribution throughout the area was observed to be directly linked with the built morphology as higher ground surface temperatures were found in densely packed neighbourhoods (UHC 1 and UHC 2), compared to low-density areas. This is due to unobstructed wind in the low-density areas that allow faster heat dissipation radiated from the urban canyon (Mavrogianni et al. 2011). Similarly, there were milder temperatures in the areas shaded by buildings for all of the selected study areas. This was observed mostly in UHC 1 and 2, which were highly dense urban areas with extremely narrow streets and minimal building separation that allows mutual shading for most of the day. The roof temperatures were also observed to be the highest for the base scenario. The concrete has poor reflectance coupled with its ability to retain heat for a long period of time. The roofs with an increase in building heights were observed to show a slight decline in LSTs.

• Green Scenario: This scenario involved strategies such as roof greening, façade greening and roadside greening. The green scenario was modelled and simulated for priority locations based on the built layout and morphology. For the densest urban layouts (UHC 1 and UHC 2), only rooftop greening was adopted, which did not have a notable effect on the ground surface temperature, but showed a significant drop in the case of roof temperature owing to evapotranspiration. A slight decrease in ground surface temperature was observed in the case of areas characterized by ground and double-storied buildings, showing the effect of rooftop greening might even extend to the ground in the case of low-rise buildings (Peng and Jim 2013). For low-density areas (UHC 3, 4 and 5), façade, and roadside greening were adopted. While façade greening had a negligibly small effect on the ground surface temperature that too only around the periphery of buildings, roadside greening in the form of trees, on the other hand, shows the highest decline in surface temperature compared to any other forms of greenery.

• Cool Scenario: The cool scenario consisted of coating the rooftops with high-reflectance paint and replacing dark asphalt roads with reflective pavement. The cool scenario shows the most significant drop in surface temperatures compared to all scenarios for all the locations. The cool roofs do not affect the ground surface temperature but show a significant drop in LSTs compared to the base scenario in the case of roof temperatures. This is due to the high albedo (0.95) of reflective paint, which reflects most solar radiation falling on it thus exhibiting lower LSTs. Similarly, the application of reflective pavements also showed a maximum reduction in the ground surface temperatures.

• Material Scenario- The material scenario was modelled as a sub-scenario of the cool scenario and involved the replacement of burnt clay bricks with fly ash bricks as a building material. However, only a slight variation in surface temperature was noted for both neighbourhood and building levels. This may be because fly ash bricks have insulating properties owing to their low thermal conductivity that keeps the indoors cool during the summer and warm during winter. However, due to the poor reflectance, there is not much change in the surface temperatures.

• Combination Scenario- The combination scenario involved integrating all the mitigation strategies to reduce the number of impervious surfaces and increase the reflectivity of sealed surfaces. 50% of the rooftops were greened, while the rest were coated with reflective paint in addition to reflective pavements, facade greening and roadside greening across the urban canyon. The mitigation strategies within the combination scenario depicted consistent temperature variations as in the other scenario. The combination scenario where all of the applied strategies were incorporated shows the most significant drop in land surface temperatures on the neighbourhood level.

Conclusion

This research has highlighted the impact of unplanned and inadvertent urbanization on the climate, specifically in the form of urban heat island formation through the temporal land use and land surface temperature analysis of Kolkata City (KMC). The study sets forth to gather a deep understanding of areas of particular land use class requiring immediate interventions in terms of mitigation strategies to curb the rising LSTs by delineating heat zones in order of priority with the help of Land use and LST maps and spatial indices such as Normalized Urban Heat Island Index (NUHII). The main purpose of the study is to understand various mitigation scenarios and study their effect on land surface temperature for each of the selected study areas (UHC). The research additionally quantified the performance and effectiveness of the mitigation strategies individually and in combination for different urban categories on the neighbourhood and building level. The temporal land use and land surface temperature analysis identified four classes namely urban, vegetation, water and others for KMC and their relationship with LST. It was delineated that the urban impervious surfaces have increased at an exponential rate over the years. A positive correlation between LST and urban LU class signifies that an increase in built-up area at the cost of other land use classes critically affects the land surface temperatures. The presence of urban heat islands over the region was further signified with the help of NUHII and five priority locations (Urban Heat Centres, UHCs) was selected based on the heat zones that were most affected. These locations were potentially at the risk in terms of rising LST and urban sprawl across the city having different urban morphology, layout and type. Various mitigation strategies were simulated using which mitigation scenarios were designed namely. The base scenario constituted the areas in as-is condition with respect to the building material and surroundings to understand the performance of all the mitigation scenarios as compared to the business as usual. The green scenario incorporated strategies such as roof, façade and roadside greening. Cool scenario incorporated the coating of roofs and pavements with high reflectance paint. The material scenario replaced burnt clay bricks with fly ash bricks, in addition to reflective material. The combination scenario comprised the effect of all mitigation strategies put together for each of the selected locations. The designed scenarios and urban areas were modelled and simulated using Envi-met software, which simulated the temporal evolution of several thermodynamic parameters on a micro-scale range by creating a 3-D, non-hydrostatic model of interaction between building, atmosphere and vegetation. The findings demonstrated that the combination scenario yields the most substantial temperature reductions at both the neighbourhood and building levels, ranging from 9.51 °C to 11.82 °C and 2.09 °C to 5.59 °C, respectively. This was followed by the material, cool, and green scenarios in descending order of effectiveness, with reductions ranging from 6.78 °C to 9.43 °C, 8.56 °C to 12.72 °C, and 9.08 °C to 12.76 °C at the neighbourhood level, and 0.02 °C to 3.41 °C, 0.09 °C to 2.99 °C, and 1.97 °C to 4.05 °C at the building level. However, the impact mitigation strategies on the land surface temperature was found to be dependent on the built morphology, building heights and density. Each of the selected study locations was situated in different areas exhibiting equally different urban canyon layouts. It is thus pertinent to understand the building typology and climate of the region where these strategies are to be applied. The micro-climate of an area is also affected by anthropogenic activities such as pollution caused due to fossil fuel-based vehicles, greenhouse gas and CFC emissions etc., which were not taken into account for the simulations. The micro-climatic simulation model successfully analysed and contrasted the effectiveness of integrated and individual mitigation scenarios in enhancing the overall thermal comfort of the residents. This simulation framework could serve as a piece of information and assist various urban planning units, policymakers and stakeholders to develop policies for optimizing the composition and pattern of urban areas by incorporating suitable mitigation strategies, while considering local climatic conditions, to have a city with a thermally comfortable and healthy environment.

Practical limitations of mitigation strategies

The mitigation strategies of urban greening as well as cool materials have been extensively studied and proven to reduces surface temperatures considerably aiding in alleviating the rising urban heat island effect in cities. This study also investigated the effects of each of these strategies on land surface temperatures on different urban levels for areas across the city of Kolkata and found similar results. However, there are some practical limitations associated with such strategies that must be addressed. Green roofing, façade greening and roadside greening have high initial cost associated with installation which may act as a barrier in widespread adoption especially in areas where budgets are limited. Green infrastructure also requires regular maintenance without which vegetation can die off leading to aesthetic degradation and structural issues. Other pertinent issues related to urban greening also include special structural requirements and water management. Cool roof and pavement strategies that have proven to provide the most significant reduction in surface temperatures also present similar limitations concerning the cost of installation and maintenance. Additionally, some cool roof and pavement materials may have durability issues and a shorter lifespan compared to traditional materials. Cool roofs may not be as effective in colder climates as the reflective properties of the material can increase heating costs during the winter. Despite the limitations, the mitigation strategies discussed above remain valuable tools for improving urban environments and promoting sustainability. Addressing these limitations through innovative design, technological advancements, and supportive policies can help maximize the benefits of these interventions.

Author contributions

Bharath Haridas Aithal contributed to formulating strategy, data collection, technical inputs, funding for the work and paper writing as a major part of his contribution to this work. Nimish Gupta contributed to data collection, analysis of data and with major inputs in paper writing. Gargi Dwivedi was responsible for the running of simulation models, application, writing and analysis of the work. All authors have read and agreed to the submitted version of the manuscript.

Ethics declaration

All authors have read, understood, and have complied as applicable with the statement on ‘Ethical responsibilities of Authors’ as found in the Instructions for Authors and are aware that with minor exceptions, no changes can be made to authorship once the paper is submitted.

Data declaration

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Disclosure statement

The authors have no competing interests to declare that are relevant to the content of this article.

Additional information

Funding

The authors are thankful to the Indian Institute of Technology Kharagpur and Department of Science and Technology & Biotechnology West Bengal for the financial and Infrastructure support.