Variations in Surface Urban Heat Island and Urban Cool Island Intensity: A Review Across Major Climate Zones

Muhammad Sadiq KHAN ,Sami ULLAH ,CHEN Liding

(1.State Key Laboratory of Urban and Regional Ecology,Research Center for Eco-Environmental Sciences,Chinese Academy of Sciences,Beijing 100085,China;2.University of Chinese Academy of Sciences,Beijing 100049,China;3.Department of Forestry and Range Management,Kohsar University Muree,Punjab 47150,Pakistan)

Abstract: The climate has an impact on the urban thermal environment,and the magnitude of the surface urban heat island (SUHI) and urban cool island (UCI) vary across the world’s climatic zones.This literature review investigated: 1) the variations in the SUHI and UCI intensity under different climatic backgrounds,and 2) the effect of vegetation types,landscape composition,urban configuration,and water bodies on the SUHI.The SUHI had a higher intensity in tropical (Af (tropical rainy climate,K?ppen climate classification),Am (tropical monsoon climate),subtropical (Cfa,subtropical humid climate),and humid continental (Dwa,semi-humid and semi-arid monsoon climate)) climate zones.The magnitude of the UCI was low compared to the SUHI across the climate zones.The cool and dry Mediterranean (Cfb,temperate marine climate;Csb,temperate mediterranean climate;Cfa) and tropical climate (Af) areas had a higher cooling intensity.For cities with a desert climate (BWh,tropical desert climate),a reverse pattern was found.The difference in the SUHI in the night-time was greater than in the daytime for most cities across the climate zones.The extent of green space cooling was related to city size,the adjacent impervious surface,and the local climate.Additionally,the composition of urban landscape elements was more significant than their configuration for sustaining the urban thermal environment.Finally,we identified future research gaps for possible solutions in the context of sustainable urbanization in different climate zones.

Keywords: urban heat island intensity (SUHI);urban cool island intensity (UCI);day-night surface urban heat island (SUHI);climate zones;landscape composition and configuration;sustainable urbanization

1 Introduction

Approximately 60% of the world’s population currently lives in cities (United Nations,2014),and approximately 3% of the earth’s surface is covered by urban area(Bechtel et al.,2015).As a result,the surface urban heat island (SUHI) phenomenon has been observed in cities across various climate zones.Early studies of the SUHI can be classified into three broad categories.The first is a theoretical approach (Fan et al.,2016),which includes the urban-rural temperature difference (Jendryke et al.,2016),differences in day-night SUHI (Cai et al.,2017),seasonal variations (Dai et al.,2018),and urban parameterizations of numerical weather and climate models (Masson et al.,2014).The second is an experimental approach often defined as the SUHI,which has been considered alongside other bio-geophysical factors using remote sensing data (Schatz and Kucharik,2014)and hydroclimate factors (Hoffmann et al.,2012;Wiesner et al.,2014).The third approach is to conduct numerical research,which has been used to quantify the SUHI and simulate the thermal environment (Brazel et al.,2009;Li et al.,2016;Zhou et al.,2016).The phenomenon of SUHI intensity developed as a specific research field (Peng et al.,2012;Kikon et al.,2016;Gaur et al.,2018),which later extended to the regional scale to investigate the spatiotemporal pattern of the city-level SUHII (Li and Zhou,2019).Several studies have investigated the SUHI in different climate zones (Wienert and Kuttler,2005),and along different latitudes and longitudes (Tran et al.,2006).For example,the SUHI has been reported to occur on a large scale (regional level)in different climatic regions (Huang et al.,2017;Xiao et al.,2018).In the desert climate zone,SUHI intensity has been reported to be higher in central rather than in downtown areas (Lazzarini et al.,2015;Polydoros et al.,2018).Coastal cities tend to have lower SUHI intensities than inland cities (Tran et al.,2006;Barbieri et al.,2018),while in the Mediterranean climate zone,the SUHI fluctuates with the changing oceanic weather conditions (Fabrizi et al.,2010;Gaur et al.,2018).

The findings of these studies were based on factors such as the shape and size of the area,density of builtup land,population density,surface characteristics,e.g.,vegetation abundance,the spatial pattern of land use and land cover (LULC),precipitation and temperature,and day-night temperature difference (Du et al.,2016;Zhou et al.,2017;Imhoff et al.,2010).

The day-night temperature variation affects the SUHI intensity in China across different climate zones (Yang et al.,2017).Previous studies identified a higher nighttime SUHI in temperate regions (Sobrino et al.,2013;Tiangco et al.,2008;Wibowo and Salleh,2018),and a higher day-time SUHI in tropical regions (Tran et al.,2006).Land surface temperature (LST) is directly linked to the physical properties of surfaces (Chen et al.,2014;Sobrino and Jiménez-Mu?oz,2014;Wu et al.,2014),and its impacts on LST have been determined(Estoque et al.,2017;Wang et al.,2018).Ogashawara and da Silva Brum (2012) investigated the relationship between LST and built-up areas,water bodies,and vegetation density.Built-up areas enhance the urban temperature,while vegetation lowers the LST.For example,in the United States Foley et al.(2005) found that landuse land cover (LULC) changes were responsible for the elevated surface temperature.

Urban green parks provide thermal comfort (Oke et al.,1989),and have been reported to be 1.5 to 3.0°C cooler than the surrounding urban area (Peng et al.,2012;Zhao et al.,2014).This difference has been reported to be 3.0°C to 4.0°C higher than the surrounding urban areas in summer (Spronken-Smith and Oke,1998;Shashua-bar and Hoffman,2000;Shashua-Bar et al.,2010;Estoque et al.,2017).The extent of the cooling effect differed among different types of parks (Dimoudi et al.,2013),with well-distributed small parks producing the greatest cooling effect (Dimoudi and Nikolopoulou,2003),and large parks with trees having a greater cooling effect during the day (Bowler et al.,2010).A study in Japan investigated 90 different parks and concluded an optimal park size of 2 ha would provide the best cooling effect (Cao et al.,2010).Some studies have shown that the shape and size of green landscape patches,can greatly affect the SUHI (Wu et al.,2014;Sun and Chen,2017),and thus green infrastructure strategies can mitigate the adverse impact of the thermal environment.

In response to these adverse conditions,research has been conducted on SUHI/urban cool islands (UCI) at different scales,but uncertainties remain regarding the existence of SUHI,UCI,and day-night temperature effects on the SUHI across the global climate zones.To provide a more scientific explanation,this systematic review attempted to strengthen the case for the existence of the SUHI and UCI in different climate zones.More specifically,the effect of day-night temperature differences on SUHI variation under different climate backgrounds was reviewed at the city scale.The effect of water bodies,and landscape composition and configuration on SUHI intensity were investigated based on the literature.Based on the results reported in the selected articles,we identified future research needs and possible solutions in the context of sustainable urbanization.

2 Materials and Methods

The methodology used in this review was adopted from the previous studies (Bowler et al.,2010;Deilami et al.,2018),and detailed descriptions of the SUHI in different climate zones were considered (Roth Matthias,2007).

2.1 Paper selection

Published peer-reviewed research articles were selected for qualitative and quantitative content analyses (Hintz et al.,2018).Articles written in English from publishers including Scopus (https://www.scopus.com),Wiley (https://www.wiley.com),Springer (https://link.springer.com),Elsevier (https://www.elsevier.com),and MDPI(https://www.mdpi.com) were selected.Based on the research objectives,we found 143 relevant research papers using keywords such as SUHI intensity in cities,rural-urban SUHI,green space cooling effect and extent,day-night time SUHI,water/ blue space effect on SUHI,and effect of landscape pattern (composition and configuration) effect on surface urban thermal environment in the past three decades.The articles were screened for selection or rejection based on their research titles,abstracts,and keywords.The selection criteria were studies that investigated the intensity of SUHI,UCI,daynight variation in the SUHI,and the effect of water bodies,and landscape composition and configuration on the SUHI at the city scale in different climate zones.Finally,117 studies that investigated 168 cities were reviewed for analysis.The results of each study were categorized based on K?ppen’s classification from the Equator to the North Pole (Tong et al.,2017;Chakraborty and Lee,2019).Additionally,the relevant literature was classified based on the different sizes of green parks,their cooling values,and the distance from the edge of green spaces in different climatic zones,all of which affect the SUHI in urban areas.Recent review papers were also studied to identify possible solutions to the increasing SUHI intensity and to highlight future research gaps.

2.2 Quantitative analysis

The mean SUHI intensity in cities from different climate zones was obtained from the literature.The daynight variations in SUHI intensity and the cooling effect of green space cooling were investigated.The role of water bodies and landscape composition and configuration were discussed based on information from the literature.The detailed methodology of the review in the form of a flowchart is given in Fig.1.

Fig.1 Flow chart of the methodology of the review.SUHI,surface urban heat island;UHI,urban heat island

3 Results and Discussion

3.1 The number of studies across climatic zones

The variation in SUHI and UCI intensity and the daynight variations in SUHI intensity differed among the different geographical locations of cities in climate zones from the equator to the North Pole.At the city scale,we categorized individual cities and SUHI intensity based on the K?ppen classification (Chakraborty and Lee,2019) and found that most cities studied were located in the Cfa (subtropical humid climate,44),Dwa(semi-humid and semi-arid monsoon climate,29),Dfb(temperate humid climate,25),Csa (subfrigid marine climate,22),and Cfb (temperate marine climate,17) climate zones followed by Aw (tropical savanna climate,6),Csb (temperate mediterranean climate,5),and BWh(tropical desert climate,4).The geographical locations of the selected cities are shown in Fig.2.

Fig.2 Geographical locations of surface urban heat island(UHI) studies across climate zones.Tropical climate: Af,tropical rainy climate;Am,tropical monsoon climate;Aw,tropical savanna climate;Subtropical climate: Cfa,subtropical humid climate;Cfb,temperate marine climate;Desert climate: Bsh,tropical semi-arid climate;Bwh,tropical desert climate;BSk,cool semiarid climate;Mediterranean climate: Csa,subfrigid marine climate;Csb,temperate mediterranean climate;temperate climate:Dfb,temperate humid climate;Dfc,subfrigid humid climate;Dwa,semi-humid and semi-arid monsoon climate

Most of the studies investigated the SUHI in cities with a subtropical climate,while cities in tropical and desert climate zones were the least frequently studied.Table 1 lists the studies selected for this review that investigated SUHI intensity at the city level.

Table 1 Studies that investigated the surface urban heat island (SUHI) in different climatic zones

A review conducted by Roth Matthias (2007) found that 20% of published studies were conducted in cities in the Cfa climate zone based on the K?ppen climate classification.Among the studies considered in the present review 27% were in the Cfa climate zone,while 18% were in Dwa,15% in Dfb,and 14% in Csa due to the more focused research in recent decades.Among a previous review of 223 cities between latitudes 43°S and 65°N (Wienert and Kuttler,2005),the SUHI intensity was investigated at different latitudes with 50% of studies in temperate climates,17% in the tropics,and 35% between 20°N and 40°S.However,our review found that most studies were conducted in the Cfa (27%,44)climate zone,followed by the Dwa (17%,29),Dfb(21%,25),and Csa (18%,22) climate zones.Our findings differed from previous review studies due to advances in this field of study.The other reason for the differences were the objectives and criteria of the present study.In the last decade,several papers were published that focused on the SUHI in mid-latitude cities,which was due to the growing population and rapid development in large regional cities.

3.2 Variation in SUHI intensity across climatic zones

At the city scale,the SUHI intensity varied among 95(42 studies) major urban climates.We found the highest mean SUHI intensities in the Cfa,Cfb,and Dwa climate zones,a moderate intensity in the Csa zone,and a low intensity in the Dfa zone.The moderate SUHI intensities were due to the changing oceanic effects in the Mediterranean climate zone.Interestingly,an inverse SUHI intensity was observed in the BWh climate zone in summer daytime,with a negative SUHI intensity due to there being little vegetation in the surrounding desert area.More details of the mean SUHI intensity in different climate zones are given Table 1.

In summary,we found variation in the SUHI intensity for every～ 10.0° increase along a latitudinal gradient from the Af to Dwa climate zones except in the desert (BWh) climate zone.In the Csa climate zone,the SUHI was～ 2.0°C lower,which could be due to the influence of the combined oceanic and seasonal effects.The intensity was lower in the Dfb and Dfc climate zones due to the altitudinal effect on the urban temperature.In the desert climate (BWh and Bsk),13 cities had a higher mean SUHI in rural areas than urban areas due to the higher evapotranspiration from vegetation in urban areas.The vegetation abundance was greater in cities than in the surrounding hot desert areas.The detailed variation in SUHI intensity across the climate zones is shown in Fig.3.The earlier research focused on several cities in a single case study.For example,Peng et al.,(2012) quantified the mean SUHI of (1.5 ± 1.2)°C in 419 cities.Yang et al.(2017) determined the magnitude of the SUHI to be (1.2 ± 0.3)°C in major biomes (tropical to alpine) and calculated a mean value of 1.4°C to 0.6°C in the dry climate of the Qinghai-Tibet Plateau,in China.In contrast,Shastri et al.(2017) calculated a negative SUHI of (-1.5 ± 1.2)°C in 48 cities in India.These variations in intensity could be due to the difference in climatic conditions and different urban landscapes across the climate zones.

Fig.3 Summary statistics of the surface urban heat island(SUHI) intensity in cites across climatic zones.Af,tropical rainy climate;Am,tropical monsoon climate;Aw,tropical savanna climate;Bwh,tropical desert climate;BSk,cool semi-arid climate;Cfa,subtropical humid climate;Csa,subfrigid marine climate;Cfb,temperate marine climate;Csb,temperate mediterranean climate;Dfb,temperate humid climate;Dfc,subfrigid humid climate;Dwa,semi-humid and semi-arid monsoon climate

A review by Roth (2007) found variations in monthly SUHI intensity of 5.7°C,～ 3.0°C,and～ 3.5°C in the Af,Bush,and Aw climate zones,respectively.Our results were slightly higher in the tropical and desert (Bwh) climate zones.This may be due to the increased urbanization and anthropogenic activities in recent years as well as more focused research in these climate zones.Similarly,the dependency of SUHI intensity on latitude was investigated by Wienert and Kuttler (2005),resulting in a lower mean SUHI in the tropics (～ 4.0°C),and higher values (～ 6.1°C) towards the temperate climate regions,which confirmed our results.The reasons for the higher SUHI in the higher latitudes were the release of the maximum anthropogenic heat,high heat transfer capacity of the surfaces,and rapid urbanization that generates differences in surface energy balance.As shown in Fig.3,a variation in the intensity of the SUHI was apparent from the tropical to humid continental climate zones.

3.3 Variation of UCI intensity across the climate zones

The UCI intensity varied among 36 (28 studies) major urban climates at the city scale.The cooling contribution of green space was related to many factors,particularly its size and distribution in the urban landscape.From the literature,we found that the extent and intensity of cooling varied among the green spaces in different climate zones.For example,in the Af climate zone,a park of 36 ha extended cooling of 1.3°C to a distance of 100-300 m beyond the green space,while a 3 ha park maintained cooling of 0.8°C for up to 100 m beyond the green space in the Cfa cooling zone.Cooling of 1.7°C was observed at a distance of 104 m from the edge of parks (Ca et al.,1998;Yu et al.,2018a).Vaz Monteiro et al.(2016) examined two parks of 0.89 and 156 ha and found that their cooling intensity and extent was 2.5°C(150 m),and 5.9°C (1100 m) in the Cwb climate zone.Jauregui (1990) investigated a large park in the Csb climate zone of 500 ha,for which the cooling effect of 3°C extended to 2000 m.In the Dfb climate,Vaz Monteiro et al.(2016) calculated a cooling intensity of 0.6°C that extended 150 m from a small park of 5 ha.Further details of the green space in different cities and its cooling intensity and extent are given in Table 2.

Table 2 Urban cooling intensity (UCI),park size,and extent across climate zones

In summary,we found the highest mean intensity of UCI in the Cfb,Csa,and Csb climate zones,with moderate cooling in the Cfb zone.The details of the mean urban cooling intensity are shown in Fig.4.The cooling efficiency of different green spaces and the extent of their cooling effect varied across the climate zones.Different cooling intensities were found in different urban areas.The cooling contribution of green parks was found to differ in several climate zones.For example,for large parks we found a higher cooling intensity in the Cfb and Csb climate zones and a moderate intensity in the Cfa and Csa climate zones.A low cooling contribution was observed in the Dfb and Dwa climate zones.The reasons for the different cooling contributions may be due to variations in size,the proportion of impervious surfaces in the park,and proximity to adjacent landscapes.

Fig.4 Urban cooling intensity (UCI) intensity of cities across the climate zones.Af,tropical rainy climate;Am,tropical monsoon climate;Aw,tropical savanna climate;Bwh,tropical desert climate;BSk,cool semi-arid climate;Cfa,subtropical humid climate;Csa,subfrigid marine climate;Cfb,temperate marine climate;Csb,temperate mediterranean climate;Dfb,temperate humid climate;Dfc,subfrigid humid climate;Dwa,semi-humid and semi-arid monsoon climate

The cooling efficiency of different green spaces and their extent varied across the climate zones.Different cooling intensities were found in different urban areas.The cooling contribution of green parks was found to differ among climate zones.For example,in the large parks,we found a higher cooling intensity in the Cfb and Csb climate zones and a moderate intensity in the Csb and Dfb climate zones.A low cooling contribution was observed in the Cfa and Af climate zones.This was likely due to the size of the green spaces in these regions and the extent of their cooling,and was also related to the urban landscapes adjacent to parks which tended to lack tall buildings.

A review by Bowler et al.(2010) quantified a mean cooling efficiency of 0.98°C for several parks.Similarly,Hintz et al.(2018) reviewed 17 different studies in various climate zones and highlighted the importance of green spaces in breaking up impervious surfaces which confirmed our results.Our results indicated the importance of urban green parks to ameliorate the ever-increasing urban thermal environment in urban areas.The appropriate size and distribution of green parks might be helpful for sustainable urbanization and will also benefit urban residents.The results also indicated the different cooling contributions of different types of parks in different climate zones.

Vegetation structures such as woodlots or even asingle evergreen or deciduous tree can significantly mitigate the SUHI,and a varied composition (e.g.,trees,shrubs,and grasses) better enables vegetation to impact the urban climate.Based on the results of previous studies,the cooling effect of trees was found to be higher than that of shrubs,and grasses (trees >shrubs >grasses).For example,Onishi et al.(2010) found that the cooling intensity due to tree cover was 2.49°C and 2.36°C in a park with a 30%∶70% tree/grass cover ratio.In the Dwa climate zone,several studies found that trees provided a higher cooling intensity than grasses(Chen et al.,2014;Dobrovolny,2013;Amani-Beni et al.,2018).For example,Scott et al.(1999) reported cooling intensity of 1.4°C,and 0.7°C for trees and grasses,respectively.Similarly,Coutts et al.(2013) reported a cooling intensity of 2.0°C for trees and 0.9°C for grasses in the urban landscape.Sun and Chen (2012)reported a cooling intensity of 7.0°C for trees and 4°C for grasses in a complex urban landscape.In a study in Japan,Park et al.(2017) found that a single coniferous tree contributed to a 1.9°C (max) and 0.8°C (mean)cooling at the roadside,while Gulyás et al.(2006) found a tree cooling effect of 2.1°C (max) and 1.6°C (mean) in Hungary.Lin and Lin (2010) found that many small lot of deciduous trees in Taipei City of China contributed to a cooling of 2.5°C (max) and 1.4°C (mean) in summer daytime.Shashua-Bar (2010) found a combined summerwinter and day-night cooling effect of 3.1°C (max) and 1.6°C (mean) for both coniferous and deciduous trees in Tel-Aviv (Israel).Based on the literature,the cooling intensity of both coniferous and deciduous vegetation differed,with combinations found to be more significant.The reason for the cooling effect provided by trees is primarily the canopy cover and shade that they provide,as well as the high transpiration rate in an urban environment.The higher cooling intensity of trees compared to shrubs and grasses in different urban climates was also reported by Bowler et al.(2010).

3.4 Effect of day-night temperature variations on SUHI intensity across climate zones

Diurnal temperature variation has a large effect on SUHI intensity across climate zones.In the daytime,the grey and impervious urban surfaces absorb sensible heat and then release it slowly at nighttime.This review investigated 27 studies that considered the diurnal temperature variation (Table 3) and found a higher SUHI intensity in the daytime than nighttime in the Af,Aw,and Dwa climate zones (Group I),while a lower intensity was found in the daytime than in the nighttime for the Bsk,Csa,and BWh climate zones (Group II).In the Cfa,Cfb,Dfb,and Bsh climate zones,the variation of SUHI intensity was less than 10% (Group III).The variation in diurnal SUHI intensity was found to be higher in the tropical rainforest (Af) and tropical savanna (Aw)climate zones.

Table 3 Differences in the day-night surface urban heat island (SUHI) intensity in different climate zones

A greater diurnal variation in SUHI was observed in the tropical climate zone due to the consistent environmental conditions over a 24 h period,with minor differences in surface temperature at night-time but high daytime temperature variations.The high intensity of incoming solar radiation in the tropics cause the impervious surfaces of tropical cities to rapidly absorb and release heat in the daytime in urban areas.In the tropical and subtropical,subtropical steppe,and Mediterranean climate zone,the SUHI intensity at the night-time is higher because there are many densely populated and congested cities,with the maximum anthropogenic heat release occurring in the subtropical region.The changing climatic conditions in the Mediterranean climate zone result in cities having large temperature variations,with a high nighttime SUHI.Some studies have reported very low variation (>10%) in the UHI intensity inthe Cfa,Cfb,Dfb,and Bsh climate zones.

In all climate zones,there was a substantial variation in the SUHI intensity between the day and night time(Fig.5).Dousset et al.(2011) reported a higher nighttime SUHI intensity in the Cfa climate zone,while Tran et al.(2006) found a higher daytime SUHI intensity in the Cfa climate zone,indicating a large degree of uncertainty.Out of the 27 studies reviewed here,we found 19 that reported a higher night-time SUHI intensity,with the others reported a higher daytime SUHI intensity.The core reasons given for this were the slow release of trapped heat by impervious surfaces,tall buildings,and large urban population,which all enhanced the nighttime SUHI intensity.The canyon effect of high-rise buildings alongside main roads may also have had an impact.

Fig.5 Day-night surface urban heat island (SUHI) intensity across the different climate zones.Af,tropical rainy climate;Am,tropical monsoon climate;Aw,tropical savanna climate;Bwh,tropical desert climate;BSk,cool semi-arid climate;Cfa,subtropical humid climate;Csa,subfrigid marine climate;Cfb,temperate marine climate;Csb,temperate mediterranean climate;Dfb,temperate humid climate;Dfc,subfrigid humid climate;Dwa,semi-humid and semi-arid monsoon climate

3.5 The effects of water bodies on SUHI intensity

Water bodies mitigate the SUHI intensity in the urban climate and their size,shape,distribution,and proximity to buildings,provide different cooling intensities.For example,Hathway and Sharples (2012) investigated a cooling of 2°C in the centre and 1.5°C in the riparian zone of a river that extended to a distance of 30 m.Similarly,other studies (Sun and Chen,2012;Theeuwes et al.,2013) investigated the association between the size,shape,and width of water bodies within the surrounding urban landscape.Li and Yu (2014)reported a cooling of 2.6°C due to the presence of a lake in a composite landscape,and a small pond (4 m × 4 m)contributed a cooling of 1.0°C that extended 30 m in an urban landscape.Similarly,Nishimura et al.(1998) reported a cooling effect of 3.0°C that extended 35 m from the leeward side of an urban fountain.Murakawa et al.(1991) found a cooling of 5.0°C at distances up to 80 m from the centre of a water body.

Based on the literature,it was apparent that water bodies,their distribution,and proximity to buildings may decrease the SUHI in urban areas (Du et al.,2016).Even an artificial water fountain could reduce the surrounding LST.Sun and Chen (2012) found that it was more effective to mitigate SUHI with water bodies in combination with green spaces in the sophisticated urban landscape of Beijing.These findings provide evidence that blue-green spaces are important landscape elements that might help to lower the LST.In the urban core area where there is limited space for vegetation patches this may not be possible and artificial fountains might help to cool the urban living environment.

3.6 The effect of landscape composition and configuration on SUHI intensity

Landscape composition and configuration may affect the SUHI intensity,with landscape composition typically making a larger contribution to mitigating the SUHI than landscape configuration.Landscape metrics,such as the percent land (PLAND),can be used to explain the negative relationships between LST and both water bodies and forest,while built-up areas and barren land were found to be positively correlated with LST in several urban climates.A significant relationship was found between LST and landscape metrics such as the patch density (PD),edge density (ED),and landscape shape index (LSI) in early studies.For example,Wu et al.(2014) investigated a significant correlation between LST and different land-use types (water bodies,forest,shrubs,cropland,grassland/agriculture,barren land,and built-up area) using landscape metrics in the city of Wuhan (China).A highly significant positive correlation was observed between LST and the built-up area,i.e.PLAND,PD,ED,and LSI,while the values for water bodies,forestland,and cropland were negatively correlated for PLAND,ED,and LSI,respectively.Similarly,Li et al.(2011) found a significant positive correlation between LST and landscape metrics such as PLAND,PD,ED,and LSI,for built-up areas,water bodies,and traffic network and a negative correlation between LST and PLAND,PD,ED,and LSI for forestland in Shanghai metropolitan city.In the densely populated city of Beijing,Sun and Chen (2012) examined the positive correlation between LST and land use by applying the landscape metrics of PLAND and core area percentage of landscape,CLAND (core area landscape metric) for water bodies,roads,and agricultural land.Naeem et al.(2018) quantified the composition and configuration of the green spaces in Beijing and Islamabad and found a negative correlation between LST and both PLAND,and ED for vegetation,and a positive correlation with both PD and LSI for built-up areas,respectively.For a better understanding,Chen et al.(2014) investigated the relationship between LST and landscape metrics before and after a cluster analysis to improve the prediction of LST in the urban landscape.

This review addressed the relationship between SUHI intensity and urban landscape composition and configuration.From the literature,we highlighted the significance of landscape composition and its changes for the relationships between mean LST and both impervious surfaces (positive) and green spaces (negative).For example,Estoque et al.(2017) found a mean difference in LST of 3°C between impervious surfaces and green spaces and LST was correlated positively with the area of impervious surfaces and negative with the area of green spaces in the complex cities of Southeast Asia.The positive correlation of LST with impervious surfaces was due to heat-absorbing materials (Imhoff et al.,2010),and the negative correlation with green spaces was due to evapotranspiration and the shading effect of trees that cool the surrounding environment (Li et al.,2012).Several studies found a negative correlation between LST and the high PLAND of forest and water bodies (Vaz Monteiro et al.,2016),while LST was positively correlated with impervious surfaces and barren land (Myint et al.,2013;Huang et al.,2017;Yang et al.,2017).It was further observed that landscape composition (PLAND) was more effective than landscape configuration for sustainable urbanization (Chen et al.,2014).A balanced proportion of vegetation and water bodies to impervious surfaces and barren land is important for the mitigation of the SUHI in an urban area.A recent review by Deilami et al.(2018) found that an impervious surface area was the key variable explaining the increase in SUHI intensity,while vegetation cover decreased the SUHI intensity.Similarly,Estoque et al.(2017) examined the effect of the shape,size,complexity,and proportion of green spaces,as well as the ratio of green spaces to impervious surfaces,and suggested a high ratio of green space to impervious surfaces would reduce the SUHI intensity in megacities.Sun and Chen(2017) found a significant correlation between land use types and landscape metrics such as PLAND,largest patch index (LPI) mean patch size (MPS),area-weighted means shape index (AWMSI),an interspersion and juxtaposition (IJI).

It was clear from the selected literature that green spaces can lower the LST,which had a negative correlation with landscape composition and configuration in the urban landscape.In built-up areas and barren land,there was a positive correlation between landscape metrics and LST.It was further observed that landscape composition metrics (e.g.,PLAND) were more important than configuration metrics for lowering the LST(Chen et al.,2014).Due to the limited amount of literature on LST and landscape metrics,further research is needed to explore the buffer contributions of adjacent land use types in the complex urban landscape,and to develop a model for a better understanding of the contribution of each type of land cover to the urban temperature.

4 Conclusions and Outlook

This study comprehensively investigated the SUHI and UCI intensity in major climatic zones based on the published literature.A higher mean SUHI intensity was found in cities located in the tropical,subtropical,and humid continental climate zones,while the maximum cooling of green spaces was noticed in the tropical and Mediterranean climate zones.This suggests that the SUHI and UCI intensity may be explained by climate zones.Although several cities in different regions have adopted green initiatives,the warming intensity in cities was about twice that of the surrounding area.

The cooling intensity was always lower than the warming intensity in all climate zones due to the limited vegetation cover in the urban landscape.For example,in tropical (Af and As) cities,more vegetation is needed to cool the hot urban thermal environment.Surprisingly,the temperature difference in cities located in the desert climate zones resulted in a negative SUHI on hot summer days.The temperature variation contributed to a higher SUHI intensity in the daytime in these climates zones.An appropriate green space distribution in a heterogeneous landscape may be a solution to the increasing SUHI,with green space composition being more important than urban configuration.Based on the above conclusions,the following future research gaps were identified.

1) Intensive studies of SUHI intensity are required in complex urban landscapes,such as Beijing,Chicago,New York,and Shanghai,to better understand and manage the urban thermal environment worldwide.

2) Research is needed to understand how the balance between green spaces and impervious surfaces in heterogeneous landscapes impacts on urban cooling and heating across the world’s climate zones.

3) An appropriate urban landscape design is necessary in all climate zones across the world,i.e.,the optimal landscape composition and configuration for different climate zones.More specifically,what is the optimal forest cover to mitigate the SUHI intensity in different urban landscapes in different climate zones?What combination of green-blue space would be appropriate in different climate zones for different landscapes?

4) Further research is required to determine how the SUHI and UCI intensity are related to the climate.For example,the optimal green space cover in cities will be different in a tropical climate from a temperate climate because of the different climatic conditions.

There are some limitations to this review that should be addressed to avoid ambiguity in the interpretation of the results,particularly in discussing the magnitude of the SUHI and UCI across the climate zones.First,there were large discrepancies among the early studies and there is considerable uncertainty regarding their results.This was largely due to the different data used to classify urban and rural sites,and the different methods used for the retrieval of temperature.Additionally,the SUHI intensity is related to many other factors in urban areas,such as meteorological and geographical factors.Despite these issues,this review attempted to explain the magnitude of SUHI and UCI intensity based on the literature for different climate zones.Second,there were different methodological approaches to the estimation of SUHI and UCI across the different climate zones.Third,we only collected data for the summer season across the different climate zones.Fourth,the estimation of cooling intensity and extent of the cooling due to urban parks/green spaces was based on data recorded by different instruments or satellite retrievals,which may also have an impact on the cooling values.Finally,there was a lack of literature on SUHI/ UCI intensity in the climatic regions of South America and southern Africa.

Conflict of Interest

The authors declare that the manuscript has no relevant financial or non-financial conflict of interests to disclose related to the content of article for publication.

Author Contributions

Muhammad Sadiq KHAN: conceptualization,methodology,software,validation,investigation,resources,data curation,and writing original draft;Sami ULLAH:formal analysis,writing review &editing,project administration &software;Yuelin Li: review and editing,funding acquisition,supervision.