Research progress on the synergies between heat waves and canopy urban heat island and their driving factors

Under the background of global warming and accelerating urbanization, the interaction between heat waves (HWs) and canopy urban heat island (CUHI) has become one of the focuses in the ﬁ eld of global climate change research. This paper comprehensively reviewed and summarized the research process on the synergies of HWs and CUHI and their in ﬂ uencing mechanism. The coupling effect between HWs and CUHI remains debated, which may be related to the use of different standards to de ﬁ ne heat wave events. The spatiotemporal differences in the synergies between HWs and CUHI was also in ﬂ uenced by climate background and local circulation. For instance, scholars have reached different conclusions regarding the stronger synergistic effect between daytime and nighttime in cities with different climate backgrounds. In addition, the modulation of urban morphological structure to the synergies between HWs and CUHI cannot be ignored. In the future, it is necessary to adopt different de ﬁ nitions of HWs to systematically study the formation mechanism of urban excess warming from different climatic backgrounds, local circulation conditions, and urban morphologies.


Introduction
HWs sually refer to the situation that the daily maximum temperature reaches or exceeds a certain threshold for several days (Tan et al., 2010;Perkins, 2015;Sun et al., 2017;Su and Dong, 2019;Zong et al., 2021).HWs are one of the important causes of diseases such as cardiovascular, respiratory, and digestive systems in the human body, and in extreme cases, they can directly lead to death (Fouillet et al., 2006;Zhang et al., 2018;He et al., 2021).Even in the period of global warming hiatus, the frequency and duration of HWs show an increasing trend, which poses a huge challenge to the health of people and the stable development of society (IPCC, 2021).In addition, urbanization has become an important driving force of regional and global warming (Seto et al., 2012;Shi et al., 2021).Canopy urban heat island is a phenomenon that the canopy urban temperature is higher than that in suburban and rural areas due to the change in the nature of the urban underlying surface and anthropogenic heat emission (Li et al., 2015;Shi et al., 2015;Yang et al., 2023).CUHI effect has become one of the most significant characteristics of urban climate, which has an important impact on human health and socio-economic development (Gershunov et al., 2009;Ren et al., 2011;Luo and Lau, 2019;Yang et al., 2019).
The interaction between climate change and urbanization and its possible synergistic effects have become the focus of global climate change research (Seto et al., 2012).Urbanization has an impact on the indicators of HWs (Zong et al., 2021), and changes the distribution and intensity of high temperature fields in summer (Founda et al., 2015;Xue et al., 2023).The HWs and CUHI may also have a superimposed effect, even more than the sum of the effects of the two on the urban area (Basara et al., 2010).The frequency of HWs in urban stations increases significantly, the increasing degree of the number of HWs and the duration of HWs is also larger than that in non-urban stations (Mishra et al., 2015;Khan et al., 2020;Mughal et al., 2020;Zinzi et al., 2020).However, some scholars have observed completely different results (Basara et al., 2008;Basara et al., 2010;Ramamurthy and Bou-Zeid, 2017;Chew et al., 2020).Scholar have found that under the HWs condition, the CUHI intensity (CUHII) at nighttime in Melbourne and Adelaide is stronger than that at daytime, while the CUHII at nighttime in Perth is weaker than that at daytime (Rogers et al., 2019).Currently, there is no consensus on whether CUHI will intensify or weaken during heat waves periods.
With the acceleration of urbanization and aggregation of population, the CUHI effect is significant (Liu et al., 2007;Zheng et al., 2018;Yang et al., 2020), which intensifies the occurrence of regional excess warming (Li et al., 2015;Yoon et al., 2018;Zong et al., 2021) and seriously affects the health of residents (Gao et al., 2015;Jiang et al., 2019).The research on HWs and CUHI has been carried out for decades.Especially in recent years, the research results are very rich.This paper comprehensively reviewed the progress on the synergies of HWs and CUHI and their driving factors, in order to provide important theoretical basis and technical support for the construction of high temperature monitoring, prediction, and early warning systems, the improvement of living environment and urban planning management.
2 Difference in the research on the synergies between HWs and CUHI Over the past few decades, major cities around the world have witnessed rapid expansion of construction land.The CUHI effect caused by urbanization is one of the typical urban climate characteristics (Ren et al., 2015;Yang et al., 2019).In parallel, climate change has led to a significant increase in the frequency of high temperature and heat wave events in most parts of the world (Su and Dong, 2019;Yang et al., 2023).The interaction between CUHI and regional extreme high temperatures cannot be overlooked.
Many scholars have carried out research on the synergies of HWs and CUHI.Some studies have demonstrated a synergistic effect between HWs and CUHI, further compounding the heat risk in urban (Jiang et al., 2019;Yang et al., 2023).Compared with the non-heat waves periods, the maximum CUHII of Beijing at nighttime during the heat waves periods increased by 1.26 °C (Jiang et al., 2019), the average CUHII in Shanghai during the heat waves periods increased by 0.9 °C (Yang et al., 2023), and the maximum CUHII in Seoul during the heat waves periods increased by 4.5 °C (Ngarambe et al., 2020).During the heat waves periods, the average CUHII of 50 cities in the United States was 0.4 °C-0.6 °C higher than that during the non-heat waves periods (Zhao et al., 2018).During the heat waves periods in the northeast of the United States in 2016, the CUHII of New York City, Washington, and Baltimore was amplified by 1 °C-2 °C (Ramamurthy and Bou-Zeid, 2017).During the heat waves periods from 2013 to 2018, the CUHII of Shanghai increased by about 0.8 °C (Ao et al., 2019).Also during the heat waves periods from 2013 to 2018, Jiang et al. (2019) ) found that the CUHII of Shanghai increased by 1.0 °C, which was close to the research results of Ao et al. (2019), and 1.2 °C and 0.9 °C were observed in Beijing and Guangzhou.Zong et al. (2021) concluded that the increase of CUHII usually reached a minimum at 0900 and 1800 LST, and the peak often occurred at noon or at midnight (as shown in Figure 1).Some studies showed that the synergies between HWs and CUHI are not obvious, and even the CUHII may decrease during heat waves periods (Richard et al., 2021).Rogers et al. (2019) found in their study on CUHI during heat waves periods and non-heat waves periods in Australian that the CUHI of nighttime in Perth weakened during heat waves periods.Compared with the non-heat waves periods, there was no CUHI enhancement in Philadelphia during the heat waves periods (Ramamurthy and Bou-Zeid, 2017).Scott et al. ( 2018) took 15 HWs in Baltimore from 2000 to 2015 as a composite event and found that there were no synergies between HWs and CUHI.After expanding the research scope to 54 cities, even the average CUHII tended to decrease during the heat waves periods.Richard et al. (2021) discovered that the maximum CUHII in Dijon (France) frequently occurs prior to or within the first few days of HWs, and subsequently diminishes during heat waves periods (As shown in Figure 2).Using ground observation data, Chew et al. (2020) discovered that, in contrast to non-heatwave periods, the CUHII does not intensify during heat waves periods.Specifically, the CUHII remains approximately 2.5 °C during both heat waves periods and non-heat waves periods, and the results from the WRF simulation are consistent with the observation results.The above findings suggest that there are significant differences in the coupling effect between CUHI and HWs across different regions, necessitating in-depth research that takes into account specific research methods, the geographical location of the research subjects, and other influencing factors.

The influence of definition of HWs on the synergies of HWs and CUHI
HWs usually refer to sustained high-temperature weather processes.Most studies only utilize a single variable to identify heat waves, such as the highest daily temperature occurring during the afternoon or the lowest daily temperature occurring in the early morning hours (Bador et al., 2017;You et al., 2017;Oswald, 2018;Ren et al., 2020).Due to the diverse geographical environments, there are differences in the definition of HWs among different countries and regions.For instance, the World Meteorological Organization (WMO) recommends a heat wave event with a daily maximum temperature exceeding 32 °C and lasting for more than 3 days.When the daytime heat wave index exceeds 40.5 °C for three consecutive hours or is expected to exceed 46.5 °C at any given time, the National Oceanic and Atmospheric Administration (NOAA) of the United States issues a high temperature warning.The Royal Netherlands Meteorological Institute (KNMI) defines a heatwave event as a daily maximum temperature exceeding 25 °C and lasting for more than 5 days, with at least 3 days of maximum temperature exceeding 30 °C.The heatwave standard set by the China Meteorological Administration (CMA) is that the maximum temperature exceeds 35 °C for three consecutive days.Fenner et al. (2019a) noted that if hot weather episodes are defined as the ten percent hottest days or nights during May-September, and identified based on daytime conditions or nighttime conditions at inner-city sites, the nighttime CUHI is exacerbated.However, if hot weather episodes are identified based on nighttime conditions at rural sites, the night-time CUHII is reduced.Fenner et al. (2019b) provided an overview of all identified HWs by any of the ten definitions, illustrating the number of definitions that identify a certain calendar day as a HW day.As shown in Figure 3, different methods to identify and define the HWs could affect the occurrence time, duration, and frequency of HWs to a certain extent.
Some scholars have attempted to use bivariate identification of heat waves, such as a method that considers both the maximum and minimum daily temperatures to be above a threshold for multiple consecutive days (Kuglitsch et al., 2010;Chen and Li, 2017;Freychet et al., 2017).Additionally, there are studies that require the occurrence of extreme daytime high temperatures and extreme nighttime high temperatures to follow a specific order (Cowan et al., 2014;Oswald and Rood, 2014).As exemplified by Cowan et al. (2014), who specified that extreme daytime high temperatures must persist continuously for more than 3 days, with extreme nighttime high temperatures occurring on the second and third days.An and Zuo (2021) conducted a study on the HWs observed by meteorological stations in China.Taking Beijing as an example, they found that some HWs only involve extreme daytime high temperatures (Figure 4A), while others are limited to extreme nighttime high temperatures (Figure 4B).Additionally, there are instances where both daytime and nighttime high temperatures are part of the HWs.Furthermore, there is a specific temporal sequence in the occurrence of daytime and nighttime high temperatures (Figures 4C, D).Currently, there is no unified standard for the definition of heat waves, which can introduce some uncertainty in the attempt to gain a deeper understanding of the interaction between HWs and CUHI.

The influence of climate background and local circulation on the synergies between HWs and CUHI
In addition to the definition of HWs, the difference in the synergies between HWs and CUHI is also related to the climate background and local circulation (meteorological elements in the boundary layer such as solar radiation, wind speed, wind direction, humidity).HWs are usually caused by synoptic scale high-pressure systems (Matsumura et al., 2015), which severely inhibit the formation of convective clouds (He et al., 2015).Therefore, the synoptic background during heat waves periods is usually breezy and cloudless.Although the synergies of HWs and CUHI has been confirmed in different regions, different results have also been observed in cities under different climate backgrounds (Xie and Zhou, 2023).Liao et al. (2018) studied the HWs in China from 1961 to 2014 and believed that the increasing trends of HWs in urban areas are greater than those in rural areas in wet climates, suggesting a positive contribution of urbanization to HW trends.However, the urbanization contribution to HW trends is smaller and even negative in wet climates.Zhao et al. (2018) found that the synergistic effect of HWs and CUHI is more significant in temperate climate regions of the United States, but not in arid regions.Basara et al. (2010) studied the HWs in Oklahoma in 2008 and observed that the HW at nighttime was stronger than that at daytime.Jiang  2019) studied the excess warming of megacities with different climate backgrounds and found that during the heat waves periods, the CUHII of Beijing (temperate semi humid monsoon climate) and Guangzhou (Marine subtropical monsoon climate) at nighttime was higher than that at daytime, while the diurnal variation of Shanghai (subtropical humid monsoon climate) heat island was completely opposite when selecting coastal stations as rural stations.
In addition to the climate background, the local circulation caused by different geographical environments has a significant impact on the temporal and spatial distribution of extreme high temperature in cities (Zhang et al., 2011;Zhou et al., 2020;Chen et al., 2022).Sea land breeze circulation is a secondary atmospheric circulation caused by uneven heating of the underlying surface, and is one of the most prominent mesoscale characteristics in coastal areas (Wagner et al., 2012;He et al., 2020).Yang et al. ( 2023) selected a day with sea-land breeze (24 September 2013) and a day without sea-land breeze (25 September 2013) and analyzed the diurnal variation of the wind vectors and the CUHII on these 2 days.The daytime CUHII was lower at the coastal sites on days with a sea-land breeze than on days without a sea-land breeze as a result of the influence of the transport of cold and wet air from the sea.The daytime CUHII at downwind urban sites was affected by heat transport and was stronger on days with a sea-land breeze (As shown in Figure 5).
The results confirm that sea breezes reduce the CUHII in coastal site of Shanghai.The formation principle of mountain-valley wind is similar to that of land and sea wind, which is caused by temperature and pressure (Xue et al., 2023).Due to the thermal difference between the valley and its surrounding air, the wind blows from the valley to the hillside at daytime and from the hillside to the valley at nighttime.Taking Lanzhou as an example, during HW periods, the WS increased at almost all moments (Figure 6A).The weather  Observation cases of HWs in the Beijing Area.(A-D) are different heat waves periodss.The red solid line and dotted line are the time series of the daily maximum temperature and its 90th quantile, while the blue solid line and dotted line are the time series of the daily minimum temperature and its 90th quantile, respectively.The heat wave process is indicated by shadows (An and Zuo, 2021).
stations were all located in the southern part of the main urban area, so the valley wind circulation of these stations in summer was mainly regulated by the southern mountains.It can be seen from Figure 6B that the local wind circulation had obvious diurnal variations, with negative anomalies during the day (northeast wind) in the u and v directions, and positive anomalies (southwest wind) at night.The anomaly amplitudes of the local wind components in the u and v directions during HW periods were larger than those in NHW periods (Figure 6B), indicating that the local wind had a greater influence in HW periods.During the heat wave, the mountain-valley wind circulation increases the wind speed, which is not conducive to the occurrence of HWs and strong CUHII (Ngarambe et al., 2020;Zong et al., 2021).But it also enhances the surface heat flux and offsets the increased advection cooling (Li and Bou-Zeid, 2013).In addition, the foehn wind is a local airflow movement caused by the mountains, that is, a Diurnal variation of the (A,B and D) wind direction at sites FX, PD and JD and (C,E) the CUHII at sites PD and JD on a day with a sea-land breeze (24 September 2013) and a day without a sea-land breeze (25 September 2013) (Yang et al., 2023).local wind in which the airflow over the mountains sinks on the leeward slope and becomes dry and hot (McGowan, 1997).Wang et al. (2012) used the observation of multi-element auto weather stations (AWSs) to analyze the foehn effect that occurred on the east slope of Taihang Mountains.The results showed that there are two foehn centers in the northern and southern parts of Taihang Mountains, causing the rapid rising temperature in Hebei Plain.On the whole, the formation mechanism of urban excess warming under different climate backgrounds and local circulation still lacks complete understanding.Schematic diagram of local climate zoning (Stewart and Oke, 2012).
Frontiers in Environmental Science frontiersin.org07 5 The influence of urban morphologies on the synergies between HWs and CUHI At present, a broad consensus has been reached that the spatial heterogeneity of urban areas (such as land cover, roughness of urban underlying surface, height of buildings, and sky view factor) can directly lead to the spatial non-uniform distribution of near surface air temperature (Fenner et al., 2017).Compared with rural areas, the urban land surface albedo is lower, which can amplify the CUHI by absorbing more solar radiation (Oke, 1982), and the reduction of evaporation in urban areas further aggravates the CUHI effect (Taha, 1997;Zhao et al., 2018).In addition, urban buildings increase surface roughness, and reduce wind speed, and urban heat is not easy to dissipate (Fujibe, 2003).In the traditional researches on the CUHI effect, the study areas were simply divided into urban and suburban types, that is, the CUHII was calculated by the average temperature difference between urban stations and reference stations.However, this single division calculation ignored the differences of meteorological elements within the city, thus affecting the accurate assessment of CUHI (Ren and Ren, 2011;Yang et al., 2013;Li et al., 2015;Shi et al., 2022).
Stewart&oke (2012) proposed a set of local climate zones (LCZs), which divides the urban underlying surface into 17 basic types according to the characteristics of the urban surface and differences in human activities (Figure 7).Using the temperature calculation of different types of LCZs can more accurately carry out the comparative study of CUHI at regional and global scales.A few scholars used LCZs to study urban excess warming and promoted the quantitative study of the synergies of HWs and CUHI from the perspective of urban morphologies (Ngarambe et al., 2020;Zheng et al., 2022;Xue et al., 2023;Yang et al., 2023).The CUHII under different LCZs was significantly different and was jointly determined by HWs, wind speed and the location of LCZs from the city center (Xue et al., 2023).For example, in LCZs with a high density of buildings that favor CUHII growth, HWs may last longer.The spatial characteristics of extreme high temperatures of different LCZs in Shanghai show that (Figure 8), the CUHII, the number and duration of HWs of highdensity LCZ1 and LCZ2 are significantly higher than those of lowdensity LCZ4, but the average CUHII of low-density LCZ5 is similar to that of high-density LCZ1 (Yang et al., 2023).High-density buildings reduce wind speed, prolong the duration of HWs, and enhance the synergies between HWs and CUHI (Gemechu, 2022).The CUHII is also affected by anthropogenic heat flux, especially in high-density and high-rise building areas (Zhang et al., 2015).In addition, when the wind speed reaches the threshold of horizontal heat transfer (He, 2018), the CUHII of low-density LCZ5 in the downwind direction increases for urban areas with a large number of heat emissions (Yang et al., 2023).It is worth noting that few studies have focused on the thermal differences between regions located in different directions within the urban area.Therefore, it is necessary to expand the research focus from a single suburb to the interior of cities, and analyze the impact of different urban morphologies indicators on the local thermal environment, so as to carry out the research on the synergies of HWs and CUHI in megacities.

Conclusion
This paper reviewed the research on the difference in the research on the synergies between HWs and CUHI and analysed their formation mechanism from the aspects of definition of HWs, climate backgrounds, local circulations, urban morphologies.In general, the interaction between HWs and CUHI is affected by natural factors (such as humidity, wind speed, and geographical location) and human factors (such as anthropogenic heat emissions, land use change and the definitions of HWs).Different methods to identify and define the HWs could affect the occurrence time, duration, and frequency of HWs to a certain extent.The synergies between HWs and CUHI is also related to the climate background and local circulation.The difference in urbanization degree (such as urban morphologies, urban scale, land cover, and anthropogenic heat flux) also affect the synergies between HWs and CUHI.In summary, influenced by different geographical locations, climate backgrounds, local circulations and urban morphologies, the spatiotemporal distribution of urban excess warming is uneven, and its influencing factors also show complexity, diversity, and variability.
In the context of global warming and rapid urbanization, the synergies between HWs and CUHI poses a significant challenge to human health, production safety, environmental protection, and sustainable urban development.The urgent need is to gain a more comprehensive understanding of the mechanisms and regulatory factors responsible for urban excessive warming caused by HWs and CUHI, which will enable better prediction and assessment of future extreme high temperature events.Table 1.

Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers.Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

FIGURE 1
FIGURE 1Diurnal variation of UHII between HW and NHW periods during summertime (June-August) 2014-2020 (A-G) and averaged value (H).HW and NHW periods are indicated by red and blue, respectively.Lines denote average UHII values and shaded areas present the standard deviation of the average UHII values according to all urban reference stations(Zong et al., 2021).

FIGURE 3
FIGURE 3The occurrences of HWs at site Potsdam from 1893 to 2017 during the warm season (May-September) for ten different HW definitions.Each colour corresponds to the number of definitions agreeing for each day to be a HWs day and not to an individual HWs definition.

FIGURE 6
FIGURE 6 Diurnal variation of WS (A) and the local wind component anomaly (B) in the summertime (July-August) of 2013-2018.Diurnal variation of PM2.5 concentration (C) in the summertime (July-August) of 2014-2018.HW and NHW periods are indicated by red and blue, respectively.Shaded areas denote the standard deviation (Xue et al., 2023).

FIGURE 8
FIGURE 8 Overview of the study area (A) and spatial distribution of Shanghai automatic weather stations (B).Box-whisker plots for the CUHII values (C) in each LCZ and (D) at different wind speeds for each urban station during 2013-2018.The black solid lines are the maximum, upper quartile, median, lower quartile and minimum values from top to bottom, the central red solid line represents the average and the red plus symbols represent outliers (Yang et al., 2023).