Introduction

The process of urbanization is accompanied by the great changes in land use structure and the increase in anthropogenic heat emissions. As a consequence, urban heat island (UHI) effects has become one of the main urban environmental problems (Memon et al. 2008; Hong et al. 2022b, a). The existence of UHI seriously affects the thermal comfort of residents, increases the energy consumptions and poses a threat to human health (Hong et al. 2022b, a). Mitigating the UHI effect has become an important issue to ensure the life quality of urban residents and achieve the sustainable development of the city (Li et al. 2015, 2014).

Studies have shown that the introduction of blue spaces can effectively improve urban thermal environment (Syafii et al. 2016, 2017) and alleviate UHI effects (Chang et al. 2007; **n et al. 2010). Blue spaces refer to outdoor urban surfaces such as lakes and ponds (Gunawardena et al. 2017). It has been found that they have a cooler surface due to their large thermal inertia and high evaporation rate (Spronken-Smith et al. 2010; Kleerekoper et al. 2012), which leads to a lower air temperature in their vicinity compared to that in built environment. This phenomenon is known as the water cooling island effect (Syafii et al. 2017). These cool islands can provide cooling to their surroundings (Oke et al. 2017; Ampatzidis et al., 2020), change the aerodynamic characteristics, and improve the transmission of air and pollutants in horizontal directions (Gunawardena et al. 2017).

At present, the vast majority of the water cool island studies (69%) were conducted in humid subtropical climates, lacking observation and research under other climatic conditions (Ampatzidis et al. 2020). Therefore, some research conclusions might be not universally applicable. For example, some researchers have found that water bodies could provide cooling during the day, but it would lead to a warming effect at night (Ampatzidis et al. 2020). However, these conclusions were mainly found in subtropical climates (e.g. Guangzhou, China (Cfa) (Yang et al. 2015); Delft, Netherland (Cfa) (Solcerova et al. 2019) and Rotterdam, Netherland (Cfb) (Heusinkveld et al. 2014)), while the thermal effects of water bodies in other climates (e.g. temperate climates) is lack of information.

To get a further understanding of the thermal effects of water bodies on their surroundings, field studies were conducted in a temperate monsoon climate city (**’an) and a humid subtropical climate city (Chengdu) during summertime. The surface temperature and air temperature above the water bodies were measured and the effects of water bodies on ambient air temperature, humidity and UTCI were investigated. This study contributes to the knowledge of the water cool island effect, and the findings can provide a scientific basis for alleviating UHI effects and improving the outdoor thermal environment.

Material and method

Study field

The measurements were conducted in **’an-Fengqing Park and Chengdu-Dayuan Central Park. **’an (33.42°-34.45°N, 107.40°-109.49°E) is located in Northwest China and the middle of Guanzhong Plain (Fig. 1). It features a temperate monsoon climate (Dwa) with four distinct seasons. It has a cold winter, warm spring, hot summer and cool autumn. The average annual rainfall in **'an is about 560 mm, most of which occurs in summer. **’an has an annual average temperature of 15.6 ℃, and a mean RH of 65%.

Fig. 1
figure 1

Location of the study fields

Full size image

Fengqing Park (34.25°N, 108.90°E) is located in the main urban area of **’an (Fig. 2). It has an area of 17.7 ha. The surrounding area is highly urbanized. The park is covered with open lawns (20%) and bushes (60%). A pool of 26,449 m2 is located in this park.

Fig. 2
figure 2

Satellite picture of Fengqing Park (a) and Dayuan Central Park (b), and locations of the measuring points (Source: Google Earth)

Full size image

Chengdu (30.08°-31.43°N, 102.90°-104.88°E) is located in Southwest China and the west of Sichuan Basin (Fig. 1). It features a humid subtropical climate (Cwa), with abundant precipitation, and abundant sunshine. It is cool in spring and autumn, hot in summer and warm in winter. The average annual rainfall in Chengdu is about 1000 mm, most of which occurs in summer. Compared to **’an, the climate in Chengdu is more humid and the variation of temperature throughout the year is smaller. It has an annual average temperature of 16℃, and a mean RH of 75%.

Dayuan Central park (30.55°N, 104.05°E) is located in the main urban area of Chengdu (Fig. 2). It has an area of 23.9 ha. The surrounding area is highly urbanized. The park is covered with open lawns (60%) and bushes (30%). A pool of 20,212 m2 is located in this park.

Weather conditions

In this study, measurements were conducted in typical summer days. The weather conditions (China Weather) are shown in Table 1.

Table 1 Weather conditions on measuring days
Full size table

The weather conditions of the measuring days are close to those of the local typical meteorological day. In this study, the typical meteorological data were the average data collected from 1995 to 2010 (Zhang et al. 2012). Therefore, the measurement results can represent the thermal conditions in the general cases.

Data collection

The air temperature (\({T}_{a}\)), humidity (RH) and global temperature (\({T}_{g}\)) at a height of 1.5 m at different distances from the edge of water bodies were measured in this study. Three points were selected in Fengqing Park and Dayuan Park which were 19 m, 103 m, 184 m and 17 m, 103 m, 183 m away from the edge of the water bodies respectively. The measuring points were named as near-points, mid-points and far points, which are illustrated with points A, B and C in Fig. 2. The prevailing wind direction was northwest on August 14th and south on September 20th. Therefore, the measuring points was selected along the southeast direction on August 14th and north direction on September 20th to capture the impacts of water bodies on the thermal environment.

Table 2 shows the basic information of the measurement instruments, including the model number, manufacturer, range, accuracy and resolution.

Table 2 Instruments used for measurements
Full size table

In this study, the first measurement was conducted around the sunrise, and the last measurement was conducted was conducted about 4 h after the sunset. The sunrise time on 2021, Augest 14th and 2021, September 20th was 6:04 and 6:50 respectively, and the sunset time on 2021, Augest 14th and 2021, September 20th was 19:33 and 19:03 respectively. The time interval between measurements was set as constant as possible. Therefore, data were recorded at 6:00, 8:00, 10:00, 12:00, 15:00, 17:00, 19:00, 21:00 and 23:00 on 2021, August 14th and at 7:00, 9:00, 11:00, 13:00, 15:00, 17:00, 19:00, 21:00 and 23:00 on 2021, September 20th.

Evaluation of the thermal comfort

This study applied the Universal thermal climate index (UTCI) to evaluate the thermal comfort of the human body. UTCI is an index proposed by the International Society for Biometeorology (ISB) based on the concept of equivalent temperature. UTCI has been proved to be suitable for outdoor thermal comfort research in various regions (Brode et al. 2011). It is divided into ten categories ranging from extreme cold stress to extreme heat stress according to the different thermal sensation of the human body, as shown in Table 3.

Table 3 UTCI values and thermal sensation
Full size table

In this study, UTCI was calculated using the RayMan pro. This model was developed by the Association of German Engineers (Matzarakis et al. 2007). For calculations, it requires basic measured meteorological data, geographic data, human profile and calculation date and time. The assumed human profile in this paper was a typical 35-year-old man with a weight of 75 kg and body height of 1.75 m, with a clothing value of 0.9clo and engages in mild activity. For meteorological data, it is required the input of the four meteorological parameters (\({T}_{a}\), RH, \(v\) (wind speed) and \({T}_{mrt}\)(mean radiant temperature)). The meteorological parameters \({T}_{a}\) and RH can be entered as directly measured data, whereas \(v\) and \({T}_{mrt}\) must be calculated.

\({T}_{mrt}\) is defined as the uniform temperature of a surrounding surface giving of blackbody radiation (\(\varepsilon\)= 1) which results in the same radiation energy gain of a human body as the prevailing radiation fluxes which are usually very varied under open space conditions The calculation of \({T}_{mrt}\) is expressed as (Cheung et al. 2018):

$${T}_{mrt}={\left[{({T}_{g}+273)}^{4}+\frac{1.10\times {10}^{8}{v}^{0.6}}{\varepsilon {D}^{0.4}}({T}_{g}-{T}_{a})\right]}^{0.25}-273$$
(1)

where \(D\) is globe diameter (set to 0.05 m in this study); \(\varepsilon\) is emissivity (set to 0.95 for a black globe);

UTCI requires \(v\) to be measured at 10 m. As it is difficult to obtain wind speeds at 10 m. Therefore, an exponential wind profile (Oke et al. 2017) can be used to approximate the wind speed at 10 m.

$${v}_{z=10}={v}_{z=1.5}{(\frac{10}{{z}_{ref}})}^{0.3}$$
(2)

where \({z}_{ref}\) is set to 1.5 m in this study and the wind speed at 1.5 m is set to 1.5 m/s because both **’an and Chengdu are in the static-wind area with a mean wind speed of approximately 1.5 m/s at pedestrian level (Zhang et al. 2012).

Results

Effects of water bodies on ambient air temperature and humidity

Figure 3 show the air temperature and humidity measured in two parks.

Fig. 3
figure 3

Air temperature and humidity in Fengqing Park (a) and Dayuan Central Park (b)

Full size image

Results showed that the change trends of air temperature and humidity were same at different measuring points. The air temperatures increased after the sun rise and reached the maximum values at about 15:00. The trend of humidity change is opposite to that of temperature change and it reached the lowest values at the same time. By comparing the air temperature and humidity between the different measuring point, it can be seen that both the water bodies could provide cooling and wetting effects in their vicinity, and the closer to the water body, the more obvious the effects.

Figure 4 shows the temperature and humidity difference between near points and far points.

Fig. 4
figure 4

temperature (a) and humidity (b) difference between near points and far points in Fengqing Park and Dayuan Central Park

Full size image

Results showed that the cooling effects of water bodies were more obvious during the daytime than in the morning and at night, while the wetting effects didn’t show obvious diurnal variation.

Thermal effects of water bodies under different climatic conditions

Figure 5 shows the surface temperature of the water bodies and the air temperature above them.

Fig. 5
figure 5

Surface temperature and air temperature of water bodies in Fengqing Park (a) and Dayuan Central Park (b)

Full size image

Results showed that the thermal effects of the water bodies were different. In park of **’an, the surface temperature of the water body was always lower than the temperature of above air. However, in park of Chengdu, the water body was cooler than the air above only in the daytime, while in the morning and during the night, it became 2 ℃ warmer than the air.

The exchange of sensible heat H is determined by:

$$\mathrm{H}={h}_{c}({T}_{s}-{T}_{a})$$
(3)

where \({T}_{s}\) is the surface temperature; and \({h}_{c}\)(W/(m2 \(\cdot\) K)) is heat convection coefficient. An empirical formula can provide reasonable accuracy for \({h}_{c}\) (Qin et al. 2014):

$${h}_{c}\left\{\begin{array}{c}5.6\times {4.0}_{v} v<5\\ 7.2\times {v}^{0.78} v\ge 5\end{array}\right\}$$
(4)

where \({\text{v}}\)(m/s) is wind speed measured at 9.0 m height.

Therefore, according to Fig. 5, the water body in **’an has negative sensible heat exchange all day, while that in Chengdu has positive sensible heat exchange with its surroundings at night, that is, the warming effect.

Effects of water bodies on thermal comfort

Figure 6 shows the UTCI in the two parks.

Fig. 6
figure 6

UTCI values in Fengqing Park (a) and Dayuan Central Park (b)

Full size image

The data show that the water bodies can significantly affect the thermal comfort of its surroundings. The closer to the edge of the water body, the lower the value of UTCI. Table 4 shows the measured daily average UTCI value and air temperature of the measuring points in the two parks.

Table 4 Daily average UTCI in two parks
Full size table

It can be seen that the daily average UTCI difference between the near points and the mid points were 0.6 ℃ and 0.4 ℃ in the two parks, while that between the mid points and far points were 0.6 ℃ and 1.1℃. The daily average air temperature differences between the near points and mid points in the two parks were 0.6 ℃ and 0.6 ℃, while that between the mid points and far points were 0.6 ℃ and 1.0℃. The temperature difference and UTCI difference were close, which implied that the comfort achieved by the cooling effect was not cancelled out by the humidity change.

According to the data, the difference in UTCI value between the two parks was not more than 1℃, which indicates a similar thermal comfort. However, this conclusion is inconsistent with the actual situation. People who live in Chengdu and **'an at the same time will always feel more uncomfortable in Chengdu during the summer because of the higher air humidity.

In fact, the actual thermal sensation of the human body is not completely consistent with the thermal perception determined by UTCI values. Many studies have shown that the neutral UTCI range are different in cities with different climates (Xu et al. 2018; Hadianpour et al., 2018).

To ensure the thermal benchmarks such as UTCI can reflect the thermal comfort more appropriately, many researchers conducted questionnaire surveys to modify the relationship between UTCI and thermal sensation. Xu et al. (2019) investigated the thermal comfort in a park in **’an by conducting a subjective thermal sensation questionnaire. The weighted means of TSV (thermal sensation vote) versus 1 °C UTCI during summer, fitted with linear model was obtained. Equation (1) shows the relationship between UTCI and MTSV.

$$y=0.102x-2.360 ({R}^{2}=0.98)$$
(5)

where \(y\) is the mean thermal sensation vote, \(x\) is the UTCI value in ℃, \({R}^{2}\) is the correlation coefficient of the regression equations.

The relationship between TSV and thermal sensation is shown in Table 5:

Table 5 TSV value and corresponding thermal sensation
Full size table

Wei et al. (2021) investigated the relationship between human thermal sensation and thermal benchmarks in a park in Chengdu, and also obtained the relationship between UTCI and MTSV during summer, which is shown in Eq. (2).

$$y=0.1807x-4.4724 ({R}^{2}=0.8339)$$
(6)

According to the above regression equations, the TSV of the measuring points in Fengqing Park and Dayuan Central Park can be calculated, which are shown in the Table 6:

Table 6 Daily average TSV in two parks
Full size table

The results show that the daily average thermal sensation of Fengqing park was warm, while that of Dayuan park was hot. Although their UTCI is close, the daily average thermal comfort of **'an park is significantly better than that of Chengdu, which is consistent with the actual body feelings.

Discussions

Some previous studies have reported the warming effect of water bodies. For example, Yang et al. (2015) compared the measured surface temperature of a pond with the meteorological air temperature in Guangzhou (Cfa), it was found that the water pond lowered the air temperature during the daytime but decreased the cooling rates of the atmosphere above it during the nighttime. Heusinkveld et al. (2014) found in Rotterdam (Cfb) that the water body had negative impact on the thermal environment at night and increase the maximum UHI intensity. In this study, we found the water body in Chengdu (Cwa) had warming effect at night while that in **’an (Dwa) did not. To make sure that the measurements in **’an-Fengqing Park is not a special case study. Figure 7 a and b show the results of two other measurements in **'an, which were the measurement of the pond in **’an-iHarhour on July 5th, 2021 and the measurement of the pool in **’an-**ngqing Park on July 13th, 2021. The data show that the surface temperatures of the water bodies were always lower than temperature of the air above, which were similar to that of the measurements on August 14th, so it can be inferred that the thermal effect of water bodies in **'an is different from that in Chengdu.

Fig. 7
figure 7

Surface temperature and air temperature of water bodies in **’an-iHarbour (a) and **’an- **ngqing Park (b) (the measuring time was selected according to the sunrise time)

Full size image

Since the maximum intensity of urban heat island often occurs at night (Martilli et al. 2020), therefore; this finding indicates that the thermal effects of local water bodies should be fully investigated before introducing water bodies to mitigate urban heat island effects. Further efforts are expected to investigate the thermal effects of water bodies under other climatic and find out how climatic conditions affect them, which will help provide more information for solving urban heat problems through water bodies.

Many studies have investigated the impact of water on the ambient temperature and humidity. However, few studies have investigated the impact of water bodies on thermal comfort in their vicinity. Theeuwes et al. (2013) and Saaroni et al. (2003) found that the water bodies could improve the thermal comfort in their vicinity, but part of the comfort achieved by the cooling effect is cancelled out by the humidity change. Zhao et al. (2017) found through simulation that when the water area covers less than 34% of the domain, the existence of the water body will lead to a worse PET (Physiological equivalent temperature) value due the increase of humidity. In this study, it was found in both **’an and Chengdu that the thermal comfort achieved by the cooling effect of water bodies was not cancelled out by the humidity change. In addition, it was found that the actual thermal sensation could be different for similar UTCI values in different cities. Therefore, it is necessary to modify the thermal benchmarks before investigating the thermal comfort. This can ensure a true reflection of the local thermal environment and provide appropriate benchmarks for thermal comfort improvement measures. In the future, it is expected to conducted further studies to investigate the influence of water bodies on surrounding thermal comfort under other climatic conditions.

Conclusions

To get a further understanding of the effects of water bodies on the surroundings, field measurements were conducted in this study. The surface temperature and air temperature above the water bodies were measured and the effects of water bodies on ambient air temperature, humidity and UTCI were investigated in a temperate monsoon climate city (**’an) and a humid subtropical climate city (Chengdu) respectively in summer.

Results showed that both the water bodies could reduce the air temperature and increase the humidity in their vicinity, and the closer to the water body the more obvious the effect. In addition, it was found that the cooling effects of water bodies were more obvious during the daytime than in the morning and at night, while the wetting effects didn’t show obvious diurnal variation. By comparing the surface temperature of the water bodies with the air temperature above them, it was found that the water body in **’an could provide cooling to its surroundings all day, while the one in Chengdu had warming effect at night, which demonstrate the fact that the local climate has a great impact on the thermal effects of the water bodies.

In this study, the UTCI around the water bodies was also investigated. It was found that the water bodies would significantly improve the thermal comfort in their vicinity, and the thermal comfort achieved by the cooling effect was not cancelled out by the humidity change. In addition, it was found that the actual average thermal sensation of the human body in **’an park was warm, while that in Chengdu park was hot when the UTCI values were close. This indicated the necessity to modify the thermal benchmarks before investigating the thermal comfort in different cities.

This study contributes to the knowledge of the thermal effects of water bodies in different climate regions. We believe our findings can provide a scientific basis for alleviating UHI effects and improving the outdoor thermal environment. Further measurements are suggested to be continuously carried out in other climates, which is lack of research, to deepen the understanding of the water cooling island effects.