Long-Term Spatiotemporal Patterns and Evolution of Regional Heat Islands in the Beijing–Tianjin–Hebei Urban Agglomeration

Abstract

With the continuous development of urbanization, the urban heat island (UHI) phenomenon is becoming increasingly prominent. Especially with the development of various large urban agglomerations and the shrinking distance between cities, the regional thermal environment has attracted extensive attention. Therefore, we used Modis land surface temperature (LST) data and employed least squares, standard deviation and spatial autocorrelation analysis methods to analyze the spatiotemporal patterns and characteristics of summer daytime regional urban heat islands (RHI) in the Beijing–Tianjin–Hebei (BTH) urban agglomeration. Our results indicated that the relative land surface temperature (RLST) in the southeastern part of BTH with a relatively high level of urbanization showed a significant and continuous upward trend. With the continuous development of the level of urbanization in the southeast, the center of gravity (GC) of RHI gradually moved to the southeast, and the development direction of RHI changed from northwest–southeast to northeast–southwest. The area transfer of RHI was concentrated in no change and little change, indicating that the evolution trend of RHI was relatively stable. The high-high aggregation areas were mainly located in the more developed areas in the southeast. In addition, the methods and results of this study can provide reasonable and effective insights into the future development and planning of the BTH.

1. Introduction

In recent years, cities in China and the whole world have undergone rapid development. As urbanization continues, a large number of natural and semi-natural surfaces are transformed into impervious surfaces, which disturb the albedo and energy balance of the city and alter the thermal properties of urban areas [1,2,3]. Although the urban area accounts for less than 1% of China’s overall landmass, some recorded urban high-temperature events were undoubtedly the consequence of rapid urban development, particularly since the late 1970s [4,5]. Rapid urbanization begins with an increase in the built-up area, which leads to the intense expansion of impervious surfaces. Impervious surfaces are often considered to be an important factor in urban environmental change. They have a strong heat storage capacity and weak evaporation capacity, which hinders the transport of airflow and thus makes the city warmer [6]. In addition, different building heights, densities, materials, colors, etc., all have an impact on the land surface temperature (LST) of the city. Meanwhile, climate change is increasingly recognized as an important factor that aggravates the urban heat island (UHI) effect [7,8,9,10]. The UHI is a phenomenon in which the urban temperature is significantly higher than that of the surrounding suburbs [11,12]. Lin et al. found that the air temperature in the northeastern United States increased from 2 °C to 5 °C with the expansion of cities [13]. Chen et al. found that the average LST difference between urban and rural areas reached 15 °C in Taipei city, China [14]. In Rennes, western France, the highest difference in temperature between the urban center and the rural area was above 3 °C [15], and heat island intensity peaks were as high as 6 °C in the Sydney metropolitan area [16].

With the continuous development of urbanization, the temperature in urban areas is also increasing. According to previous studies, the average daily temperature in summer in urban areas of more than 500 square kilometers rose by 4.7 °C from 2003 to 2005 [17,18]. Urban air temperatures in the United States have risen by 0.24 °C per decade [19]. Cities in South and Southeast Asian countries, on the other hand, have risen by 0.23–0.57 °C per decade in recent decades [20,21,22]. The continuous rise in urban temperature has made the UHI problem increasingly prominent. It not only causes air pollution, increasing energy and water consumption [23,24,25,26], but also causes respiratory, cerebrovascular, heart and other related diseases that endanger the health of residents and even increases mortality in urban areas [19,27,28,29]. In addition, area transfer will occur between non-heat island areas and heat island areas, and heat island areas at all levels. For example, non-heat island areas are converted to heat island areas, and low-grade UHI is converted to high-grade UHI. This further exacerbates the UHI problem and affects the health and life of urban residents. Therefore, the UHI phenomenon has gained the attention of researchers around the world, such in as North America [3,30,31,32], South America [33], Europe [15,34,35], Africa [36], South Asia [37,38], Southeast Asia [39,40,41], East Asia [42,43,44] and many other cities.

Urban agglomeration, generally refers to a chain of roughly adjacent metropolitan areas that may be separated to some extent or connected to form a contiguous urban area. With the development of urbanization, urban agglomeration has become the most prominent feature of global urbanization in recent decades, representing compact spatial organization and close economic connections in a specific geographic area [45,46,47]. Due to the shortening and disappearance of the distances between cities, urban agglomerations can drastically change the regional thermal environment [48,49]. Regional energy and material flows make it impossible to solve the heat island problem in urban agglomerations by relying solely on a single city. Therefore, the study of UHI in urban agglomerations is a good starting point to address the regional thermal environment. Yu et al. proposed that the traditional UHI appears as a regional urban heat island (RHI) on an urban agglomeration scale [50]. An increasing number of scholars have conducted studies on the thermal environment of larger urban agglomerations around the world; for example, the studies of UHI in China’s three fastest-growing urban agglomerations, the Yangtze River Delta (YRD) [48], the Pearl River Delta (PRD) [50] and the Beijing–Tianjin–Hebei urban agglomeration (BTH) [51,52]. Cities in Japan’s Pacific Rim cluster (JP) were struggling with the heat island effect of intense metropolitan areas, especially Tokyo, which is a fast-growing city with rising day and night temperatures [53]. As the earliest region to study the UHI effect, the London metropolitan area (LD) has also become a hot spot for heat island impacts [54]. Many cities in the North American Atlantic Coastal agglomeration (NAAC) and North American Great Lakes agglomeration (NAGL) also suffer from strong urban heat island problems [55,56].

BTH has a typical temperate monsoon climate, in which summer is hot and rainy, with average summer temperatures above 30 °C and the highest temperatures reaching 40 °C. The continuous high temperature has strengthened the UHI effect in BTH [7]. This has led to a high incidence of heat events, which continues to draw the attention of the government and the scientific community [57]. Furthermore, in previous studies of UHI in the BTH, the UHI in summer was obviously stronger than in other seasons [58,59], and the UHI was the strongest during the daytime in summer [59,60]. In addition, with the rapid growth of the population and economy in BTH, the relationship between UHI and socioeconomic development in this region has also received increasing attention. For instance, Hou et al. examined the influence of natural and socioeconomic factors (average summer air temperature; total summer precipitation; and the proportion of impervious surfaces in the research units) on summer surface UHI [58]. Zhao et al. selected the BTH as a study area to study the surface UHI intensity based on remotely sensed LST data; different patterns of the seasonal variations were found in daytime and nighttime surface UHI intensity [60].

Many studies have been conducted on the relationship between UHI and urbanization in BTH, the urban thermal environmental risks and the factors influencing RHIs [57,61,62]. Similarly, many scholars have studied the diurnal, monthly and seasonal differences in UHI in the BTH from different time scales [58,59,60]. However, we found that in the previous studies on UHI of the BTH, firstly, its time span was relatively small, mostly 5–10 years; secondly, its spatial scale was also small, and it mainly studied the characteristics of heat islands in urban areas, and rarely study the evolution of RHIs in the entire urban agglomeration from the perspective of the BTH as a whole. Therefore, MODIS-LST data from 2001 to 2020 were used to analyze the spatial and temporal patterns of the summer daytime RHI in the BTH and its spatiotemporal evolutionary characteristics. There are two major aims for this paper: (1) to explore the spatial distribution and evolutionary trend of RHIs and (2) to reveal the area transfer and spatial autocorrelation of the RHI in the BTH from 2001 to 2020.

2. Study Area

The BTH is a highly representative regional urban agglomeration of large, medium and small cities and is also a hot area of regional thermal environmental research in recent years. The BTH consists of two municipalities, Beijing and Tianjin, and the eleven cities (Shijiazhuang, Tangshan, Qinhuangdao, Handan, Xingtai, Baoding, Zhangjiakou, Chengde, Cangzhou, Langfang and Hengshui) of Hebei Province (Figure 1). The BTH is one of the most developed regions in China. In 2019, the BTH, which only accounts for 2.3% of China’s total area, has a population of 102.2 million and a GDP of 8447.6 billion yuan (Table 1) [51]. Beijing, Baoding, Tianjin and Shijiazhuang have a total population of more than 10 million. Among them, Beijing has the largest population of about 13.87 million. In addition, Beijing and Tianjin have the highest GDP at 3537.1 billion yuan and 1410.4 billion yuan respectively. In recent years, with the continuous development of urbanization, the built-up area of the BTH has also increased. In 2019, the built-up area of the BTH reached 4228 square kilometers, of which the built-up area in Beijing accounted for about one-third. In addition, the land use/cover types of the BTH are mainly cultivated land (43.73%), forest land (24.93%) and grassland (14.93%). The cultivated land is mainly located in the southeast plain area and the Zhangjiakou in the northwest area, the forest land is mainly distributed in the Taihang Mountains and Yanshan Mountains in the northwest, and the grassland is mainly located in Zhangjiakou and other places in the west of the Taihang Mountains. The overall terrain of the BTH is high in the northwest and low in the southeast. The Taihang Mountains and Yanshan Mountains are in the northwest, and the southeast is dominated by plains. Most cities are located in the southeastern plain area, belonging to a temperate continental monsoon climate, with hot and rainy summers and cold and dry winters, while Zhangjiakou and Chengde are mainly located in the northwestern mountainous area with relatively low temperatures [63]. In addition, the northwestern part of Beijing is Yanshan, and the mountainous area occupies 62% of the total area, while the plains suitable for urban development regions account for only 38%.

3. Data and Methods

3.1. LST Data

Previous studies on UHI in the BTH have shown that summer daytime is the most obvious time of year for UHI [58,59,60]. Therefore, in order to better study the spatiotemporal characteristics of UHI in the BTH, the data we selected were the average pixel-by-pixel land surface temperature and emissivity (LST&E) provided by the MOD11A2 Version 6 product, with a spatial resolution of 1 km. Each pixel value in MOD11A2 was a simple average of all corresponding (MOD11A1) (https://doi.org/10.5067/MODIS/MOD11A1.006, accessed on 6 December 2021) LST pixels collected over an 8-day period. The product contained LST at night and during the day, based on research needs we selected daytime data. In addition, we filtered surface temperature images that were heavily polluted by clouds, and excluded LST images with more than 30% invalid pixels in the study area.

3.2. Methods

The spatiotemporal pattern and characteristics of the RHI in the BTH were analyzed from three aspects (Figure 2): the spatiotemporal distribution, evolution trend, and spatiotemporal characteristics of the RHI in this paper. Statistical analysis methods were used to calculate the mean and coefficient of variation (CV) in the LST of each city to describe the spatial distribution of the RHI. The least-squares model, standard deviation ellipse (SDE) and transition matrix were used to describe the spatiotemporal evolution of RHIs. Finally, spatial autocorrelation methods, including global Moran’s I and local Moran’s I, were used to analyze the spatiotemporal characteristics of RHIs. We adopted a series of methods and measures to explore the spatiotemporal pattern and evolution trend of the RHI in the BTH as a whole, so as to have a more comprehensive understanding of the overall development of the RHI.

3.2.1. Relative Land Surface Temperature (RLST)

The RLST was used to study the spatial variation in LST in the study area and its temporal evolution [64]. It characterizes the intensity of heat islands in the region by the magnitude of temperature differences in different geographic locations and thus analyzes the contribution of the region to the overall regional thermal environmental dynamics. The RLST equation is as follows:

(1)$RLS{T}_j=LS{T}_j-\overline{{LS}_{T_j}}$

where RLST_j is the RLST in year j; $LS{T}_{j }$ is the LST of each grid in year j; $\overline{LS{T}_j}$ is the mean value of LST for the entire study area in year j; and j is the year used in this study, which is 2001 to 2020. The average temperature of the whole BTH was used as the background value, and the added value obtained by subtracting the background value from the LST of each grid was the RLST. The contribution of each region to the overall regional thermal environment was judged by the RLST, and according to Qiu et al., the area of 2 °C < RLST was called a high-temperature zone, while Yu et al. defined it as the RHI in their study of the PRD [6,50]. In order to understand the spatiotemporal pattern and evolution of RHIs in the BTH more deeply, we divided the part of 2 °C < RLST into three classes: 2 °C < RLST ≤ 4 °C, 4 °C < RLST ≤ 6 °C and 6 °C < RLST, and expressed as class I, class II and class III, respectively.

3.2.2. Trend Analysis of RHI Development

To demonstrate the trend of RHIs, we used the least-squares model. The linear regression of the model can reflect the trend of RHIs over time. The slope in the regression analysis indicates the change in RHIs in the BTH from 2001 to 2020. A positive value indicates a continuing trend of increasing RLST in the area. A negative value, on the other hand, indicates a continuous downward trend in RLST in the area. The slope is expressed as:

(2)$slope=\frac{n\ast {\sum }_{i=1}^n(i\ast RLS{T}_i)-\left({\sum }_{i=1}^{n}i\right)\left({\sum }_{i=1}^{n}RLS{T}_i\right)}{n\ast {\sum }_{i=1}^n{i}^2-{\left({\sum }_{i=1}^{n}i\right)}^2}$

where RLST_i is the value of the RLST of each grid, n (taken as 20 in this study) is the time span, and i is the time unit ranging from 1 to 20.

The least-squares method was used to describe the change trend of RHIs, and in order to further explore the significance of the change trend, Pearson correlation was used to test the statistical significance. The formula is as follows:

(3)$r=\frac{{\sum }_{i=1}^n\left(RLS{T}_i-\overline{RLST}\right)\left(t_i-\overline{t}\right)}{\sqrt{{\sum }_{i=1}^n{\left(RLS{T}_i-\overline{RLST}\right)}^2}\sqrt{{\sum }_{i=1}^n{\left(t_i-\overline{t}\right)}^2}}$

where r is the correlation coefficient and t is the time series from 1 to 20. The correlation coefficients range from −1 to 1. According to the p-value table corresponding to the Pearson correlation, our sample size was 20, so the correlation is significant at p < 0.05 when |r| ≥ 0.42.

SDE is an effective method that can correctly reveal the overall characteristics of the spatial distribution of geographic elements [65]. This method uses center of gravity (GC) of the geographic element distribution as the center, the main trend direction as the orientation angle, and the standard deviation in the x-axis direction and y-axis direction as the long and short axes of the ellipse, respectively. Thus, a spatial distribution ellipse is constructed to describe and explain the characteristics such as directionality and spatial distribution pattern of geographic elements [66,67]. The SDE is calculated by

(4)$SD{E}_x=\sqrt{\frac{{\sum }_{i=1}^n{\left(x_i-\overline{X}\right)}^2}{n}}$

(5)$SD{E}_y=\sqrt{\frac{{\sum }_{i=1}^n{\left(y_i-\overline{Y}\right)}^2}{n}}$

where SDEx is the standard deviation of the x-axis; $SD{E}_y$ is the standard deviation of the y-axis; x_i and y_i are the x and y coordinates of each point element, respectively; and $\overline{X}$ and $\overline{Y}$ are the average x and y coordinates of point elements across the study area, respectively. The SDE was used to describe the spatial distribution of RHIs, as well as the evolutionary trend and development direction of RHIs in the BTH during the two decades from 2001 to 2020.

3.2.3. Area Transfer Analysis

In order to better understand RHI changes in the BTH from 2001 to 2020, starting from the interior of RHI, we explored the area transfer between the RHI at different classes. In addition, in order to explore the evolution degree of RHI, we divided the transfer of RHI at each level into three degrees: large increase, little/no change (little changes include little increase and little decrease), and large decrease (Table 2) [68]. Among them, “Value” refers to the level over which the RHI is transferred at all levels. A positive value is the transfer from a low level to a high level, 0 is no transfer, and a negative value is the transfer from a high level to a low level.

3.2.4. Spatial Autocorrelation Analysis

According to Tobler’s first geographic law, the closer the spatial distance between objects is, the greater the correlation between their attribute values, i.e., the stronger the spatial dependence [69]. In this study, global spatial autocorrelation (global Moran’s I) and local spatial autocorrelation (local Moran’s I) were used to quantitatively measure the spatial heterogeneity of RLST. The global Moran’s I reflects the spatial autocorrelation of the RLST and ranges from −1 to 1. Then, high positive values of global Moran’s I indicate that high-density areas are closely clustered [33], and local Moran’s I is used to reveal the agglomeration and differentiation characteristics of geographic features in spatial locations, reflecting the spatial correlation of the RLST of the spatial unit and the adjacent unit [70]. The local Moran’s I is divided into four types: low-high aggregation (areas with low RLST surrounded by high values), high-low aggregation (areas with high RLST surrounded by low values), high-high aggregation (areas with high RLST surrounded by high values) and low-low aggregation (areas with low RLST surrounded by low values). The calculations are shown in Equations (6) and (7).

(6)$Global Mora{n}^{\prime }s I=\frac{n{\sum }_{i=1}^n{\sum }_{j\ne i}^n{w}_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}{{\sum }_{i=1}^n{\sum }_{j\ne i}^n{w}_{ij}{\sum }_{i=1}^n{\left(x_i-\overline{x}\right)}^2}$

(7)$Local Mora{n}^{\prime }s I=\frac{n\left(x_i-\overline{x}\right){\sum }_{j=1}^n{w}_{ij}\left(x_j-\overline{x}\right)}{{\sum }_{i=1}^n{\left(x_i-\overline{x}\right)}^2}$

where x is the mean observed value for all n positions (areas), $w_{ij}$ is the spatial weight matrix, and $x_i$ and $x_j$ represent the observed values for spatial positions i and j, respectively.

4. Results

4.1. Spatial Pattern of LST

Table 3 presented the maximum, minimum, average values, standard deviation, and CV of LST from 2001 to 2020. The highest average LST was Langfang (32.6 °C), and the lowest was Chengde (27.4 °C). The standard deviations of average LST in Zhangjiakou (3.2 °C) and Beijing (3.1 °C) were higher than other cities. Moreover, comparing the temperature changes in each city, it can be found that the CV values of LST of each city were less than 30%, which also demonstrated that the results of LST were reliable [71]. Among them, the CV values of Zhangjiakou and Beijing were larger, 10.42% and 10.25%, respectively, while the values of Hengshui, Langfang and Cangzhou were smaller, 2.74%, 3.57% and 3.59%, respectively. The LST of Hengshui, Langfang and Cangzhou was relatively stable, and was less affected by other factors; while the temperature difference between Zhangjiakou and Beijing was relatively large, and it was strongly affected by surrounding factors.

4.2. RHI Pattern

Spatial trends of RLST development at the pixel scale were analyzed by ordinary least-squares regression, and the results showed strong spatial heterogeneity (Figure 3a). The RLST in approximately 48.9% of the study area gradually decreased from 2001 to 2020 (slope < 0), while 51.1% of the area had a gradual increase (slope > 0). Pixels with increasing RLST trends from 2001 to 2020 were mainly in the southeast, especially the southeastern part of Beijing, the eastern part of Baoding, and cities such as Tangshan, Tianjin, Langfang, Shijiazhuang, Xingtai and Handan. The northwest illustrated a trend of continuous decrease, especially in Zhangjiakou and Chengde, which is the ecological conservation area of BTH in recent years.

The results indicated that the trends of RLST in most areas of the BTH were small differences between significant (p < 0.05) and non-significant (p ≥ 0.05) (Figure 3b). RLST trends were significant in 51.2% of the area. In the entire study area, the significant increasing area (p < 0.05) was larger than the significant decreasing area (p < 0.05), with area ratios of 27.8% and 23.4%, respectively. The significant increasing areas were mainly located in eastern cities (except for Cangzhou and Qinhuangdao) of the Taihang Mountains and Yanshan, the cities with higher levels of urbanization. The significant decrease was mainly located in the northwestern part (Zhangjiakou and Chengde) of the BTH.

Figure 4 presented the evolution direction of RHI in the BTH and the result of the shift in GC. From 2001 to 2020, the GC of RHI gradually shifted to the southeast. In 2001, GC of RHI was on the border between Zhangjiakou and Baoding in 2001, which gradually moved to the south and was located in Baoding from 2001 to 2010. However, from 2010 to 2020, the GC gradually shifted to the southeast. Table 4 also shows that the rotations from 2001 to 2010 were all approximately 170°, which indicates that the main direction of RHI development was northwest−southeast. However, the value of rotation decreased from 178.58° in 2010 to 30.46° in 2020, which indicates that the spatial pattern of RHIs gradually shifted from the northwest–southeast direction to the northeast−southwest direction with the rapid development of urbanization in the southeastern of BTH.

4.3. RHI Area Transfer Results

Figure 5 and Figure 6 visualize the transformation trajectories among RHI at three levels in the study area from 2001 to 2020. The total area of RHI in BTH had an overall increasing trend with rapid urbanization from 2001 to 2020 (Figure 5). Changes in RHI were mainly located in the southeast and concentrated in little change/no change, its area transfer ratio was as high as 97.01% (Figure 6). Among them, the highest proportion of no change was about 69.45%, the little change was 27.57%, including little increase (value = 1) and little decrease (value = −1), and their area transfer accounted for 13.76% and 13.8%, respectively. Little changes are mainly located in Beijing, Tianjin, and Shijiazhuang, Xingtai, Handan and other cities in the east of Taihang Mountain. The area transfer of large increase and large decrease accounted for a smaller proportion, 1.52% and 1.47%, respectively. The distribution of large increase and large decrease was relatively sparse. From 2018 to 2019, there was a large increase in Xingtai and Handan, and from 2019 to 2020, there was a significant decrease in this region. This revealed that the evolution trend of RHI in BTH was relatively stable and the development trend was relatively flat.

4.4. RLST Spatial Autocorrelation

The spatial heterogeneity of the RHI pattern of the BTH represented by the value of Moran’s I was shown in Figure 7. The global Moran’s I of RLST in the BTH showed an overall increasing trend. The global Moran’s I values were all greater than 0.95, showing a significant positive spatial correlation (Figure 7), and also revealed that the RLST was strongly affected by spatial agglomeration with adjacent grids. In addition, from 2001 to 2020, the global Moran’s I values showed a gradual upward trend (R² = 0.4841), indicating that the positive spatial autocorrelation of the RLST in the BTH continued to increase.

Spatial clusters of RLST from 2001 to 2020 are shown in Figure 8. The local spatial autocorrelation of RLST was mainly concentrated in two clustering patterns, high-high aggregation and low-low aggregation. In contrast, the distribution of low-high aggregation and high-low aggregation was relatively small. Before 2012, high-high aggregation was mainly distributed in the northwest corner of Zhangjiakou, Beijing, Cangzhou, Langfang, Shijiazhuang, Xingtai and Handan in the east of Taihang Mountain. In recent years (after 2012), high-high aggregation was mainly concentrated in areas with a high level of urbanization in the southeast, such as Beijing, Tianjin, Langfang, and Shijiazhuang, Xingtai and Handan, and with the development of urbanization, the area of high-high aggregation in the southeast continues to increase. For the whole study area, the non-significant clusters were widely distributed, indicating that the thermal homogeneity of the entire urban agglomeration on the RLST level is relatively weak.

5. Discussion

5.1. Characteristics and Influencing Factors of RHI in the BTH

According to the above-mentioned analysis, RHI in the BTH was mainly distributed in the southeastern region. With the rapid urban development in the southeast, the area of RHI was gradually increasing. The GC of the RHI gradually shifted to the southeast, and the area transfer among the RHI at all levels in the southeast region was relatively smooth; the RHI in the BTH showed a smooth increasing trend. In contrast, the area of RHI in the northwest was relatively small, and there was a significant downward trend in the RHI. The results were consistent with a study of surface UHI in the BTH by Liu et al. [59], and they pointed out the emergence of a cold island phenomenon in the northwestern part of the BTH and a strong surface UHI phenomenon in the southeast. In addition, they also suggested that the cold islands in the northwestern part of BTH occur because the northwestern part is dominated by mountains with high forest cover, where in recent years, due to the trial of ecological protection measures such as returning cultivated land to forest land and grassland in the northwest, the RHI effect was gradually weakening. However, the continuous enhancement of the RHI in the southeast was closely related to the development of urbanization. Estoque et al. investigated, with the employment of urban–rural gradient analysis and multiresolution analysis, three cities (Bangkok, Jakarta and Manila) in Southeast Asia and found a strong correlation between the mean LST and the density of impervious surfaces (positive) and green spaces (negative) along the urban–rural gradients in the three cities [40]. The UHI study in Greater Cairo indicated that built-up areas had a warming effect during the daytime [36]. Yu et al. also found that the RLST in a forest area was low, with negative values between −3 and 0 °C, and the RLST of the impervious surfaces reached 3–8 °C [50]. This indicates that forests have a significant cooling effect, while impervious surfaces have a warming effect.

In addition to the influence of different land use/cover types on the heat island effect of the BTH, meteorological conditions and socioeconomic factors also have a certain impact on the heat island. Hou et al. analyzed the influencing factors of the heat island effect in the BTH and pointed out that the surface UHI intensity was significantly negatively correlated with the total precipitation, and the average temperature contributed greatly to the interpretation of the summer surface UHI intensity [58]. Liu et al. found that surface UHI was significantly positively correlated with GDP, urban population, electricity consumption, and built-up area. In the study on the influencing factors of surface UHI in China’s three major urban agglomerations, urban population and electricity consumption were socioeconomic factors affecting urban heat islands in urban agglomerations [59]. In addition, Zhao et al. found that the monthly average daytime surface UHI intensity of the BTH was closely related to the monthly average plant leaf area index and albedo. Among them, the monthly average daytime surface UHI intensity was significantly negatively correlated with the monthly average leaf area index (r = 0.90, p < 0.001), while the correlation between albedo and surface UHI intensity was similar to that of the leaf area index [60].

5.2. Definition of Regional Urban Heat Islands

As cities continue to expand, the distances between them are shrinking, coupled with the booming development of many urban agglomerations in China. An increasing number of studies have begun to focus on urban agglomerations, especially the three major urban agglomerations in China, the BTH, the YRD and the PRD. For example, Yu et al. used multiscale analysis and landscape analysis to analyze the spatial and temporal characteristics of the RHI in the PRD [50]. Du et al. used remote sensing and other data to study the UHI and its relationship with types of land, meteorological conditions, anthropogenic heat sources and urban areas in the YRD [48]. Li et al. used the Weather Research and Forecasting (WRF) model combined with remote sensing to explore the urbanization impacts on local circulations in the BTH [30]. Liu et al. compared the temporal and spatial characteristics of surface UHI within three major urban agglomerations (BTH, YRD and PRD) [59]. The study of UHI in the traditional sense of individual cities has gradually transformed into RHI in urban agglomerations. Yu et al. proposed that the traditional UHI appears as RHI on an urban agglomeration scale [50]. Rao et al. systematically elucidated the definition of RHI. At the entire regional scale, interconnected urbanization processes and the area surrounding the urban site are also affected by the increased temperature of the urban site, leading to an increase in surface temperature over a large area and forming an RHI [50,72].

In contrast to the traditional UHI, the RHI covers a broader range and provides a more comprehensive overview of the development of the urban thermal environment in a region. Moreover, as cities continue to expand and the distance between them shrinks, the development of a city’s internal heat island can no longer be attributed solely to the internal structure and layout of the city. In this paper, we found that the RHI around Beijing, Langfang and Tianjin and around Shijiazhuang, Xingtai and Handan in the eastern part of the Taihang Mountains are developing rapidly and gradually connecting into a “heat region”. Therefore, to mitigate the development of UHIs and RHI, urban planners should consider the isolation of heat islands between cities in future urban planning; for example, building a green belt between Tianjin and Beijing and enhancing green infrastructure between Shijiazhuang, Xingtai and Handan. Although there have been relatively few studies on RHI, the study of RHI will become a hotspot in the future.

5.3. Limitations and Contributions

Certain limitations were observed in this paper. First, the precision of the selected data is not high. MODIS LST (1 km) has a low-spatial resolution and does not represent the local geothermal conditions well. Second, in this study, only LST in summer daytime was chosen for consideration. This does not provide a comprehensive overview of the development of RHI in BTH and does not reveal the differences in RHI over seasons.

In this study, the traditional calculation method of UHI was not used to study the thermal environment of the city. Instead, the spatial and temporal patterns and evolutionary characteristics of RHI from 2001 to 2020 were explored with the whole BTH as the study area, and by improving some of the previous research methods, we were able to analyze RHI more effectively and accurately. In addition, there were relatively few studies on RHI, and therefore, our methodology and results could be helpful to subsequent researchers. Likewise, it could also provide scientific advice to city planners and managers.

6. Conclusions

In this study, we used the least-squares model, SDE, area transfer matrix and Moran’s I to illustrate the spatiotemporal patterns and evolution of RHI in BTH during the two decades from 2001 to 2020. The following conclusions can be summarized as follows: (1) The average summer daytime temperature in cities in the southeast was relatively high, especially in Langfang, Cangzhou, Xingtai and Handan. This was mainly because of the development of urbanization in the southeast region, the continuous expansion of urban areas, and the increase in impervious surfaces, which makes the ground temperature rise. (2) The continuous rise of the surface temperature in the southeastern region also makes the RLST in this area present a significant and continuous upward trend. In the northwestern, due to the implementation of ecological protection measures such as returning farmland to forests and grasslands, the RLST showed a significant downward trend. (3) With the development of urbanization in the southeast, the GC of RHI (2 °C < RLST) gradually moved to the southeast, and the evolution direction of RHI also changed from northwest–southeast to northeast–southwest. (4) Through the evolution direction of RHI at each level, it can be found that the area transfer of heat islands at all levels was concentrated in no change and little change. This indicates that the internal development trend of RHI was relatively stable. (5) Through the calculation of Moran’s I, it was found that RLST has a very significant positive spatial correlation. Moreover, high-high agglomeration areas were mainly located in the more developed urban areas in the southeast. The low-low agglomeration was mainly located in the northwestern mountainous area. Non-significant regions had a wider distribution, which indicated that the thermal homogeneity on the RLST level was relatively weak for the BTH as a whole. The results of this paper can provide a scientific basis for the future urban development and planning of the BTH.

Author Contributions

Conceptualization, H.X. and C.L.; methodology, H.X. and R.Z.; software, H.W.; writing—original draft preparation, H.X.; writing—review and editing, C.L., M.L. and Y.H. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. Location of the BTH.

Figure 2. Schematic diagram of the research method.

Figure 3. Trends of RHI development in BTH during 2001−2020. (a) Trends of RHI development in BTH; (b) Significance of RHI development. BJ: Beijing; TJ: Tianjin; SJZ: Shijiazhuang; TS: Tangshan; BD: Baoding; CZ: Cangzhou; HD: Handan; HS: Hengshui; LF: Langfang; XT: Xingtai; QHD: Qinhuangdao; CD: Chengde; and ZJK: Zhangjiakou.

Figure 4. The Standard Deviation Ellipse of high-temperature zones (RLST > 2) for 2001, 2005, 2010, 2015 and 2020. BJ: Beijing; TJ: Tianjin; SJZ: Shijiazhuang; TS: Tangshan; BD: Baoding; CZ: Cangzhou; HD: Handan; HS: Hengshui; LF: Langfang; XT: Xingtai; QHD: Qinhuangdao; CD: Chengde; and ZJK: Zhangjiakou.

Figure 5. Spatial distribution of area transfer degree of RHI at three levels.

Figure 6. Annual area proportion transfer of RHIs.

Figure 7. Global Moran’s I of RLST in the BTH from 2001 to 2020.

Figure 8. Spatial clusters and spatial outliers of RLST from 2001 to 2020. BJ: Beijing; TJ: Tianjin; SJZ: Shijiazhuang; TS: Tangshan; BD: Baoding; CZ: Cangzhou; HD: Handan; HS: Hengshui; LF: Langfang; XT: Xingtai; QHD: Qinhuangdao; CD: Chengde; and ZJK: Zhangjiakou.

Table 1

The socioeconomic indicators of the BTH and 13 cities.

City	Population(Million)	GDP(Billion Yuan)	Built-Up Area(Square Kilometer)
Beijing	13.87	3537.1	1469
Tianjin	10.87	1410.4	1151
Shijiazhuang	10.49	581	309
Tangshan	7.57	689	249
Qinhuangdao	3.01	161.2	142
Handan	10.59	348.6	188
Xingtai	7.99	212	108
Baoding	12.11	377.2	199
Zhangjiakou	4.65	155.1	101
Chengde	3.82	147.1	78
Cangzhou	7.85	358.8	87
Langfang	4.81	319.6	71
Hengshui	4.57	150.5	76
BTH	102.2	8447.6	4228

Table 2

The RHI levels and degree of changes.

Degree		Value	Changes
Large increase		2	I → III
Little/No change	Little increase	1	I → II; II → III
	No change	0	Same RHI category
	Little decrease	−1	III → II; II → I
Large decrease		−2	III → I

Table 3

Documentation for remotely sensed LST of the BTH from 2001 to 2020.

	Remotely Sensed LST (°C)
	Tmax	Tmin	Tmean	StdDev	CV (%)
Baoding	39.4	19.2	30.9	2.3	7.45
Beijing	40.9	18.7	30.3	3.1	10.25
Cangzhou	38.7	24.0	32.5	1.2	3.59
Chengde	39.1	18.9	27.4	2.1	7.72
Handan	44.1	23.5	32.1	1.9	5.73
Hengshui	38.4	25.8	32.0	0.9	2.74
Langfang	42.4	28.8	32.6	1.2	3.57
Qinhuangdao	37.8	20.2	29.9	1.8	6.02
Shijiazhuang	40.8	20.6	31.8	2.2	7.02
Tangshan	39.5	24.6	31.0	1.7	5.47
Tianjin	40.4	24.1	31.5	1.9	6.06
Xingtai	42.4	23.2	32.2	1.8	5.46
Zhangjiakou	44.5	15.9	30.2	3.2	10.42

Table 4

The parameters of the directional distribution (standard deviation ellipse) analysis of high-temperature zones (RLST > 2 °C) for 2001, 2005, 2010, 2015 and 2020 (XstdDist: X-axis standard deviation distance; and YStdDist: Y-axis standard deviation distance).

	CenterX	CenterY	XStdDist (km)	YStdDist (km)	Rotation
2001	854,655.12	4,833,581.89	138.74	242.49	171.39
2005	868,252.05	4,716,504.20	126.57	211.72	177.28
2010	872,052.04	4,789,658.19	147.69	236.72	178.58
2015	908,202.66	4,680,562.95	111.68	197.85	20.15
2020	930,401.48	4,699,935.44	109.34	217.52	30.46

References

1. Foley, J.A.; DeFries, R.; Asner, G.P.; Barford, C.; Bonan, G.; Carpenter, S.R.; Chapin, F.S.; Coe, M.T.; Daily, G.C.; Gibbs, H.K. et al. Global Consequences of Land Use. Science; 2005; 309, pp. 570-574. [DOI: https://dx.doi.org/10.1126/science.1111772]

2. Song, X.-P.; Hansen, M.C.; Stehman, S.V.; Potapov, P.V.; Tyukavina, A.; Vermote, E.F.; Townshend, J.R. Global land change from 1982 to 2016. Nature; 2018; 560, pp. 639-643. [DOI: https://dx.doi.org/10.1038/s41586-018-0411-9]

3. Zhao, L.; Lee, X.; Smith, R.B.; Oleson, K. Strong contributions of local background climate to urban heat islands. Nature; 2014; 511, pp. 216-219. [DOI: https://dx.doi.org/10.1038/nature13462]

4. Jones, P.; Lister, D.; Li, Q. Urbanization effects in large-scale temperature records, with an emphasis on China. J. Geophys. Res. Atmos.; 2008; 113, D16. [DOI: https://dx.doi.org/10.1029/2008JD009916]

5. Ren, G.; Li, J.; Ren, Y.; Chu, Z.; Zhang, A.; Zhou, Y.; Zhang, L.; Zhang, Y.; Bian, T. An integrated procedure to determine a reference station network for evaluating and adjusting urban bias in surface air temperature data. J. Appl. Meteorol. Climatol.; 2015; 54, pp. 1248-1266. [DOI: https://dx.doi.org/10.1175/JAMC-D-14-0295.1]

6. Qiu, G.Y.; Zou, Z.; Li, X.; Li, H.; Guo, Q.; Yan, C.; Tan, S. Experimental studies on the effects of green space and evapotranspiration on urban heat island in a subtropical megacity in China. Habitat Int.; 2017; 68, pp. 30-42. [DOI: https://dx.doi.org/10.1016/j.habitatint.2017.07.009]

7. Chapman, S.; Watson, J.E.; Salazar, A.; Thatcher, M.; McAlpine, C.A. The impact of urbanization and climate change on urban temperatures: A systematic review. Landsc. Ecol.; 2017; 32, pp. 1921-1935. [DOI: https://dx.doi.org/10.1007/s10980-017-0561-4]

8. Coumou, D.; Robinson, A. Historic and future increase in the global land area affected by monthly heat extremes. Environ. Res. Lett.; 2013; 8, 034018. [DOI: https://dx.doi.org/10.1088/1748-9326/8/3/034018]

9. Seneviratne, S.I.; Lüthi, D.; Litschi, M.; Schär, C.J.N. Land–atmosphere coupling and climate change in Europe. Nature; 2006; 443, pp. 205-209. [DOI: https://dx.doi.org/10.1038/nature05095]

10. Robaa, S. Urban-suburban/rural differences over Greater Cairo, Egypt. Atmósfera; 2003; 16, pp. 157-171.

11. Cheval, S.; Popa, A.-M.; Șandric, I.; Iojă, I.C. Exploratory analysis of cooling effect of urban lakes on land surface temperature in Bucharest (Romania) using Landsat imagery. Urban Clim.; 2020; 34, 100696. [DOI: https://dx.doi.org/10.1016/j.uclim.2020.100696]

12. Oke, T.R. Boundary Layer Climates; Routledge: Oxfordshire, UK, 2002.

13. Lin, Y.; Liu, A.; Ma, E.; Li, X.; Shi, Q. Impacts of future urban expansion on regional climate in the Northeast Megalopolis, USA. Adv. Meteorol.; 2013; 2013, 362925. [DOI: https://dx.doi.org/10.1155/2013/362925]

14. Chen, Y.; Yu, S. Impacts of urban landscape patterns on urban thermal variations in Guangzhou, China. Int. J. Appl. Earth Obs. Geoinf.; 2017; 54, pp. 65-71. [DOI: https://dx.doi.org/10.1016/j.jag.2016.09.007]

15. Foissard, X.; Dubreuil, V.; Quénol, H. Defining scales of the land use effect to map the urban heat island in a mid-size European city: Rennes (France). Urban Clim.; 2019; 29, 100490. [DOI: https://dx.doi.org/10.1016/j.uclim.2019.100490]

16. Santamouris, M.; Haddad, S.; Saliari, M.; Vasilakopoulou, K.; Synnefa, A.; Paolini, R.; Ulpiani, G.; Garshasbi, S.; Fiorito, F. On the energy impact of urban heat island in Sydney: Climate and energy potential of mitigation technologies. Energy Build.; 2018; 166, pp. 154-164. [DOI: https://dx.doi.org/10.1016/j.enbuild.2018.02.007]

17. Stone, B.; Vargo, J.; Habeeb, D. Managing climate change in cities: Will climate action plans work?. Landsc. Urban Plan.; 2012; 107, pp. 263-271. [DOI: https://dx.doi.org/10.1016/j.landurbplan.2012.05.014]

18. Carter, J.G.; Cavan, G.; Connelly, A.; Guy, S.; Handley, J.; Kazmierczak, A. Climate change and the city: Building capacity for urban adaptation. Prog. Plan.; 2015; 95, pp. 1-66. [DOI: https://dx.doi.org/10.1016/j.progress.2013.08.001]

19. Limaye, V.S.; Vargo, J.; Harkey, M.; Holloway, T.; Patz, J.A. Climate change and heat-related excess mortality in the Eastern USA. EcoHealth; 2018; 15, pp. 485-496. [DOI: https://dx.doi.org/10.1007/s10393-018-1363-0]

20. Rihan, M.; Naikoo, M.W.; Ali, M.A.; Usmani, T.M.; Rahman, A. Urban Heat Island Dynamics in Response to Land-Use/Land-Cover Change in the Coastal City of Mumbai. J. Indian Soc. Remote Sens.; 2021; 49, pp. 2227-2247.

21. Dewan, A.; Kiselev, G.; Botje, D.; Mahmud, G.I.; Bhuian, H.; Hassan, Q.K. Surface urban heat island intensity in five major cities of Bangladesh: Patterns, drivers and trends. Sustain. Cities Soc.; 2021; 71, 102926. [DOI: https://dx.doi.org/10.1016/j.scs.2021.102926]

22. Xiang, Z.; Qin, H.; He, B.-J.; Han, G.; Chen, M. Heat vulnerability caused by physical and social conditions in a mountainous megacity of Chongqing, China. Sustain. Cities Soc.; 2022; 80, 103792. [DOI: https://dx.doi.org/10.1016/j.scs.2022.103792]

23. Churkina, G.; Kuik, F.; Bonn, B.; Lauer, A.; Grote, R.; Tomiak, K.; Butler, T.M. Effect of VOC emissions from vegetation on air quality in Berlin during a heatwave. Environ. Sci. Technol.; 2017; 51, pp. 6120-6130. [DOI: https://dx.doi.org/10.1021/acs.est.6b06514] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28513175]

24. Santamouris, M.; Kolokotsa, D.J.E. On the impact of urban overheating and extreme climatic conditions on housing, energy, comfort and environmental quality of vulnerable population in Europe. Energy Build.; 2015; 98, pp. 125-133. [DOI: https://dx.doi.org/10.1016/j.enbuild.2014.08.050]

25. Sun, Y.; Zhang, X.; Zwiers, F.W.; Song, L.; Wan, H.; Hu, T.; Yin, H.; Ren, G. Rapid increase in the risk of extreme summer heat in Eastern China. Nat. Clim. Change; 2014; 4, pp. 1082-1085. [DOI: https://dx.doi.org/10.1038/nclimate2410]

26. Yu, Z.; Xu, S.; Zhang, Y.; Jorgensen, G.; Vejre, H. Strong contributions of local background climate to the cooling effect of urban green vegetation. Sci. Rep.; 2018; 8, pp. 1-9. [DOI: https://dx.doi.org/10.1038/s41598-018-25296-w]

27. Anderson, G.B.; Bell, M.L. Heat waves in the United States: Mortality risk during heat waves and effect modification by heat wave characteristics in 43 US communities. Environ. Health Perspect.; 2011; 119, pp. 210-218. [DOI: https://dx.doi.org/10.1289/ehp.1002313]

28. Montero, J.; Mirón, I.; Criado-Álvarez, J.; Diaz, J. Influence of local factors in the relationship between mortality and heat waves: Castile-La Mancha (1975–2003). Sci. Total Environ.; 2012; 414, pp. 73-80. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2011.10.009]

29. Tan, J.; Zheng, Y.; Tang, X.; Guo, C.; Li, L.; Song, G.; Zhen, X.; Yuan, D.; Kakstein, A.J.; Li, F. et al. The urban heat island and its impact on heat waves and human health in Shanghai. Int. J. Biometeorol.; 2010; 54, pp. 75-84. [DOI: https://dx.doi.org/10.1007/s00484-009-0256-x]

30. Rinner, C.; Hussain, M. Toronto’s urban heat island—Exploring the relationship between land use and surface temperature. Remote Sens.; 2011; 3, pp. 1251-1265. [DOI: https://dx.doi.org/10.3390/rs3061251]

31. Brazel, A.; Gober, P.; Lee, S.-J.; Grossman-Clarke, S.; Zehnder, J.; Hedquist, B.; Comparri, E. Determinants of changes in the regional urban heat island in metropolitan Phoenix (Arizona, USA) between 1990 and 2004. Clim. Res.; 2007; 33, pp. 171-182. [DOI: https://dx.doi.org/10.3354/cr033171]

32. Sun, R.; Lü, Y.; Yang, X.; Chen, L. Understanding the variability of urban heat islands from local background climate and urbanization. J. Clean. Prod.; 2019; 208, pp. 743-752. [DOI: https://dx.doi.org/10.1016/j.jclepro.2018.10.178]

33. Carneiro, E.; Lopes, W.; Espindola, G. Linking Urban Sprawl and Surface Urban Heat Island in the Teresina–Timon Conurbation Area in Brazil. Land; 2021; 10, 516. [DOI: https://dx.doi.org/10.3390/land10050516]

34. Anniballe, R.; Bonafoni, S.; Pichierri, M. Spatial and temporal trends of the surface and air heat island over Milan using MODIS data. Remote Sens. Environ.; 2014; 150, pp. 163-171. [DOI: https://dx.doi.org/10.1016/j.rse.2014.05.005]

35. Kaplan, G.; Avdan, U.; Avdan, Z.Y. Urban heat island analysis using the landsat 8 satellite data: A case study in Skopje, Macedonia. Proceeding; 2018; 2, 358.

36. El Kenawy, A.M.; Hereher, M.; Robaa, S.M.; McCabe, M.F.; Lopez-Moreno, J.I.; Dominguez-Castro, F.; Gaber, I.M.; Al-Awadhi, T.; Al-Buloshi, A.; Al Nasiri, N. et al. Nocturnal surface urban heat island over Greater Cairo: Spatial morphology, temporal trends and links to land-atmosphere influences. Remote Sens.; 2020; 12, 3889. [DOI: https://dx.doi.org/10.3390/rs12233889]

37. Abrar, R.; Sarkar, S.K.; Nishtha, K.T.; Talukdar, S.; Shahfahad,; Rahman, A.; Towfiqul Islam, A.R.; Mosavi, A. Assessing the Spatial Mapping of Heat Vulnerability under Urban Heat Island (UHI) Effect in the Dhaka Metropolitan Area. Sustainability; 2022; 14, 4945. [DOI: https://dx.doi.org/10.3390/su14094945]

38. Dutta, I.; Das, A. Exploring the Spatio-temporal pattern of regional heat island (RHI) in an urban agglomeration of secondary cities in Eastern India. Urban Clim.; 2020; 34, 100679. [DOI: https://dx.doi.org/10.1016/j.uclim.2020.100679]

39. Masoudi, M.; Tan, P.Y. Multi-year comparison of the effects of spatial pattern of urban green spaces on urban land surface temperature. Landsc. Urban Plan.; 2019; 184, pp. 44-58. [DOI: https://dx.doi.org/10.1016/j.landurbplan.2018.10.023]

40. Estoque, R.C.; Murayama, Y.; Myint, S.W. Effects of landscape composition and pattern on land surface temperature: An urban heat island study in the megacities of Southeast Asia. Sci. Total Environ.; 2017; 577, pp. 349-359. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2016.10.195]

41. Keeratikasikorn, C.; Bonafoni, S. Satellite images and Gaussian parameterization for an extensive analysis of urban heat islands in Thailand. Remote Sens.; 2018; 10, 665. [DOI: https://dx.doi.org/10.3390/rs10050665]

42. Tran, H.; Uchihama, D.; Ochi, S.; Yasuoka, Y. Assessment with satellite data of the urban heat island effects in Asian mega cities. Int. J. Appl. Earth Obs. Geoinf.; 2006; 8, pp. 34-48. [DOI: https://dx.doi.org/10.1016/j.jag.2005.05.003]

43. Yang, Q.; Huang, X.; Tang, Q. The footprint of urban heat island effect in 302 Chinese cities: Temporal trends and associated factors. Sci. Total Environ.; 2019; 655, pp. 652-662. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2018.11.171]

44. Wang, J.; Huang, B.; Fu, D.; Atkinson, P.M. Spatiotemporal variation in surface urban heat island intensity and associated determinants across major Chinese cities. Remote Sens.; 2015; 7, pp. 3670-3689. [DOI: https://dx.doi.org/10.3390/rs70403670]

45. Fang, C.; Zhou, C.; Gu, C.; Chen, L.; Li, S. A proposal for the theoretical analysis of the interactive coupled effects between urbanization and the eco-environment in mega-urban agglomerations. J. Geogr. Sci.; 2017; 27, pp. 1443-1449. [DOI: https://dx.doi.org/10.1007/s11442-017-1445-x]

46. Wu, J. Urban ecology and sustainability: The state-of-the-science and future directions. Landsc. Urban Plan.; 2014; 125, pp. 209-221. [DOI: https://dx.doi.org/10.1016/j.landurbplan.2014.01.018]

47. Zhou, D.; Bonafoni, S.; Zhang, L.; Wang, R. Remote sensing of the urban heat island effect in a highly populated urban agglomeration area in East China. Sci. Total Environ.; 2018; 628, pp. 415-429. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2018.02.074] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29448025]

48. Du, H.; Wang, D.; Wang, Y.; Zhao, X.; Qin, F.; Jiang, H.; Cai, Y. Influences of land cover types, meteorological conditions, anthropogenic heat and urban area on surface urban heat island in the Yangtze River Delta Urban Agglomeration. Sci. Total Environ.; 2016; 571, pp. 461-470. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2016.07.012] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27424113]

49. Zhou, D.; Li, D.; Sun, G.; Zhang, L.; Liu, Y.; Hao, L. Contrasting effects of urbanization and agriculture on surface temperature in eastern China. J. Geophys. Res. Atmos.; 2016; 121, pp. 9597-9606. [DOI: https://dx.doi.org/10.1002/2016JD025359]

50. Yu, Z.; Yao, Y.; Yang, G.; Wang, G.; Vejre, H. Spatiotemporal patterns and characteristics of remotely sensed region heat islands during the rapid urbanization (1995–2015) of Southern China. Sci. Total Environ.; 2019; 674, pp. 242-254. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2019.04.088]

51. Wang, J.; Huang, B.; Fu, D.; Atkinson, P.M.; Zhang, X. Response of urban heat island to future urban expansion over the Beijing–Tianjin–Hebei metropolitan area. Appl. Geogr.; 2016; 70, pp. 26-36. [DOI: https://dx.doi.org/10.1016/j.apgeog.2016.02.010]

52. Wang, E.; Nakagoshi, N. Urban heat island change in the Beijing–Tianjin–Hebei metropolitan area driven by land cover change. J. Meteorol. Soc. Jpn.; 2010; 47, pp. 77-89.

53. Liu, F.; Hou, H.; Murayama, Y. Spatial interconnections of land surface temperatures with land cover/use: A case study of Tokyo. Remote Sens.; 2021; 13, 610. [DOI: https://dx.doi.org/10.3390/rs13040610]

54. Jones, P.D.; Lister, D.H. The urban heat island in Central London and urban-related warming trends in Central London since 1900. Weather; 2009; 64, pp. 323-327. [DOI: https://dx.doi.org/10.1002/wea.432]

55. Hardin, A.; Liu, Y.; Cao, G.; Vanos, J. Urban heat island intensity and spatial variability by synoptic weather type in the northeast US. Urban Clim.; 2018; 24, pp. 747-762. [DOI: https://dx.doi.org/10.1016/j.uclim.2017.09.001]

56. Lowe, S.A. An energy and mortality impact assessment of the urban heat island in the US. Environ. Impact Assess. Rev.; 2016; 56, pp. 139-144. [DOI: https://dx.doi.org/10.1016/j.eiar.2015.10.004]

57. Li, M.; Wang, T.; Xie, M.; Zhuang, B.; Li, S.; Han, Y.; Cheng, N. Modeling of urban heat island and its impacts on thermal circulations in the Beijing–Tianjin–Hebei region, China. Theor. Appl. Climatol.; 2017; 128, pp. 999-1013. [DOI: https://dx.doi.org/10.1007/s00704-016-1903-x]

58. Hou, L.; Yue, W.; Liu, X. Spatiotemporal patterns and drivers of summer heat island in Beijing-Tianjin-Hebei Urban Agglomeration, China. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens.; 2021; 14, pp. 7516-7527. [DOI: https://dx.doi.org/10.1109/JSTARS.2021.3094559]

59. Liu, Y.; Fang, X.; Xu, Y.; Zhang, S.; Luan, Q. Assessment of surface urban heat island across China’s three main urban agglomerations. Theor. Appl. Climatol.; 2018; 133, pp. 473-488. [DOI: https://dx.doi.org/10.1007/s00704-017-2197-3]

60. Zhao, G.; Dong, J.; Liu, J.; Dong, J.; Liu, J.; Zhai, J.; Cui, Y.; He, T.; Xiao, X. Different patterns in daytime and nighttime thermal effects of urbanization in Beijing-Tianjin-Hebei urban agglomeration. Remote Sens.; 2017; 9, 121. [DOI: https://dx.doi.org/10.3390/rs9020121]

61. Sining, C.; Jun, G. Natural Hazards, Risk. Climatic risks of Beijing–Tianjin–Hebei urban agglomeration and their changes. Geomat. Nat. Hazards Risk; 2021; 12, pp. 1298-1314. [DOI: https://dx.doi.org/10.1080/19475705.2021.1924296]

62. National Bureau of Statistics of China (NBSC). China Statistical Yearbook; China Statistics Press: Beijing, China, 2020.

63. Feng, Z.; Liu, D.; Zhang, Y. Water requirements and irrigation scheduling of spring maize using GIS and CropWat model in Beijing-Tianjin-Hebei region. Chin. Geogr. Sci.; 2007; 17, pp. 56-63. [DOI: https://dx.doi.org/10.1007/s11769-007-0056-3]

64. Sun, R.; Chen, L. Effects of green space dynamics on urban heat islands: Mitigation and diversification. Ecosyst. Serv.; 2017; 23, pp. 38-46. [DOI: https://dx.doi.org/10.1016/j.ecoser.2016.11.011]

65. Furfey, P.H. A note on Lefever’s “standard deviational ellipse”. Am. J. Sociol.; 1927; 33, pp. 94-98. [DOI: https://dx.doi.org/10.1086/214336]

66. Zhao, L.; Zhao, Z.Q. Spatial Agglomeration of the manufacturing industry in China. J. Quant. Tech. Econ.; 2014; 31, pp. 110-121.

67. Zhao, L.; Zhao, Z.Q. Projecting the spatial variation of economic based on the specific ellipses in China. Sci. Geogr. Sin.; 2014; 34, pp. 979-986.

68. Wang, Y.; Yao, Y.; Chen, S.; Ni, Z.; Xia, B. Spatiotemporal evolution of urban development and surface urban heat island in Guangdong-Hong Kong-Macau greater bay area of China from 2013 to 2019. Resour. Conserv. Recycl.; 2022; 179, 106063. [DOI: https://dx.doi.org/10.1016/j.resconrec.2021.106063]

69. Tobler, W.R. A computer movie simulating urban growth in the Detroit region. Econ. Geogr.; 1970; 46, pp. 234-240. [DOI: https://dx.doi.org/10.2307/143141]

70. Guo, A.; Yang, J.; Xiao, X.; Xia, J.; Jin, C.; Li, X. Influences of urban spatial form on urban heat island effects at the community level in China. Sustain. Cities Soc.; 2020; 53, 101972. [DOI: https://dx.doi.org/10.1016/j.scs.2019.101972]

71. Brown, C.E. Coefficient of variation. Applied Multivariate Statistics in Geohydrology and Related Sciences; Springer: Berlin/Heidelberg, Germany, 1998; pp. 155-157.

72. Rao, S.; Zhang, H.; Jin, T.; Dou, H.Y. The spatial character of regional heat island in Pearl River Delta using MODIS remote sensing data. Geogr. Res.; 2010; 29, pp. 127-136.