Evolution Analysis of the Coupling Coordination of Microclimate and Landscape Ecological Risk Degree in the Xiahuayuan District in Recent 20 Years

Abstract

Understanding the degree of interaction between microclimate and landscape risk in urban development is essential. This study analyzed the degree of interaction between microclimate and landscape ecological risk in 2000, 2010, and 2020 in the Xiahuayuan District in Zhangjiakou City, Hebei Province, China, using a coupled coordination degree model. The results show that the degree of landscape ecological risk in the Xiahuayuan District from 2000 to 2020 was mainly medium-high and high risk; the risk transfer area was 69.75 km² and 107.76 km² from 2000 to 2010 and 2010 to 2020, respectively. The surface temperature gradually decreased from west to east; the area of the middle temperate zone suitable for human habitation in 2000, 2010 and 2020 was 42.96%, 36.03% and 47.05%, respectively. The landscape ecological risk degree and surface temperature were closely related during the study period and interacted significantly. The coupling degree and coordination degree were dominated by high coupling degree and mutual coordination. The area of high coupling degree in 2000, 2010 and 2020 accounted for 79.53%, 78.07% and 85.06%, respectively; the area of mutual coordination degree accounted for 78.80%, 80.97% and 83.13%, respectively. The interaction between landscape ecological risk degree and surface temperature in the Xiahuayuan District was more evident, with strong coupling coordination.

1. Introduction

Resource-exhausted cities are integral to China’s urban development. Since 2008, China has established three batches of resource-exhausted cities. These cities continue to export a large number of resources during periods of abundant natural resources. The ecology, environment, health and climate in the region have declined due to a troublesome change in the type of underlying surface. Natural resources are not inexhaustible. With a lack of natural resources, climate and landscape ecology problems caused by the change in underlying surface types still exist. To improve the industrial structure of resource-exhausted cities, continually advance regional urbanization and rationally construct an ecological civilization, the Chinese government has proposed the “sustainable development” transformation strategy.

As resource-exhausted cities have steadily developed, artificial surfaces have become the core element of regional surface types. The urban microclimate has witnessed significant changes due to deforestation, water pollution and emission of toxic gases. The land surface temperature (LST), an indicator of the typical regional climate, can better reflect the distribution of a regional microclimate. In recent years, scholars worldwide have determinedly explored the factors associated with land surface temperature (LST), such as Ramaiah et al. [1], who explored the correlation between LST and the Enhanced Built-up and Bareness Index, Modified Normalized Difference Water Index, and Soil Adjusted Vegetation Indexes, and found that both urban green spaces and water bodies could reduce LST. In another study, Hu et al. [2] associated LST with three types of urban centers and explored the changing LST laws in the different urban center types. Guo et al. [3] adopted an integration of the spatial error, spatial lag and ordinary least squares models to explore the influence of spatial factors on LST. Meanwhile, Qiao et al. [4] argue that the rise and fall of urban LST may be controlled by changes in population, building and landscape, suggesting urban evolution can influence the regional microclimate.

As urban microclimates change, humans must make changes to production and life methods to facilitate sustainability. At the same time, natural ecosystems have come under pressure from human activities [5], causing continuous changes in the landscape ecological risk degree. Consequently, through urban landscape ecological risk and scientific management, protecting urban areas from ecological risks can be achieved. Research on the factors influencing the landscape ecological risk degree both in China and abroad has made gradual progress, focusing on different aspects of urban expansion [6,7,8,9], landscape patterns [10,11] and land use [12]. For instance, Mo et al. [13] assert that the spatial correlation between road expansion and urban ecological risk level makes their changes consistent, while Lina et al. [14] report that the spatial distribution of ecological risk in Northeast China is closely related to the landscape pattern. Furthermore, Wang et al. [15] concluded that changes in land use cover can influence ecological risk levels.

Various driving factors continuously affect the urban microclimate and landscape ecological risk represented by surface temperature. Previous studies have shown that alleviating two factors of the driving force has a common place but also a different one. The lack of research on coupling, coordination and the correlation between the two, leads to a reduction in the connectivity between their driving factors. In this study, we used Landsat image series and land-use data to integrate landscape risk degree assessment and LST, and built a coupling coordination degree model to provide reference for minimizing landscape ecological risk and LST change.

2. Research Region and Research Method

2.1. Profile of the Research Region

The selected study area, the Xiahuayuan District (Figure 1) in Zhangjiakou City, Hebei Province, is located at 115°16′ E and 40°29′ N. This region has a temperate continental monsoon climate, with four distinct seasons, and hosts abundant mineral resources, including reserves of coal, shale, granite and other resources used throughout the territory. The region also hosts ample symbiotic mineral resources, including hematite and multi-metals. By March 2009, the Xiahuayuan District, owing to its long-time singular industrial structure, inadequate economic aggregates and serious resource shortages, was classified by the State Council of China as a second batch resource-exhausted city. Since 2000, the Xiahuayuan District has provided continuous resource support for coal-related industries, including electricity, building materials, chemicals and metallurgy. Considerable manpower has focused on the coal mining industry, bringing great changes to the underlying surfaces in the region and significantly impacting landscape ecological risk and LST. With the depletion of mineral resources and other resources, the economic growth of the Xiahuayuan District slowed, and the high-intensity destruction of the ecological environment during resource development had a great impact on the regional climate and landscape components. Therefore, the Xiahuayuan District has representative significance as the study area.

2.2. Research Methods

a.

Landscape risk evaluation unit

In this study, the grid is used to sample land-use data, which is used as the landscape ecological risk assessment unit in the study area. Early researchers determined the grid region should be 2–5 times the average region of the patch size [16,17,18]. According to data of the Xiahuayuan District for 2000, 2010 and 2020, the average patch size of the landscape was 0.17, 0.18 and 0.19 km², respectively. Based on the data resolution and accuracy, the research region comprises 800 m × 800 m grids, with a total of 568 evaluation units.

b.

Calculation of landscape ecological risk index

This study used the landscape risk index (Equation (1)) that reflects the comprehensive relative loss of different landscape types within the evaluated grid unit, which is mainly composed of the landscape loss degree index (Equation (2)) [19,20]. The landscape loss index expresses the loss degree of ecosystem attributes when the natural ecosystem is disturbed, which is mainly composed of the landscape vulnerability index and the landscape structure index (Equation (3)). The former shows the degree to which the landscape is affected by the external environment; the degree of vulnerability of landscape under human influence is indirectly expressed as the degree of perceived stress of landscape components. For this study, the landscape vulnerability index was determined based on previous experience and actual research, as follows: artificial land surface (1), woodland (2), grassland (3), cultivated land (4) and water bodies (5), and the values were normalized to obtain the landscape vulnerability index of each land type [21]. The landscape structure index is a weighted representation of the degree of landscape separation (Equation (4)), dominance degree (Equation (5)) and fragmentation degree (Equation (6)) [22]. These indicate the degree of separation, importance and fragmentation of landscape patches, respectively. Equations (1)–(6) are described below.

$$ER{I}_k={\displaystyle \sum _{i=1}^n\frac{A_{ki}}{A_k}}{T}_i,$$

(1)

where $ER{I}_k$ is the landscape ecological risk degree index; $A_{ki}$ refers to the area of each landscape type in the kth evaluation unit (ha); $A_k$ refers to the total region of the landscape type in the evaluation unit (ha); and $T_i$ refers to the index of landscape loss degree.

$$T_i=\sqrt{S_i\cdot W_i},$$

(2)

where $S_i$ refers to the landscape structure index, and $W_i$ refers to the landscape vulnerability index.

$$S_i=x{C}_i+y{D}_i+z{N}_i,$$

(3)

where $C_i$ refers to the fragmentation index, $D_i$ refers to the landscape dominance index and $N_i$ refers to the degree of landscape separation. Based on previous research and the actual situation in the research region, x, y and z are taken as 0.5, 0.3 and 0.2, respectively [23].

$$N_i=\frac{1}{2}\sqrt{\frac{n_i}{A}}+A/A_i,$$

(4)

$$D_i=\frac{Q_i+M_i}{4}+\frac{L_i}{2},$$

(5)

$$C_i=n_i/A_i,$$

(6)

where $n_i$ refers to the number of patches in the landscape in the evaluation unit; $Q_i$ refers to the proportion of the number of grids on the patch to the total grid amounts; $M_i$ refers to the proportion of the patch amount in the total patch amounts; and $L_i$ refers to the proportion of the area of the patches to the area of the evaluation unit.

c.

Land surface temperature (LST) inversion

Several methods of LST inversion have been reported for Landsat data, including the atmospheric correction method, single window algorithm and split window algorithm. In this study, the atmospheric correction method was used to calculate the LST of Landsat data, wherein the Landsat TM data use the sixth band and the Landsat OLI data adopt the tenth band [24,25,26]. The following operations were conducted on the preprocessed original images:

First, the radiance $L_b$ was calculated as follows:

$$L_b=G_z\times DN+B_p,$$

(7)

where $G_z$ is the slope corresponding to the response function, $B_p$ is the intercept corresponding to the response function, and $DN$ is the gray value of the pixels in the thermal infrared band.

Second, the blackbody radiance $T_a$ was calculated:

$$T_a=\frac{L_{\lambda }-L_s-p(1-\epsilon )L_x}{p\times \epsilon },$$

(8)

where $L_{\lambda }$ is the thermal infrared radiance (units in $\mathrm{W}\cdot {\mathrm{m}}^{-2}\cdot \mathrm{s}{\mathrm{r}}^{-1}\cdot \mathsf{\mu }{\mathrm{m}}^{-1}$); $p$ is the atmospheric transmittance in the thermal infrared band; $L_s$ and $L_x$ indicate the upward and downward radiance of the atmosphere, respectively (units in $\mathrm{W}\cdot {\mathrm{m}}^{-2}\cdot \mathrm{s}{\mathrm{r}}^{-1}\cdot \mathsf{\mu }{\mathrm{m}}^{-1}$); and $\epsilon$ is the atmospheric transmittance.

Finally, the LST was calculated:

$$LST=\frac{K_2}{\ln (K_1/T_a+1)}-273,$$

(9)

where $K_1$ and $K_2$ are constants.

d.

Coupling coordination degree model

The relationship model between variables has two important factors: coordination and evolution. Evolution describes the change from disorder to order between variables, and coordination describes the process in which variables complement and promote each other. The coupling coordination degree model describes the coordination of evolution level between variables. At present, it is widely used in geography, economics, management and other disciplines [27,28,29,30]. Therefore, we introduced the coupling coordination degree model to explore the coupling coordination degree between the elements studied herein. The proposed model was used to comprehensively analyze the interactions and coordination between landscape ecological risk and LST. The model calculation method is shown in Equations (10)–(12) as follows [31,32]:

First, a comprehensive coordination index was built. It is necessary to determine the comprehensive coordination coefficient of landscape ecological risk index and LST. In this study, they were considered to have the same effect, so they are both 0.5.

$$T=\alpha ERI+\beta LST,$$

(10)

where $ERI$ is the landscape ecological risk index and $LST$ is the land surface temperature.

Then, the coupling degree model of surface temperature and landscape ecological risk was constructed:

$$C=\sqrt{\frac{ERI\times LST}{{\left(\frac{ERI+LST}{2}\right)}^2}},$$

(11)

Because the amplitude change of the coupling degree model was inconsistent with the comprehensive coordination index, the coupling coordination degree model was finally constructed:

$$D=\sqrt{C\times T},$$

(12)

where $D$ refers to the coupling coordination degree, $C$ refers to the coupling degree and $T$ is the comprehensive coordination index.

2.3. Data Sources

The Landsat remote sensing data, including the original images of the summers of 2000, 2010 and 2020, are available at the official database websites of the United States Geological Survey (http://glovis.usgs.gov/, accessed on 1 October 2021, considering the cloud cover and data availability, we choose the data of 2001 and 2009.) and Geospatial Data Cloud (http://www.gscloud.cn/, accessed on 1 October 2021). The original images can be preprocessed for radiometric calibration and atmospheric correction. After removing systematic errors, surface temperature inversion can be carried out so that a more accurate surface temperature value can be obtained. Land use classification data for the years 2000, 2010 and 2020 (http://www.globallandcover.com/, accessed on 1 October 2021) was prepared and released by the Ministry of Natural Resources of the People’s Republic of China. The land use data of the research region were cropped to obtain the target land use types of cultivated land, woodland, grassland, shrubland, wetland, water bodies and artificial surfaces. According to the research requirements, the natural conditions of the research region and the relevant regulations of China’s Land Use Classification (GB/T21010-2017) classified the shrubland into forestland and classified the wetlands into water bodies. The above operation used the reclassification function of ArcGIS 10.7.

3. Results

3.1. Analysis of the Spatial–Temporal Pattern Evolution of Landscape Ecological Risk Degree

Figure 2 shows the time series and spatial distribution characteristics of landscape ecological risk in the Xiahuayuan District. On the whole, high-risk areas are mainly distributed in the north and less in the south. The medium-low and low-risk categories were mainly distributed in the eastern and southern regions. The regional proportions of low risk, medium risk and high risk steadily decreased, while the regional proportions of medium-high risk and medium-low risk gradually increased. In 2000, the coal industry witnessed a soaring development in the Xiahuayuan District, and the regions with medium-high and high levels of landscape ecological risk were significant due to over-exploitation. From 2010 to 2020, most coal industries were closed in response to national policy calls and resource exhaustion. During this time, the Xiahuayuan District steadily advanced its green economy. Due to the moderate land-use structure and the uniform distribution of blue and green spaces, the low-medium-risk regions gradually expanded, and the high-risk regions gradually decreased. The reason for the expansion of the area of medium and high-risk areas may be that the “denounce” left by mining in the area has not been completely removed, and the risk sharing of landscape ecology in the area still exists.

Landscape ecological risk is divided into five categories, as shown in Figure 2, which can imitate land-use type data to describe the transfer situation. As shown in Figure 3, the transfer regions for 2000–2010 and for 2010–2020 totaled 69.75 and 107.76 km², respectively, and can be described as follows. For 2000–2010, 7.94 km² of the low-risk region was transferred to the medium-low-risk region; 8.57 km² of the medium-low-risk region was transferred to the medium-risk region; 20.77 km² of the medium-risk region was transferred to the medium-high risk region; 6.89 km² of the medium-high risk region was transferred to the high-risk region and 0.00 km² of the high-risk region was transferred back to the low-risk region. For 2010–2020, 17.85 km² of the low-risk region was transferred to the medium-low-risk region; 11.55 km² of the medium-low-risk region was transferred to the medium-risk region; 33.86 km² of the medium-risk region was transferred to the medium-high risk region; 8.60 km² of the medium-high risk region was transferred to the high-risk region and 0.00 km² of the high-risk region was transferred back to low-risk region. Generally, the risk transfer degree in the latter period was relatively high, and there was no region of transfer from the high-risk region to the low-risk region within the research region.

3.2. Analysis of the Evolution of the Spatial–Temporal Pattern of LST

We used remote sensing data from 2000, 2010 and 2020 to calculate the LST of Landsat data and obtained the time series spatial distribution data of LST. The average value standard deviation method was used to classify the surface temperature (

Table 1. Grading Scale of Land Surface Temperature.

Year	Low-Temperature Region	Sub-Low-Temperature Region	Medium-Temperature Region	Sub-High-Temperature Region	High-Temperature Region
2000	14.95%	13.26%	42.69%	16.08%	13.02%
2010	15.92%	10.75%	36.03%	25.49%	11.81%
2020	14.51%	7.66%	47.05%	19.51%	11.27%

). As shown in Table 1, in the three research stages from 2000 to 2020, the proportion of low and medium-high temperature areas in the Xiahuayuan District shows an increase and then decrease in values. In addition, the proportion of medium- and low-temperature areas and high-temperature areas shows a downward trend. In contrast, the medium-temperature region showed values that changed from low to high. In the Xiahuayuan District, the proportion of overall LST in the medium-temperature regions was relatively large, indicating that the regional LST was relatively suitable with high habitability; meanwhile, the proportion of high-temperature regions continued to decrease and the LST was moderately adjusted.

Figure 4 shows the spatial distribution of LSTs for 2000–2020. In 2000, the LST decreased from the southwest to the northeast, with the eastern region reaching the lowest values. In 2010, due to the relatively low values of small settlements in the eastern and western areas, the LST displayed a relatively low distribution trend, unlike the other regions which saw a high distribution trend. In 2020, the LST reached the lowest value on average, with the highest value being 39.55 °C and the lowest value being 19.14 °C. The distribution trend was roughly the same as that in 2000. In general, the data of the three periods show that the distribution trend decreased from western to eastern areas, and that the LST in regions with higher artificial ground surface areas had a higher surface temperature, with lower surface temperatures near forest land and water bodies.

3.3. Coupling Degree Analysis

The coupling degree model of landscape ecological risk degree and LST was established as per the equation, and the coupling degree values of 568 risk areas in 2000, 2010 and 2020 were calculated. The natural breakpoint method was then used to divide the coupling degree into four coupled zones; the classification is shown in

Table 2. Coupling Degree Grading Scale.

Year	Low-Level Coupling	Antagonistic Coupling	Running-in Coupling	High-Level Coupling
2000	0.96%	1.20%	18.31%	79.53%
2010	0.96%	2.17%	18.80%	78.07%
2020	2.41%	1.93%	10.60%	85.06%

. Table 2 shows the proportion of the coupling degree of risk areas in each year. We can conclude from our analysis that as the proportion of low-level coupled zones between landscape ecological risk degree and LST increased, a slight fluctuation occurred. Generally, there were fewer low-level coupled zones and fewer zones with weaker interaction between the two factors. Compared with the low-level coupled zones, the proportions of antagonistic interactions between coupled zones in 2000 and 2010 were slightly higher, and the interaction between landscape ecological risk degree and LST was strengthened, with a correlation in the antagonistic stage. By 2020, the proportion of the antagonistic coupled zones had decreased; in 2010, the running-in coupled zone had the highest value. The landscape ecological risk degree and LST had a close correlation, which was in a constant running-in stage. The high-level coupled zone had the largest area proportion, which indicates that the landscape ecological risk degree and LST had a strong correlation and maintained a synchronized change trend. The proportion of high-level coupling in 2020 was the highest, indicating that a portion of the running-in coupled zones in 2010 had been transformed into high-level coupled zones. Based on this analysis, it can be stated that for the Xiahuayuan District, the year 2000 was the resource export year, 2010 was the initial transformation period and 2020 was the year that essentially completed the transformation. In 2000, economic development was relatively prosperous in the Xiahuayuan District, the underlying surface types were constantly being altered by the industry and the LST had changed; thus, landscape ecological risks also increased. As resource exhaustion continued, the underlying surface types transformed slowly. By 2010, the LST reached a phased peak and the ecological risks in some areas in the region decreased. Then, up to 2020, the Xiahuayuan District continued building an “ecological, livable, and beautiful” city, such that the LST and the landscape ecological risks decreased.

3.4. Coordination Analysis

The coordination degree model of landscape ecological risk degree and LST was constructed based on the equation and the coordination value of 568 risk areas in 2000, 2010 and 2020 was calculated. The natural breakpoint method was then used to classify the degree of coordination into five grades, as shown in

Table 3. Coordination Degree Grading Scale.

Year	Severe Imbalance	Moderate Imbalance	Basic Coordination	Moderate Coordination	Highly Coordinated
2000	7.95%	13.25%	27.47%	37.35%	13.98%
2010	8.19%	10.84%	26.02%	34.70%	20.25%
2020	7.47%	9.40%	29.16%	40.24%	13.73%

. Table 3 displays the proportion of the coordination degree of risk areas in each year; for the entire period of 2000–2020, about 8% of the landscape ecological risk degree and LST in the Xiahuayuan District failed to mutually promote, and they were mutually excluded and restricted. Moreover, compared with the severe imbalance regions, the proportion of moderately imbalanced areas was slightly higher. The landscape ecological risk degree and LST in these areas were mutually restrictive but not excluded. As per the proportion of the basic coordination regions, the analysis results suggest that the landscape ecological risk degree and LST are not apparently promoting each other, and the basic coordination region had the largest proportion in 2020. The moderate coordination regions had the largest proportion among the five grades. The landscape ecological risk degree and LST were mutually promoted, and the increase or decrease in the risk value was interconnected with the increase or decrease in the surface temperature; in the highly coordinated regions, the landscape ecological risk degree and LST complemented each other with significant mutual promotion. Generally, the imbalance proportion of the landscape ecological risk degree and LST in the Xiahuayuan District was relatively small, which mainly displayed the state of coordination, and the proportion of moderate and high coordination was particularly prominent, reaching more than 50% in three periods. The degree of coordination was relatively evenly distributed; in the early stage of the transition from 2000 to 2010, small- and medium-sized mining land turned into industrial wastelands with poor ecological environment management capacity. The degree of landscape ecological risk worsened as the LST constantly increased. During the transition period from 2010 to 2020, the concept of “ecological construction centered to promote green rising” in the Xiahuayuan District had been continuously implemented, reducing both the landscape ecological risk and LST and producing a highly coordinated level in most regions.

3.5. Coupling Coordination Analysis

The coupling coordination degree of each evaluation unit was calculated (Figure 5), Kriging interpolation was conducted on the data and ArcGIS 10.7 was utilized to implement a visual representation. The natural breakpoint method was adopted to divide the results into four categories of coupling coordination: low, moderate, high and extreme coupling coordination. In 2000, the degree of coupling coordination expanded from eastern to western regions, with extreme and high coupling coordination found in most of the research regions. In 2010, the extreme coupling coordination was distributed in the central region in a cluster, and the units with low and medium coupling coordination degrees were scattered among the evaluated units and distributed in the eastern forest area and some southern regions. The research region was characterized mostly by high and extreme coupling coordination, with both categories collectively accounting for over 80%. Both low and medium coupling coordination were distributed across smaller and sporadic portions of the eastern and southern areas. In 2000, mineral exploration was required by government expropriation of arable land, forestland and grassland, which became abandoned land after mining, causing an increase in LST and landscape ecological risk degree. At that time, high and extreme coupling coordination were distributed in most regions. In 2010, the mining industry slowed down considerably, with some waste land being unused. Furthermore, the government had failed to implement environmental and ecological management at this time. The “legacy” remaining from mining continued to erode the Xiahuayuan District’s landscape ecology and LST, which appeared to be mutually reinforcing. By 2020, the Xiahuayuan District was controlling the ecology with effective regional transformation efforts; both the LST and overall landscape ecological risk were reduced. Pursuant to the distribution trend, low-medium coupling coordination was distributed in the eastern forest area and part of the construction land area. This area features the same characteristics of distribution in an area with a weaker landscape ecological risk degree and lower LST. Nevertheless, the area displayed a distribution of low and medium coupling coordination, which may be due to the singular land-use structure in the construction land area and the large forest coverage area. These ecological problems affected the Xiahuayuan District’s landscape ecological risk degree, and thus the response degree between the low-risk degree change and the diffusion effect of LST were not high.

4. Discussion

In 2013, the China State Council printed and distributed a notice on the National Sustainable Development Plan for Resource-Based Cities (2013–2020), stating that the sustainable development of resource-based cities cannot be ignored. By 2020, the transformation of national resource-based cities made significant progress due to many efforts made in China. During the transformation, the urban microclimate and landscape ecological risks significantly changed, which affected urban development, the ecological environment and the well-being of human life. The previous literature has shown that landscape ecological risk and LST are associated with road distance [25,33], urban expansion [34,35,36,37], climate change and human activities [38,39,40]. However, mitigating effects on both factors have also been identified that are merely related to their own inherent qualities. For instance, landscape ecological risk degree has been associated with the spread of construction land [41] and the distance of the ecological protection zone [42], while LST has been associated with dominant tree species [43], crop yield [44] and urban 3D characteristics [45]. There is a lack of research on the coupling, coordination and correlation between the two. Accordingly, using the Xiahuayuan District as an example, this study used the grid unit to construct a landscape risk assessment to explore the coupling and coordination relationships between the long-time series landscape ecological risk and microclimate in resource-exhausted cities. This study provides both a theoretical reference for the sustainable development of resource-exhausted cities as well as new concepts for mitigating the urban landscape ecological risk degree and microclimate changes represented by LST.

This study has a few limitations. First, the selected grid scale of the evaluation unit may have affected the calculated value of the landscape ecological risk degree and the sampling value of the LST, potentially impacting the coupling coordination degree between the two. Second, most of the microclimate and landscape ecology of resource-exhausted cities, counties, and districts are highly complex, thus the study of the Xiahuayuan District likely does not fully represent the actual situation of many resource-exhausted cities, counties and districts. Future research urgently needs to explore the linkage between landscape ecological risk and microclimate in other resource-exhausted cities, counties and districts.

5. Conclusions

In this study, we explored the degree of coupling coordination between microclimate and landscape risks for transforming resource-exhausted cities based on surface temperature and landscape risks. For this, we integrated Landsat image series and land use data for 2000, 2010 and 2020, using the Xiahuayuan District as an example and dividing it into 568 evaluation units. We investigated the coupling coordination between microclimate and landscape ecological risk based on the coupling model of coordination. The main conclusions are as follows:

(1)

When studying the distribution of landscape ecological risk in the three stages, it was found that the high-risk areas are mainly distributed in the north, and the low-risk areas and medium low-risk areas are mainly distributed in the South and East. The risk transfer area from 2000 to 2010 is 69.75 km², and the risk transfer area from 2010 to 2020 is 107.76 km².

(2)

From 2000 to 2020, the distribution scope of LST gradually decreased from the western region to the eastern region. The proportion of medium-temperature areas in the research region was relatively large, accounting for 42.96%, 36.03%, and 47.05% in three periods, respectively, and the proportion of high-temperature areas gradually decreased.

(3)

The proportion of low-level coupled zones and the antagonistic coupled zones was relatively small, and the high-level coupled zones had the largest proportion difference over the entire study period of 2000–2020, accounting for 79.53%, 78.07%, and 85.06%, respectively. One portion of the running-in coupled zone in 2010 was transferred to high-level coupled zones in 2020. Furthermore, the landscape ecological risk degree in most areas of the Xiahuayuan District was closely associated with the LST and had a significant interaction.

(4)

Considering the coordination from 2000 to 2020, the imbalance proportion between landscape ecological risk degree and LST in the Xiahuayuan District was relatively low and was mainly concentrated in the coordinated regions, accounting for 78.80%, 80.97% and 83.13%, respectively. Overall, the landscape ecological risk degree and LST complemented each other, with fewer areas of mutual restriction and exclusion.

(5)

The researched region was mainly characterized by high and extremely high coupling coordination. Low and medium coupled coordination was collectively distributed in a few smaller areas in the eastern and southern parts of the region. This demonstrates that, from 2000 to 2020, the landscape ecological risk degree and LST in the Xiahuayuan District had a significant interactive relationship with strong coupling coordination.

According to our conclusions, there is strong coupling coordination between surface temperature and landscape ecological risk in the Xiahuayuan District. The research period spanned from the resource output period to the resource depletion period, then to the transformation and development period. Our data show that after more than 20 years of development, the surface temperature and the landscape ecological risk complement each other, with a consistent dynamic trend in the region. The results also indirectly show that the influencing factors driving the changes are not affecting only a single aspect; the spatial differences of coupling coordination degree in different regions also influence the driving factor. This leads to a high correlation between urban climate development and landscape ecological risk at the macro level. In the process of urban development, there are certain spatial differences between the change in urban climate and the degree of landscape ecological risk. In future research, it will be important to study the coupling, coordination and correlation between microclimate and landscape ecological risk in different cities at different scales while ensuring a rational scale.