Next Article in Journal
Organic Mulching Alters the Soil Microclimate, Increases Survival and Growth of Tree Seedlings in Restoration Planting
Previous Article in Journal
ALS-Based, Automated, Single-Tree 3D Reconstruction and Parameter Extraction Modeling
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Effects of the Spatial Pattern of Forest Vegetation on Urban Cooling in Large Metropolitan Areas of China: A Multi-Scale Perspective

College of Horticulture and Landscape Architecture, Yangzhou University, Yangzhou 225009, China
*
Author to whom correspondence should be addressed.
Forests 2024, 15(10), 1778; https://doi.org/10.3390/f15101778
Submission received: 1 September 2024 / Revised: 5 October 2024 / Accepted: 8 October 2024 / Published: 10 October 2024
(This article belongs to the Section Forest Meteorology and Climate Change)

Abstract

:
Urban forests are expected to mitigate the urban heat island (UHI) effect in megacities. The mechanism and factors influencing the cooling effect of urban forest have been extensively discussed; however, the spatial scale effect of cooling heterogeneity of the urban forest is still uncertain. Based on Landsat 8/9 OLI/TIRS imagery, the relationship between land surface temperature (LST) and the spatial patterns of forest vegetation in Beijing, Shanghai, and Tianjin was investigated at different spatial scales, including patch level, rural–urban gradient, and multiple spatial extents. The results indicated that the cooling effect of forest vegetation is stronger than that of grassland. The combination of the three indicators—Area, Normalized Difference Vegetation Index (NDVI), and the percentage of neighboring greenspace (NGP)—can largely explain the differences in cooling intensity between forest vegetation patches. The results suggest that the cooling effect of forest vegetation was affected by air humidity, and the cooling intensity of forest vegetation is stronger in coastal cities than in inland cities. In dry cities, the impact of the patch area on the cooling intensity of forest patches is greater than the NDVI, while the opposite is true in humid coastal cities. The LST variations in the urban–rural gradient can largely be explained by the landscape composition. This study proposes to apply larger spatial extents (e.g., 450 m × 450 m grid in this study) to investigate the relationship between landscape configuration metrics (e.g., Aggregation and Cohesion in this study) and the LST; and to use smaller spatial extents (e.g., 90 m × 90 m grid in this study) to reveal the relationship between area and shape related metrics. This study extends our scientific understanding of scaling effects to the relationship between landscape metrics and the LST.

1. Introduction

At present, over half of the world’s population lives in urban areas, and this figure will reach around 68% by 2050 [1]. Urbanization has a significant impact on the process and pattern evolution of the urban landscape. For example, the impervious area of urban surface increases, which is usually reflected in the transformation of natural surfaces (vegetation and soil, etc.) to impervious surfaces such as concrete and asphalt, and buildings of various heights and densities [2]. The change of land surface and the concentration of human activities brought by urbanization affect the processes and services of natural ecosystems, and further cause many social and environmental problems [3,4]. One is the urban heat island (UHI) effect, characterized by higher temperatures in urban areas than in rural areas [5,6]. The UHI effect not only worsens urban water and air quality [7], but accelerates urban energy consumption [8] and poses a potential threat to human health [9]. Globally, 9.1 out of every 1000 deaths are reported to be attributable to extreme heat [10]. As a result, the UHI effect has attracted considerable attention from urban decision-makers and academics worldwide [11,12,13,14,15,16]. In summary, the development of thermal mitigation measures is essential to improve the quality of life and the environment of urban residential areas and to facilitate the formulation of planning measures related to sustainable development.
The UHI effect is widely observed in cities regardless of their size and location, and the magnitude of the UHI often increases with increasing city size [11]. Li et al. [12] found that city size and population were significantly correlated with UHI intensity in 100 major cities across China mainland. Manoli et al. [13] indicated that the UHI intensity can be explained mainly by mean annual precipitation and city population size, and that increasing vegetation cover and albedo are effective strategies in dry regions but not in tropical cities. Similarly, the study by Zhou et al. [14] claimed that climatic conditions have the greatest influence on cross-city variations in UHI intensity. Based on a global report by Clinton and Gong [15], the magnitude of the UHI effect in different cities was largely determined by the amount of vegetation and the size of the urban metropolis. Thus, the risks associated with the UHI effect are greater in populous metropolitan cities around the world [16].
Urban green spaces (UGSs) reduce solar radiation on the surface through the shading of plants, and the synergistic mechanism of evapotranspiration and photosynthesis to help cool the air and mitigate the UHI effect [17,18], therefore, is considered as an urban “cold island” [19,20,21]. In addition, UGSs can improve human health [22,23], prompt social inclusion [24], and contribute to sustainable development. Since it is cost-effective, environmentally friendly, and politically acceptable, the creation of a UGS is increasingly considered to be a promising nature-based solution to mitigate the UHI [25,26]. The relationship between the spatial patterns of UGSs and local land surface temperature (LST), and the associated planning and management strategies have been continuously studied [27,28,29,30]. It is well known that an increase in the amount and percentage of a UGS leads to a reduction of temperatures, and this relationship is very consistent [28,31,32,33,34]. However, the effect of the spatial configuration of UGSs on the cooling effect is contradictory in some cases [27]. For example, a UGS with regular and compact geometry resulted in a greater reduction of LST in Addis Ababa, Ethiopia [35], Fuzhou, China [36] and Singapore [37]; while others claimed the opposite [38,39]. In addition, many researchers claimed that the patch density of a UGS was negatively correlated with LST [34,39,40,41]; but others hold the opposite view [42,43]. These inconsistencies prevent the application of the results to the planning and management of UGSs.
The reasons for this inconsistency could be that these studies (1) were conducted in cities with different climatic conditions [36,44,45,46,47]; (2) used satellite images with different spatial resolutions from 1 m to 1000 m [41,42,48,49]; and (3) applied different sizes of analytical units, such as grids, city blocks or self-defined polygons [32,34,39,41,50]. The studies by Wu et al. [51] and Min et al. [52] confirmed that the relationships between the LST and the driving factors are scale-dependent. However, most previous studies have revealed the landscape pattern–LST relationship at a single spatial scale or location. Therefore, the question arises whether the spatial pattern of a UGS affects LST differently in cities with different climatic conditions? Are the current inconsistent conclusions due to the different spatial extent of the analysis?
To address these insufficiencies, this study selected three megacities, Beijing, Tianjin, and Shanghai, as study areas, to quantify the relationship between the spatial patterns of urban forest vegetation and the cooling effect at multiple scales. The study area, the data used, the processing of the data (land cover classification and LST calculation), the selection of cooling indicators and landscape patterns, and the applied scales are described in Section 2, and followed by Section 3 showing the calculation results. The results in Section 3 mainly include the following: (1) the cooling intensity of different land cover types in three cities; (2) the combined effect of landscape indicators on the cooling intensity of urban forest vegetation and the differences among three cities; (3) the correlation between the spatial pattern metrics (composition and configuration) of urban forest vegetation and LST by different spatial extents (the grid scale in this study); and (4) the effects of changing spatial extents on the relationship between LST and spatial pattern metrics of urban forest vegetation. The discussion and application of the results, as well as the limitations of this study, are presented in Section 4, followed by Section 5 showing the main findings of this study.

2. Methodology

2.1. Study Area

The study areas included the urban areas and some rural areas of three megacities in China, Beijing, Shanghai, and Tianjin, each with a permanent urban population of more than 10 million [53]. Each study area covered a 50 km-grid landscape, with a radius of 25 km around the urban core (Figure 1). Beijing (115°25′–117°30′E, 39°28′–41°05′N) covers an area of about 16,410.54 km2; it had a permanent population of about 21.84 million at the end of 2022 [54]. Shanghai covers a total area of about 6340 km2, with the central city occupying an area of 664 km2. At the end of 2022, it had a population of 24.75 million [55]. The total area of Tianjin is 11,966.45 km2, and the permanent population of Tianjin was approximately 13.63 million at the end of 2022 [56]. All three cities are densely populated and socially and economically highly developed regions in China.
Beijing, Tianjin, and Shanghai are all located on the east coast of China, but in different climate zones, with different climate types and ecological environments. To be specific, Beijing has a typical sub-humid continental monsoon climate in the northern temperate zone, with a hot and rainy summer and a cold and dry winter. In 2022, the maximum temperature in Beijing was 39.2 °C and the average annual precipitation was 585.4 mm [54]. Shanghai has a north subtropical monsoon climate with four distinct seasons, plenty of sunshine, and abundant rainfall. In 2022, the highest temperature in Shanghai was 40.0 °C and the average annual precipitation was 1044.1 mm [55]. The climate here belongs to a warm temperate semi-humid monsoon climate with four distinct seasons, hot and rainy in summer, and cold and dry in winter. In 2022, the highest temperature in Tianjin was 40.7 °C, and the average annual precipitation was 589.9 mm [56].

2.2. Data Sources

In this study, we used three Landsat 8/9 Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS) images; the band information is shown in Supplementary Table S1. The Landsat 8/9 OLI/TIRS images were provided by the United States Geological Survey (https://earthexplorer.usgs.gov/ (accessed on 10 October 2023)) [57], all with less than 10% cloud cover. Detailed image information is shown in Supplementary Table S2. The OLI images (combined multispectral and panchromatic bands) were used for land cover classification and landscape pattern analysis, as well as for the calculation of Normalized Difference Vegetation Index (NDVI). The TIRS images were used for the LST calculation. The meteorological data on the acquisition date of the image is shown in Supplementary Table S3.

2.3. Land Cover Classification and LST Retrieval

In this study, the maximum likelihood method was applied to classify land cover types using the ENVI (the Environment for Visualizing Images) platform; the land cover map is shown in Supplementary Figure S1. Green space was simply categorized into two types based on the NDVI value to include forest vegetation and other vegetation. The study area was roughly classified into four types for research purposes to include: impervious surface, water, forest vegetation, and other vegetation; a description of land cover types can be found in Supplementary Table S4. The thermal band of the Landsat 8/9 images were used to calculate the LST maps of three cities using the mono-window algorithm (MWA) (Supplementary Figure S1) [58], which is based on the following equations:
T s = [ a 10 ( 1 C 10 D 10 ) + ( b 10 ( 1 C 10 D 10 ) + C 10 + D 10 ) T 10 D 10 T a ] / C 10
where a 10 and b 10 are coefficients corresponding to −67.9542 and 0.45987, respectively [58]. T10 stands for the at-sensor brightness temperature, and Ta stands for the mean atmospheric temperature. In addition, C10 and D10 are defined, respectively, by Formulas (2) and (3):
C 10 = τ 10 ε 10
D 10 = ( 1 τ 10 ) [ 1 + ( 1 ε 10 ) τ 10 ]
where τ 10 is the total atmospheric transmissivity; ε 10 represents the land surface emissivity.

2.4. Definition of Cooling Indicators

As in previous studies [32,59], we take the average LST of the study area (i.e., the average LST of the entire grid) as the reference LST (Tref). The cooling intensity (CI) of a landscape patch is equal to CI = Tref − Ti, where Ti is the average LST of the patch. Therefore, a land cover patch is considered an effective cold island (ECI) if Ti is smaller than Tref (CI > 0). Meanwhile a land cover patch is defined as a potential cold island (PCI) if Ti is greater than or equal to Tref (CI ≤ 0). Since the LST of the water body is significantly lower than that of other land cover types and its proportion varies greatly in the three cities (see Supplementary Table S5), the water body was excluded from the calculation of the reference LST.

2.5. Selection and Definition of Landscape Indicators

Eight landscape metrics were used to quantify the landscape composition and configuration of the study areas in Beijing, Shanghai, and Tianjin at both class- and patch-level (Table 1). The class-level landscape metrics, including percentage of landscape (PLAND), mean patch size (MPS), largest patch index (LPI), patch density (PD), mean patch shape index (Shape_mn), aggregation index (AI), and patch cohesion index (Cohesion), were used to describe the overall landscape patterns of Beijing, Shanghai, and Tianjin to better understand the results of this research; the corresponding results are shown in Supplementary Table S5. The patch area (Area) and percentage of neighboring greenspace (NGP) are two patch-level indicators applied to quantify the spatial characteristics of forest vegetation patches. NGP is a spatial metric proposed by Zhou et al. [59] for quantifying the proportion of a UGS in the vicinity of a landscape patch. The calculation of landscape metrics was performed based on the land cover maps using FRAGSTATS 4.2 [60].

2.6. Spatial Analysis

2.6.1. Patch Scale Analysis

In this study, the proportion of ECI and PCI for each land cover type was calculated. We confirmed the differences in the impact of each land cover type on the urban LST, and further explored the cross-city differences in the cooling performance of each land cover type. In addition, a predictive model for the CI of forest vegetation was constructed through the stepwise multiple linear regression (SW-MLR) method to quantify the combined effects of Area, NDVI, and NGP. The effects of the three landscape indicators on the CI were analyzed and compared, and the cross-city differences in the impact of landscape indicators on the CI were further investigated.

2.6.2. Rural–Urban Gradient Analysis

The heterogeneity of the rural–urban gradient of the LST is a common method for defining the UHI. We studied the relationship between landscape components and the mean LST across the urban–rural gradients. Considering that the spatial resolution of Landsat 8/9 TIRS images is 30 m, this study established multiple buffer zones at 300 m distance intervals in each urban center and determined the spatial variability of the LST gradient in urban and rural areas and its relationship with an impervious surface and the proportion of forest vegetation in three megacities by bivariate correlation analysis and regression analysis.

2.6.3. Multi-Spatial Extent Analysis

To study the influence of the varying spatial extent on the relationship between landscape metrics of forest vegetation LST, a multi-spatial extent-based analysis is needed. In this study, 9 different sizes of polygon grids were applied, including 60 m × 60 m, 90 m × 90 m, 180 m × 180 m, 270 m × 270 m, 360 m ×360 m, 450 m × 450 m, 540 m × 540 m, 630 m × 630 m, and 720 m × 720 m. The correlation coefficients between the landscape metric of forest vegetation and LST were then calculated under different sizes of polygon grids, and the results were further analyzed and compared between cities. Since the image resolution of LST is 30 m, the 9 selected grids are all integer multiples of 30 m.

3. Results

3.1. The Cooling Intensity of Different Land Cover Types

Among the three cities, Beijing and Shanghai had a similar PLAND of impervious surfaces (54.69% and 54.26%, respectively), and Tianjin had the lowest (49.14%) (Supplementary Table S5). The mean LST of impervious surface (39.28 °C) was highest in Shanghai, followed by Beijing (36.4 °C) and Tianjin (35.91 °C) (Table 2). In terms of the PLAND of forest vegetation, Tianjin had the highest value (32.69%), followed by Beijing (27.18%) and Shanghai (23.14%), respectively. The average LST of forest vegetation was highest in Shanghai (34.43 °C), followed by Beijing (32.01 °C) and Tianjin (30.56 °C). The results indicated that increasing the percent cover of forest vegetation could effectively reduce its mean LST. The PLAND of other vegetation was highest in Beijing, followed by Tianjin and Shanghai, while the average LST was highest in Shanghai and the lowest in Tianjin, suggesting that the association between PLAND and LST of other vegetation was not as strong as that of forest vegetation. The mean LST of water (30.27 °C, 31.01 °C, and 29.9 °C) was quite stable, although the PLAND of water bodies in the three cities was significantly different (1.99%, 13.28%, and 7.03%).
As shown in Table 2, the CI of water (4.43 °C, 6.69 °C, and 3.9 °C) was strongest, followed by forest vegetation (2.69 °C, 3.27 °C and 3.24 °C) and other vegetation (1 °C, 1.6 °C, and 1.8 °C). In addition, a similar pattern was observed when compared to the temperature of impervious surfaces. Different from other land cover types, the mean LST of impervious surfaces in the three study areas was all greater than the Tref, indicating that impervious surfaces were the main contributor to urban warming among different land cover types. Comparing different cities, the temperatures of the four land cover types in Shanghai were all higher than the other two cities. As shown in Figure 2, the ECI of water was the greatest for all cities, followed by forest vegetation and other vegetation. The difference in the ECI between forest vegetation and other vegetation in Beijing and Tianjin was similar, but the difference between the two was significantly smaller in Shanghai, which meant that the difference between the cooling performance of forest and other vegetation in Shanghai was relatively small.

3.2. Modelling the Relationship between Landscape Indicators and CI of Forest Vegetation in Three Cities

Table 3 summarizes the results of the SW-MLR analysis. The variance inflation factor (VIF) values ranged from 1.086 to 1.280, indicating a low degree of collinearity of the three landscape indicators in the three cities. The results showed that the Area, NGP, and NDVI combining together could explain a significant proportion of the variations of the CI of forest vegetation in Beijing (R2 = 0.670, p < 0.001), Shanghai (R2 = 0.729, p < 0.001), and Tianjin (R2 = 0.739, p < 0.001). In addition, the Area of forest vegetation was the most important variable affecting the CI in Beijing, followed by the NDVI and NGP, while the NDVI had a higher explanatory power in Shanghai and Tianjin, followed by the Area and NGP.

3.3. Impervious Surface and Forest Vegetation vs. LST along Urban–Rural Gradient

The results showed that the LST in the three cities, as a whole, presents a “peak-valley” distribution along the urban–rural gradient (Figure 3). As the distance from the urban center increased, the percentage of impervious surfaces of the three cities generally declined, and a valley value appeared in the first gradient (0 km~0.3 km). The percentage of forest vegetation generally increased, and the mean LST decreased along the urban–rural gradient. The average LST of Beijing, Shanghai, and Tianjin reached its peak at similar locations for the three cities, i.e., 0.20 km and 0.25 km away from the city center. Overall, the LST changed along the urban–rural gradient of the three cities all reflected the typical UHI phenomenon; that is, the LST decreased gradually along the urban–rural gradient. The results suggested Tianjin had the highest LST drop (6.1 °C), followed by Beijing (3.2 °C) and Shanghai (2.1 °C) along the urban–rural gradient.
As can be seen from Figure 3, the PLAND of impervious surface and forest vegetation both showed significant linear relationships with the LST (p < 0.001) along the urban–rural gradient. Specifically, the PLAND of the impervious surface could explain 89.1% (Beijing), 75.1% (Shanghai), and 95.9% (Tianjin) of the variations of LST. In addition, the PLAND of forest vegetation could explain 95.4% (Beijing), 85.3% (Shanghai), and 98.5% (Tianjin) of the LST variations. The results suggested that the LST variation along the urban–rural gradient could be largely explained by the composition of land cover types.

3.4. Effects of Changing Spatial Extent on the Relationship between Urban Forest Patterns and LST

As shown in Figure 4, all landscape metrics of forest vegetation in this study were negatively correlated with the average LST at different sizes of grids. The relationships between the PLAND, LPI, MPS, PD, Shape_mn and LST along an increasing spatial extent followed a similar pattern, and the correlations between the LST with PLAND and LPI were the strongest, followed by the MPS, PD, and was lowest with the Shape_mn. Similarly, the relationships between the AI, Cohesion and LST along increasing spatial extent followed a similar pattern. When the applied size of the grid was less than 360 m × 360 m, the correlations between the LST with the configuration metrics (i.e., AI and Cohesion in this study) were significantly weaker than with other landscape metrics.
In general, the correlation of the PLAND, MPS, LPI, PD, and Shape_mn with the average LST showed a similar downward trend with the increase of grid size, whereas the AI and Cohesion showed an upward trend. With the increase in grid size, the correlation between the AI and Cohesion and the average LST first increased and then decreased. Specifically, the AI and Cohesion had no significant correlation with the LST in a smaller spatial extent (e.g., 90 m and 180 m grids), but the correlation increased significantly when the applied grid scale exceeded 270 m, reaching its highest value at a certain spatial extent (i.e., 450 m or 540 m grid in this study). Similar patterns were observed in the three cities.

4. Discussion

Both vegetation and water bodies had a considerable cooling effect during the day in summer in the three cities. Impervious surfaces were the main source of heat in the cities, which was consistent with previous studies [61,62]. There were regional differences in the magnitude of the cooling effect of forest vegetation, other vegetation, and water bodies compared to different cities. The cooling effect of water bodies in summer daytime was the strongest, and this conclusion was consistent [63,64]. However, it was worth noting that, although the proportion of water bodies in the three cities was very different, there was no significant difference in the average temperature. The results suggested that increasing the area of a water body did not effectively reduce its temperature, but only affected the cumulative benefit of cooling. Therefore, it was recommended to replace the aggregated arrangement with a scattered and evenly distributed pattern to benefit more areas when the proportion is fixed. The cooling effect of forest vegetation in the three cities was stronger than that of grassland, with Shanghai having the strongest effect, followed by Tianjin and Shanghai, which was consistent with previous studies [64,65]. This result suggested that the cooling effect of plants was somewhat related to local climate conditions. When the temperature is at a similar level, coastal cities often have a stronger cooling effect on their plants due to high humidity.
The patch area, NDVI, and NGP can explain most of the changes in the cooling intensity of forest vegetation, but there were some differences between the cities. The cooling effect of forest vegetation in Beijing was most affected by the Area, followed by NDVI and NGP. In Shanghai and Tianjin, the NDVI was the most important indicator of cooling intensity between forest vegetation patches, followed by the Area and NGP. This once again confirmed the influence of humidity on the cooling effect of forest vegetation, and cities with higher humidity also had a stronger cooling effect. The results suggested that targeted planning strategies based on the climatic background of different cities should be developed to better utilize the cooling effect of urban forest vegetation. Strategies such as planting more trees to replace grasslands, selecting tree species with larger leaf areas, and improving the health level of existing vegetation can effectively enhance the cooling effect of UGSs. Although it was difficult to increase the area of existing forest vegetation patches in highly urbanized areas, planting as much greenery as possible, even vertical greenery, on spare land near a vegetation patch can increase the NGP and thus enhance the cooling effect of existing forest patches.
The LST variations in urban–rural gradients can be largely explained by the composition of land cover, with impervious surfaces and forest vegetation playing a dominant role. The results confirmed that the replacement of natural areas by impervious surfaces was the main landscape change that triggered the UHI effect. In general, the correlation between the percentage of impervious surface and LST was higher than that of forest vegetation. Therefore, strategies related to reducing the temperature of impervious surfaces, such as building surfaces, squares, and roads, should be taken seriously. The cooling effect of water bodies was stronger than that of forest vegetation and other vegetation (Figure 2), which was consistent with previous studies [59]. Because water bodies made up a smaller part of the city, they were not the main “decider” for the urban LST. Therefore, urban planners need to pay more attention to landscape optimization around water bodies to maximize their cooling effect.
This study selected seven class-level landscape metrics to quantify the relationship between the spatial pattern of forest vegetation and LST and link the relationship with changing spatial extent. The magnitude of the correlation between component and configuration metrics of forest vegetation and the mean LST varied greatly in changing the spatial extent (Figure 4). Specifically, the correlation between the PLAND and mean LST in this study showed a decreasing trend with the increase of grid size, which was consistent with the results of previous studies [62,66]. This may be because it was easier to obtain the landscape proportion of forest vegetation distributed from 0% to 100% under a small grid unit, and therefore the correlation between the spatial proportion of forest vegetation and the LST was easier to obtain. However, under the larger grid size, the difference in the PLAND of forest vegetation among each unit was relatively small, which hindered the exploration of the correlation between the two. This study confirmed that increasing the amount of forest vegetation could effectively reduce the regional LST. Consistent with the PLAND, MPS and LPI were significantly correlated with the mean LST at a small spatial extent, and their correlation decreased with the increase of grid size, which was consistent with previous studies [28]. This was largely because, with the increase of spatial extent, the value intervals of MPS and LPI shrank, and the sample distribution was too concentrated. On the contrary, the correlation between the AI, Cohesion, and LST was not found at a small spatial extent but increased with the increase of grid size. In addition, the shape index of forest vegetation and the regional LST showed a low correlation under different spatial extents, which was consistent with some previous conclusions [67,68]. The results indicated that forest vegetation patches with regular and compact geometry have greater cooling effects than irregular and elongated ones when the total amount was fixed; however, the impact of shape complexity was limited.
It can be concluded that there was no uniform and optimal spatial extent for studying the relationship between landscape patterns and the LST, which is consistent with previous studies (see Supplementary Table S6). Different landscape metrics were applicable to different spatial units, and some landscape metrics were suitable to be measured under a smaller spatial extent. For example, in this study, 90 m × 90 m is the best spatial extent for analyzing the relationship between the PLAND, LPI, MPS, PD and Shape_mn of forest vegetation and the mean LST. In addition, 450 m × 450 m was the optimal scale to quantify the relationship between the LST and spatial configuration metrics (e.g., AI and Cohesion in this study). The results of this study suggested that the spatial composition of the landscape accounts for more variation in the urban LST than spatial configuration, which is consistent with previous studies [37,41,43].
There were some limitations of this study that should be addressed. First, the spatial resolution of the Landsat 8/9 imagery is limited, which restricts the application of conclusions based on this resolution in urban planning. Second, this research was based on remote-sensed imagery; therefore, the findings are dependent on the collection time and seasons. Finally, this research was only conducted in three cities in China, so the results may not be transferable to other cities even with similar climatic conditions. Therefore, more cities with different climatic backgrounds should be considered in future studies to extend the understanding of the influence of local climate conditions on the cooling effect of urban forest vegetation.

5. Conclusions

Based on Landsat 8/9 OLI and TIRS imagery, the cooling effect of urban forest vegetation under patch level, rural–urban gradient, and the multi-grid level was studied, and the relationship between abundance and spatial patterns of forest vegetation and LST was discussed in Beijing, Shanghai, and Tianjin. More evidence was provided on the influence of landscape composition and configuration on the LST variations in the megacities of China. In addition, this study emphasized the importance of selecting the appropriate spatial extent, which meant that the landscape pattern–LST relationship may be underestimated or overestimated at certain research scales. Overall, the main conclusions of this study are as follows:
(1)
The cooling effect of water bodies was strongest, followed by forest vegetation and grassland. The average cooling effect of forest vegetation was 2.69 °C (Beijing), 3.27 °C (Shanghai), and 3.24 °C (Tianjin). On average, the mean LST of forest vegetation was about 4.39 °C, 4.85 °C, and 5.35 °C lower than that of impervious surfaces in Beijing, Shanghai, and Tianjin, respectively.
(2)
The LST variations in urban–rural gradients can be largely explained by the landscape composition, and the proportion of impervious surfaces and forest vegetation played a dominant role. More attention should be paid to those areas between 0.2 km and 0.25 km away from the city center since the average LST was higher there than at other locations, and this was consistent for all three cities.
(3)
Combining the Area, NDVI and NGP can explain a significant amount of heterogeneity of the cooling effect of forest vegetation. The patch area was the most important indicator influencing the cooling effect of forest vegetation in Beijing, while the NDVI had the greatest explanatory power in Shanghai and Tianjin. This difference was likely caused by differences in air humidity between the cities.
(4)
Changing the spatial extent had a great impact on the relationship between the spatial pattern of forest vegetation and LST, and the effects were basically consistent among the three urban areas in this study. A larger spatial extent (i.e., 450 m grid) was suggested to reveal the relationship between spatial configuration metrics (e.g., Aggregation and Cohesion) and the mean LST; meanwhile, a small spatial extent (i.e., 90 m grid) was recommended to quantify the correlation between the LST and area-, density- and shape-related metrics (e.g., PLAND, LPI, MPS, PD and Shape_mn) in this study.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/f15101778/s1, Table S1. Band information of Landsat 8/9 Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS). Table S2. Descriptions of the Landsat 8/9 OLI/TIRS images used. Table S3. The meteorological data at the acquisition time of the image. Table S4. Land cover categories and their descriptions. Table S5. General landscape characteristics in Beijing (2022), Shanghai (2022), and Tianjin (2023). Table S6. Multi-scale studies on the impact of landscape pattern on temperature. Figure S1. LST and land cover maps of Beijing, Shanghai, and Tianjin. References [69,70,71,72,73,74,75,76,77] are cited in Supplementary Materials.

Author Contributions

Conceptualization, J.X. and Y.Y.; methodology, J.X. and W.Z.; software, J.X. and Y.Y.; validation, W.Z.; formal analysis, J.X.; investigation, W.Z. and W.Y.; resources, W.Z.; data curation, W.Z.; writing—original draft preparation, J.X.; writing—review and editing, J.X., Y.Y. and W.Y.; visualization, J.X.; supervision, T.W.; project administration, W.Z.; funding acquisition, W.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Natural Science Foundation of China (Grant Number: 32101577; Representative: Wen Zhou) and Scientific Research Foundation for Advanced Talents, Yangzhou University (Grant Number: 137012167; Representative: Wen Zhou).

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors on request.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. United Nations Department of Economic and Social Affairs. World Urbanization Prospects: The 2018 Revision; United Nations Publications; United Nations: New York, NY, USA, 2019. [Google Scholar]
  2. Akbari, H.; Pomerantz, M.; Taha, H.J.P. Cool surfaces and shade trees to reduce energy use and improve air quality in urban areas. Sol. Energy 2001, 70, 295–310. [Google Scholar] [CrossRef]
  3. Peng, J.; Xie, P.; Liu, Y.X.; Ma, J. Urban thermal environment dynamics and associated landscape pattern factors: A case study in the Beijing metropolitan region. Remote Sens. Environ. 2016, 173, 145–155. [Google Scholar] [CrossRef]
  4. Sun, R.H.; Lü, Y.H.; Yang, X.J.; Chen, L.D. Understanding the variability of urban heat islands from local background climate and urbanization. J. Clean. Prod. 2019, 208, 743–752. [Google Scholar] [CrossRef]
  5. Oke, T.R. The energetic basis of the urban heat-island. Q. J. R. Meteorol. Soc. 1982, 108, 1–24. [Google Scholar] [CrossRef]
  6. Zhou, D.C.; Bonafoni, S.; Zhang, L.X.; Wang, R.H. Remote sensing of the urban heat island effect in a highly populated urban agglomeration area in East China. Sci. Total Environ. 2018, 628–629, 415–429. [Google Scholar] [CrossRef]
  7. Grimm, N.B.; Faeth, S.H.; Golubiewski, N.E.; Redman, C.L.; Wu, J.G.; Bai, X.M.; Briggs, J.M. Global change and the ecology of cities. Science 2008, 319, 756–760. [Google Scholar] [CrossRef]
  8. Li, X.M.; Zhou, Y.Y.; Yu, S.; Jia, G.S.; Li, H.D.; Li, W.L. Urban heat island impacts on building energy consumption: A review of approaches and findings. Energy 2019, 174, 407–419. [Google Scholar] [CrossRef]
  9. Gong, P.; Liang, S.; Carlton, E.J.; Jiang, Q.W.; Wu, J.Y.; Wang, L.; Remais, J.V. Urbanisation and health in China. Lancet 2012, 379, 843–852. [Google Scholar] [CrossRef]
  10. Zhao, Q.; Guo, Y.M.; Ye, T.T.; Gasparrini, A.; Tong, S.L.; Overcenco, A.; Urban, A.; Schneider, A.; Entezari, A.; Vicedo-Cabrera, A.M.; et al. Global, regional, and national burden of mortality associated with non-optimal ambient temperatures from 2000 to 2019: A three-stage modelling study. Lancet Planet. Health 2021, 5, E415–E425. [Google Scholar] [CrossRef]
  11. Soydan, O. Effects of landscape composition and patterns on land surface temperature: Urban heat island case study for Nigde, Turkey. Urban Clim. 2020, 34, 100688. [Google Scholar] [CrossRef]
  12. Li, K.N.; Chen, Y.H.; Wang, M.J.; Gong, A. Spatial-temporal variations of surface urban heat island intensity induced by different definitions of rural extents in China. Sci. Total Environ. 2019, 669, 229–247. [Google Scholar] [CrossRef]
  13. Manoli, G.; Fatichi, S.; Schläpfer, M.; Yu, K.L.; Crowther, T.W.; Meili, N.; Burlando, P.; Katul, G.G.; Bou-Zeid, E. Magnitude of urban heat islands largely explained by climate and population. Nature 2019, 573, 55–60. [Google Scholar] [CrossRef]
  14. Zhou, D.C.; Zhao, S.Q.; Liu, S.G.; Zhang, L.X.; Zhu, C. Surface urban heat island in China’s 32 major cities: Spatial patterns and drivers. Remote Sens. Environ. 2014, 152, 51–61. [Google Scholar] [CrossRef]
  15. Clinton, N.; Gong, P. MODIS detected surface urban heat islands and sinks: Global locations and controls. Remote Sens. Environ. 2013, 134, 294–304. [Google Scholar] [CrossRef]
  16. Department of Health. Heatwave Plan for England. Protecting Health and Reducing from Extreme Heat and Heatwaves. In NHS Best Practice Guidance; Department of Health: London, UK, 2008; Volume 32. [Google Scholar]
  17. Gunawardena, K.R.; Wells, M.J.; Kershaw, T. Utilising green and bluespace to mitigate urban heat island intensity. Sci. Total Environ. 2017, 584, 1040–1055. [Google Scholar] [CrossRef]
  18. Xue, Z.S.; Hou, G.L.; Zhang, Z.S.; Lyu, X.G.; Jiang, M.; Zou, Y.C.; Shen, X.J.; Wang, J.; Liu, X.H. Quantifying the cooling-effects of urban and peri-urban wetlands using remote sensing data: Case study of cities of Northeast China. Landsc. Urban Plan. 2019, 182, 92–100. [Google Scholar] [CrossRef]
  19. Oke, T.R. City size and urban heat island. Atmos. Environ. 1973, 7, 769–779. [Google Scholar] [CrossRef]
  20. Peng, J.; Liu, Q.Y.; Xu, Z.H.; Lyu, D.N.; Du, Y.Y.; Qiao, R.L.; Wu, J.S. How to effectively mitigate urban heat island effect? A perspective of waterbody patch size threshold. Landsc. Urban Plan. 2020, 202, 103873. [Google Scholar] [CrossRef]
  21. Wong, N.H.; Tan, C.L.; Kolokotsa, D.D.; Takebayashi, H. Greenery as a mitigation and adaptation strategy to urban heat. Nat. Rev. Earth Environ. 2021, 2, 166–181. [Google Scholar] [CrossRef]
  22. Bratman, G.N.; Anderson, C.B.; Berman, M.G.; Cochran, B.; de Vries, S.; Flanders, J.; Folke, C.; Frumkin, H.; Gross, J.J.; Hartig, T.; et al. Nature and mental health: An ecosystem service perspective. Sci. Adv. 2019, 5, eaax0903. [Google Scholar] [CrossRef]
  23. Spotswood, E.N.; Benjamin, M.; Stoneburner, L.; Wheeler, M.M.; Beller, E.E.; Balk, D.; McPhearson, T.; Kuo, M.; McDonald, R.I. Nature inequity and higher COVID-19 case rates in less-green neighbourhoods in the United States. Nat. Sustain. 2021, 4, 1092–1098. [Google Scholar] [CrossRef]
  24. Wolch, J.R.; Byrne, J.; Newell, J.P. Urban green space, public health, and environmental justice: The challenge of making cities ‘just green enough’. Landsc. Urban Plan. 2014, 125, 234–244. [Google Scholar] [CrossRef]
  25. Santamouris, M.; Ban-Weiss, G.; Osmond, P.; Paolini, R.; Synnefa, A.; Cartalis, C.; Muscio, A.; Zinzi, M.; Morakinyo, T.E.; Ng, E.; et al. Progress in urban greenery mitigation science—Assessment methodologies advanced technologies and impact on cities. J. Civ. Eng. Manag. 2018, 24, 638–671. [Google Scholar] [CrossRef]
  26. Yu, Z.W.; Guo, X.Y.; Zeng, Y.X.; Koga, M.; Vejre, H. Variations in land surface temperature and cooling efficiency of green space in rapid urbanization: The case of Fuzhou city, China. Urban For. Urban Green. 2018, 29, 113–121. [Google Scholar] [CrossRef]
  27. Zhou, W.Q.; Wang, J.; Cadenasso, M.L. Effects of the spatial configuration of trees on urban heat mitigation: A comparative study. Remote Sens. Environ. 2017, 195, 1–12. [Google Scholar] [CrossRef]
  28. Zhou, W.; Cao, F.L. Effects of changing spatial extent on the relationship between urban forest patterns and land surface temperature. Ecol. Indic. 2020, 109, 105778. [Google Scholar] [CrossRef]
  29. Du, H.Y.; Cai, W.B.; Xu, Y.Q.; Wang, Z.B.; Wang, Y.Y.; Cai, Y.L. Quantifying the cool island effects of urban green spaces using remote sensing Data. Urban For. Urban Green. 2017, 27, 24–31. [Google Scholar] [CrossRef]
  30. Kong, F.H.; Nakagoshi, N. Spatial-temporal gradient analysis of urban green spaces in Jinan, China. Landsc. Urban Plan. 2006, 78, 147–164. [Google Scholar] [CrossRef]
  31. Guerri, G.; Crisci, A.; Congedo, L.; Munafò, M.; Morabito, M. A functional seasonal thermal hot-spot classification: Focus on industrial sites. Sci. Total Environ. 2022, 806, 151383. [Google Scholar] [CrossRef]
  32. Kong, F.H.; Yin, H.W.; James, P.; Hutyra, L.R.; He, H.S. Effects of spatial pattern of greenspace on urban cooling in a large metropolitan area of eastern China. Landsc. Urban Plan. 2014, 128, 35–47. [Google Scholar] [CrossRef]
  33. Marando, F.; Heris, M.P.; Zulian, G.; Udías, A.; Mentaschi, L.; Chrysoulakis, N.; Parastatidis, D.; Maes, J. Urban heat island mitigation by green infrastructure in European Functional Urban Areas. Sustain. Cities Soc. 2022, 77, 103564. [Google Scholar] [CrossRef]
  34. Zhou, W.; Cao, F.L.; Wang, G.B. Effects of Spatial Pattern of Forest Vegetation on Urban Cooling in a Compact Megacity. Forests 2019, 10, 282. [Google Scholar] [CrossRef]
  35. Feyisa, G.L.; Dons, K.; Meilby, H. Efficiency of parks in mitigating urban heat island effect: An example from Addis Ababa. Landsc. Urban Plan. 2014, 123, 87–95. [Google Scholar] [CrossRef]
  36. Yu, Z.W.; Guo, X.Y.; Jorgensen, G.; Vejre, H. How can urban green spaces be planned for climate adaptation in subtropical cities? Ecol. Indic. 2017, 82, 152–162. [Google Scholar] [CrossRef]
  37. 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, 44–58. [Google Scholar] [CrossRef]
  38. Chen, A.L.; Yao, X.A.; Sun, R.H.; Chen, L.D. Effect of urban green patterns on surface urban cool islands and its seasonal variations. Urban For. Urban Green. 2014, 13, 646–654. [Google Scholar] [CrossRef]
  39. Dugord, P.A.; Lauf, S.; Schuster, C.; Kleinschmit, B. Land use patterns, temperature distribution, and potential heat stress risk—The case study Berlin, Germany. Comput. Environ. Urban Syst. 2014, 48, 86–98. [Google Scholar] [CrossRef]
  40. Li, J.X.; Song, C.H.; Cao, L.; Zhu, F.G.; Meng, X.L.; Wu, J.G. Impacts of landscape structure on surface urban heat islands: A case study of Shanghai, China. Remote Sens. Environ. 2011, 115, 3249–3263. [Google Scholar] [CrossRef]
  41. Zhou, W.Q.; Huang, G.L.; Cadenasso, M.L. Does spatial configuration matter? Understanding the effects of land cover pattern on land surface temperature in urban landscapes. Landsc. Urban Plan. 2011, 102, 54–63. [Google Scholar] [CrossRef]
  42. Li, X.M.; Zhou, W.Q.; Ouyang, Z.Y. Relationship between land surface temperature and spatial pattern of greenspace: What are the effects of spatial resolution? Landsc. Urban Plan. 2013, 114, 1–8. [Google Scholar] [CrossRef]
  43. Li, X.M.; Zhou, W.Q.; Ouyang, Z.Y.; Xu, W.H.; Zheng, H. Spatial pattern of greenspace affects land surface temperature: Evidence from the heavily urbanized Beijing metropolitan area, China. Landsc. Ecol. 2012, 27, 887–898. [Google Scholar] [CrossRef]
  44. 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, 349–359. [Google Scholar] [CrossRef]
  45. Fan, H.Y.; Yu, Z.W.; Yang, G.Y.; Liu, T.Y.; Liu, T.Y.; Hung, C.H.; Vejre, H. How to cool hot-humid (Asian) cities with urban trees? An optimal landscape size perspective. Agric. For. Meteorol. 2019, 265, 338–348. [Google Scholar] [CrossRef]
  46. Jauregui, E. Influence of a large urban park on temperature and convective precipitation in a tropical city. Energy Build. 1990, 15, 457–463. [Google Scholar] [CrossRef]
  47. Zhou, W.; Yu, W.D.; Zhang, Z.Y.; Cao, W.; Wu, T. How can urban green spaces be planned to mitigate urban heat island effect under different climatic backgrounds? A threshold-based perspective. Sci. Total Environ. 2023, 890, 164422. [Google Scholar] [CrossRef]
  48. Wu, H.; Ye, L.P.; Shi, W.Z.; Clarke, K.C. Assessing the effects of land use spatial structure on urban heat islands using HJ-1B remote sensing imagery in Wuhan, China. Int. J. Appl. Earth Obs. Geoinf. 2014, 32, 67–78. [Google Scholar] [CrossRef]
  49. Rhee, J.; Park, S.; Lu, Z.J.G.; Sensing, R. Relationship between land cover patterns and surface temperature in urban areas. GIScience Remote Sens. 2014, 51, 521–536. [Google Scholar] [CrossRef]
  50. Park, J.; Kim, J.H.; Lee, D.K.; Park, C.Y.; Jeong, S.G. The influence of small green space type and structure at the street level on urban heat island mitigation. Urban For. Urban Green. 2017, 21, 203–212. [Google Scholar] [CrossRef]
  51. Wu, Q.; Tan, J.X.; Guo, F.X.; Li, H.Q.; Chen, S.B. Multi-Scale Relationship between Land Surface Temperature and Landscape Pattern Based on Wavelet Coherence: The Case of Metropolitan Beijing, China. Remote Sens. 2019, 11, 3021. [Google Scholar] [CrossRef]
  52. Min, M.; Lin, C.; Duan, X.J.; Jin, Z.F.; Zhang, L.C. Spatial distribution and driving force analysis of urban heat island effect based on raster data: A case study of the Nanjing metropolitan area, China. Sustain. Cities Soc. 2019, 50, 101637. [Google Scholar] [CrossRef]
  53. State Statistics Bureau. Statistical Chart of Economic and Social Development: Basic Information of the Population of Megacities and Megacities in the Seventh National Population Census. Available online: http://www.qstheory.cn/dukan/qs/2021-09/16/c_1127863567.htm (accessed on 16 September 2021).
  54. Beijing Municipal Bureau Statistics. Beijing Statistical Yearbook 2023. Available online: https://nj.tjj.beijing.gov.cn/nj/main/2023-tjnj/zk/indexch.htm (accessed on 23 November 2023).
  55. Shanghai Municipal Bureau Statistics. Shanghai Statistical Yearbook 2023. Available online: https://tjj.sh.gov.cn/tjnj/20240321/5a35a44acace471f87c75393133fa142.html (accessed on 23 November 2023).
  56. Tianjin Municipal Bureau Statistics. Tianjin Statistical Yearbook 2023. Available online: https://stats.tj.gov.cn/nianjian/2023nj/zk/indexch.htm (accessed on 23 November 2023).
  57. USGS (United States Geological Survey). Available online: https://earthexplorer.usgs.gov/ (accessed on 10 October 2023).
  58. Qin, Z.; Karnieli, A.; Berliner, P. A mono-window algorithm for retrieving land surface temperature from Landsat TM data and its application to the Israel-Egypt border region. Int. J. Remote Sens. 2001, 22, 3719–3746. [Google Scholar] [CrossRef]
  59. Zhou, W.; Yu, W.L.; Wu, T. An alternative method of developing landscape strategies for urban cooling: A threshold-based perspective. Landsc. Urban Plan. 2022, 225, 104449. [Google Scholar] [CrossRef]
  60. McGarigal, K.; Marks, B.J. FRAGSTATS: Spatial Pattern Analysis Program for Quantifying Landscape Structure; Environmental Science; US Department of Agriculture, Forest Service, Pacific Northwest Research Station: Washington, DC, USA, 1995. [Google Scholar]
  61. Yuan, F.; Bauer, M.E. Comparison of impervious surface area and normalized difference vegetation index as indicators of surface urban heat island effects in Landsat imagery. Remote Sens. Environ. 2007, 106, 375–386. [Google Scholar] [CrossRef]
  62. Maimaitiyiming, M.; Ghulam, A.; Tiyip, T.; Pla, F.; Latorre-Carmona, P.; Halik, Ü.; Sawut, M.; Caetano, M. Effects of green space spatial pattern on land surface temperature: Implications for sustainable urban planning and climate change adaptation. ISPRS J. Photogramm. Remote Sens. 2014, 89, 59–66. [Google Scholar] [CrossRef]
  63. Zhou, W.; Cao, W.; Wu, T.; Zhang, T. The win-win interaction between integrated blue and green space on urban cooling. Sci. Total Environ. 2023, 863, 160712. [Google Scholar] [CrossRef]
  64. Yang, G.Y.; Yu, Z.W.; Jorgensen, G.; Vejre, H. How can urban blue-green space be planned for climate adaption in high-latitude cities? A seasonal perspective. Sustain. Cities Soc. 2020, 53, 101932. [Google Scholar] [CrossRef]
  65. Yu, Z.W.; Yao, Y.W.; Yang, G.Y.; Wang, X.R.; 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, 242–254. [Google Scholar] [CrossRef]
  66. Li, X.X.; Li, W.W.; Middel, A.; Harlan, S.L.; Brazel, A.J.; Turner, B.L. Remote sensing of the surface urban heat island and land architecture in Phoenix, Arizona: Combined effects of land composition and configuration and cadastral-demographic-economic factors. Remote Sens. Environ. 2016, 174, 233–243. [Google Scholar] [CrossRef]
  67. Asgarian, A.; Amiri, B.J.; Sakieh, Y. Assessing the effect of green cover spatial patterns on urban land surface temperature using landscape metrics approach. Urban Ecosyst. 2015, 18, 209–222. [Google Scholar] [CrossRef]
  68. Bao, T.L.G.; Li, X.M.; Zhang, J.; Zhang, Y.J.; Tian, S.Z. Assessing the Distribution of Urban Green Spaces and its Anisotropic Cooling Distance on Urban Heat Island Pattern in Baotou, China. ISPRS Int. J. Geo-Inf. 2016, 5, 12. [Google Scholar] [CrossRef]
  69. Xiao, R.B.; Ouyang, Z.Y.; Zheng, H.; Li, W.F.; Schienke, E.W.; Wang, X.K. Spatial pattern of impervious surfaces and their impacts on land surface temperature in Beijing, China. J. Environ. Sci. 2007, 19, 250–256. [Google Scholar] [CrossRef]
  70. Myint, S.W.; Brazel, A.; Okin, G.; Buyantuyev, A. Combined Effects of Impervious Surface and Vegetation Cover on Air Temperature Variations in a Rapidly Expanding Desert City. Giscience Remote Sens. 2010, 47, 301–320. [Google Scholar] [CrossRef]
  71. Ma, Q.; Wu, J.G.; He, C.Y. A hierarchical analysis of the relationship between urban impervious surfaces and land surface temperatures: Spatial scale dependence, temporal variations, and bioclimatic modulation. Landsc. Ecol. 2016, 31, 1139–1153. [Google Scholar] [CrossRef]
  72. Liu, W.R.; Jia, B.Q.; Li, T.; Zhang, Q.M.; Ma, J. Correlation Analysis between Urban Green Space and Land Surface Temperature from the Perspective of Spatial Heterogeneity: A Case Study within the Sixth Ring Road of Beijing. Sustainability 2022, 14, 13492. [Google Scholar] [CrossRef]
  73. Yuan, S.F.; Xia, H.; Yang, L.X. How changing grain size affects the land surface temperature pattern in rapidly urbanizing area: A case study of the central urban districts of Hangzhou City, China. Environ. Sci. Pollut. Res. 2021, 28, 40060–40074. [Google Scholar] [CrossRef]
  74. Hou, H.; Estoque, R.C. Detecting Cooling Effect of Landscape from Composition and Configuration: An Urban Heat Island Study on Hangzhou. Urban For. Urban Green. 2020, 53, 126719. [Google Scholar] [CrossRef]
  75. Song, J.; Du, S.; Feng, X.; Guo, L. The relationships between landscape compositions and land surface temperature: Quantifying their resolution sensitivity with spatial regression models. Landsc. Urban Plan. 2014, 123, 145–157. [Google Scholar] [CrossRef]
  76. Zhao, J.C.; Zhao, X.; Liang, S.L.; Zhou, T.; Du, X.Z.; Xu, P.P.; Wu, D.H. Assessing the thermal contributions of urban land cover types. Landsc. Urban Plan. 2020, 204, 103927. [Google Scholar] [CrossRef]
  77. He, M.W.; Yuan, C.; Zhang, X.; Wang, P.C.; Yao, C.H. Impacts of green-blue-grey infrastructures on high-density urban thermal environment at multiple spatial scales: A case study in Wuhan. Urban Clim. 2023, 52, 101714. [Google Scholar] [CrossRef]
Figure 1. (a) Location of the study area. (b) Landsat 8/9 images of the study area in Beijing, Shanghai in 2022, and Tianjin in 2023 displayed in false color composite (Red—band 5; Green—band 4; and Blue—band 3).
Figure 1. (a) Location of the study area. (b) Landsat 8/9 images of the study area in Beijing, Shanghai in 2022, and Tianjin in 2023 displayed in false color composite (Red—band 5; Green—band 4; and Blue—band 3).
Forests 15 01778 g001
Figure 2. The proportion of effective cool island (ECI) and potential cool island (PCI) of each land cover type. IS, FV and OV represent impervious surface, forest vegetation and other vegetation, respectively.
Figure 2. The proportion of effective cool island (ECI) and potential cool island (PCI) of each land cover type. IS, FV and OV represent impervious surface, forest vegetation and other vegetation, respectively.
Forests 15 01778 g002
Figure 3. The relationship between the PLAND of impervious surface and forest vegetation and average LST along the urban–rural gradients of Beijing, Shanghai, and Tianjin.
Figure 3. The relationship between the PLAND of impervious surface and forest vegetation and average LST along the urban–rural gradients of Beijing, Shanghai, and Tianjin.
Forests 15 01778 g003
Figure 4. Correlation coefficients between landscape metrics of forest vegetation and mean LST under different spatial extents.
Figure 4. Correlation coefficients between landscape metrics of forest vegetation and mean LST under different spatial extents.
Forests 15 01778 g004
Table 1. Definitions of landscape metrics.
Table 1. Definitions of landscape metrics.
Landscape MetricsAbbreviationDescriptionRange
Percentage of LandscapePLANDThe proportion of total area occupied by a particular patch type; a measure of landscape composition and dominance of patch types (%)0 < PLAND < 100
Mean patch sizeMPSThe sum of area across all patches of the corresponding patch type divided by the number of patches of the same type (ha)MPS > 0
Largest patch indexLPIThe area (m2) of the largest patch of the corresponding patch type divided by total landscape area (m2), multiplied by 100 (to convert to a percentage) (%)0 < LPI < 100
Patch densityPDThe number of patches in the landscape for patch typePD > 0
Mean patch shape indexShape_mnMean value of shape indexShape_mn > 0
Aggregation indexAIThe number of like adjacencies involving the corresponding class, divided by the maximum possible number of like adjacencies involving the corresponding class, which is achieved when the class is maximally clumped into a single, compact patch; multiplied by 100 (to convert to a percentage) (%)0 ≤ AI ≤ 100
Patch cohesion indexCohesion1 minus the sum of patch perimeter (in terms of number of cell surfaces) divided by the sum of patch perimeter times the square root of patch area (in terms of number of cells) for patches of the corresponding patch type, divided by 1 minus 1 over the square root of the total number of cells in the landscape, multiplied by 100 to convert to a percentage0 ≤ Cohesion < 100
Patch areaAreaThe area of the patch (ha)Area > 0, no limit
Table 2. Results of the mean LST of different land cover types and the cooling effect (Unit: °C).
Table 2. Results of the mean LST of different land cover types and the cooling effect (Unit: °C).
CityLand Cover TypeWaterImpervious SurfaceForest VegetationOther Vegetation
BeijingMean LST30.2736.432.0133.7
CI4.43−1.72.691
TD compared with the mean LST of impervious surface6.1304.392.7
ShanghaiMean LST31.0139.2834.4336.1
CI6.69−1.583.271.6
TD compared with the mean LST of impervious surface8.2704.853.18
TianjinMean LST29.935.9130.5632
CI3.9−2.113.241.8
TD compared with the mean LST of impervious surface6.0105.353.91
Table 3. Results of the SW-MLR analyses (Dependent Variable is cooling intensity (CI)).
Table 3. Results of the SW-MLR analyses (Dependent Variable is cooling intensity (CI)).
Dependent VariableVariablesUnstandardized CoefficientsStandardized Coefficients (β)Sig.VIF
βStd. Error
Beijing(Constant)−3.7511.137 0.001
Area0.0260.0070.3760.0001.280
NDVI11.3113.5710.2940.0021.205
NGP2.4521.1500.2030.0361.269
R2 = 0.670; Adjusted R2 = 0.450
Shanghai(Constant)−5.5620.558 0.000
NDVI15.7691.7090.4880.0001.086
Area0.0260.0050.2840.0001.223
NGP3.4560.8270.2290.0001.163
R2 = 0.729; Adjusted R2 = 0531
Tianjin(Constant)−5.8630.579 0.000
NDVI16.6221.7800.5000.0001.092
Area0.0250.0050.2780.0001.225
NGP3.5310.8330.2340.0001.158
R2 = 0.739; Adjusted R2 = 0.546
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Xu, J.; Yu, Y.; Zhou, W.; Yu, W.; Wu, T. Effects of the Spatial Pattern of Forest Vegetation on Urban Cooling in Large Metropolitan Areas of China: A Multi-Scale Perspective. Forests 2024, 15, 1778. https://doi.org/10.3390/f15101778

AMA Style

Xu J, Yu Y, Zhou W, Yu W, Wu T. Effects of the Spatial Pattern of Forest Vegetation on Urban Cooling in Large Metropolitan Areas of China: A Multi-Scale Perspective. Forests. 2024; 15(10):1778. https://doi.org/10.3390/f15101778

Chicago/Turabian Style

Xu, Jie, Yiqi Yu, Wen Zhou, Wendong Yu, and Tao Wu. 2024. "Effects of the Spatial Pattern of Forest Vegetation on Urban Cooling in Large Metropolitan Areas of China: A Multi-Scale Perspective" Forests 15, no. 10: 1778. https://doi.org/10.3390/f15101778

APA Style

Xu, J., Yu, Y., Zhou, W., Yu, W., & Wu, T. (2024). Effects of the Spatial Pattern of Forest Vegetation on Urban Cooling in Large Metropolitan Areas of China: A Multi-Scale Perspective. Forests, 15(10), 1778. https://doi.org/10.3390/f15101778

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop