Green Space Cooling Effect and Contribution to Mitigate Heat Island Effect of Surrounding Communities in Beijing Metropolitan Area

Abstract

With the rapid process of urbanization and global warming, many metropolises are vulnerable to high temperatures in summer, threatening the health of residents. However, green spaces can generate a cooling effect to mitigate the urban heat island effect in big cities. They can also help to improve the living quality and wellbeing of surrounding residents. In this paper, we utilized the radiative transfer equation algorithm, k-means clustering algorithm, big data crawling, and spatial analysis to quantify and map the spatial distribution, cooling capacity, and cooling contribution for surrounding communities of 1,157 green spaces within Beijing Fifth Ring Road, a typical metropolitan area. The findings showed that (1) the area proportion of the heat island in the study area is larger than that of the cooling island. Accounting for only about 30% area in the study area, the green spaces reduce the average land surface temperature by 1.32°C. (2) The spatial features of green space, such as area and shape complexity, have a significant influence on its cooling effect. (3) Four clusters of green spaces with specific spatial features and cooling capacity were identified. And there were differences among these clusters in green space cooling contribution for the surrounding communities. (4) The differences in green space cooling contribution also existed in different urban zones. Specifically, the middle zone performed significantly better than the inner and outer zones. (5) We furthered in finding that some green spaces with medium and high cooling contributions need to improve their cooling capacity soon, and some green spaces with low cooling contributions or no contributions have a good potential for constructing new communities in the future. Our study could help planners and government understand the current cooling condition of green spaces, to improve their cooling capacity, mitigate the urban heat island effect, and create a comfortable and healthy thermal environment in summer.

Keywords: cooling contribution; cooling effect; green space; metropolitan area; resident health; urban thermal environment.

Figure 1

Location of the study area in Beijing metropolitan area, China (A–C). The Landsat8 remote sensing image (D) is displayed in false color mode, which is composed of band 5 (near-infrared), band 4 (red), and band 3 (green).

Figure 2

Spatial distribution of LST (A) and cold-hot islands at different levels (B) in the study area. Where ALST is the average LST; std is the standard deviation of LST in the study area.

Figure 3

Spatial distribution of land use in the study area (A), green spaces selected as the study objects (B).

Figure 4

Average LST schematic diagram of green space and surrounding impervious surface (A), the spatial distribution of 1,157 green spaces with different degrees of GCI (B).

Figure 5

Relationship between dependent (GCI) and independent [GA: (A), GSI: (B)], relationship between dependent (GCA) and independent [GA: (C), C_G: (D), GSI: (E)], relationship between dependent (GCG) and independent [GA: (F), GSI: (G)], relationship between dependent (L_MAX) and independent [GSI: (H)]. The regression line using the ordinary least squares method is presented in red. The regression models were selected among linear, logarithmic, and power functions with the largest R² value.

Figure 6

Spatial distribution of 4 clusters green spaces in different cooling effect (A), Radar diagrams of landscape and cooling metrics for 4 clusters green spaces (B–E).

Figure 7

Spatial distribution of green spaces with different GCC levels (A), examples of green spaces with different GCC levels (B–E).

Figure 8

The proportion of green space with low, medium, and high GCI in different GCC groups (A), the proportion of green space serving low, middle, and high population density in different GCC groups (B).

Figure 9

The box diagram of GCC distribution in different urban zones (A), proportion of green spaces with different levels of GCC in different zones (B). Hollow circles ° represent mild outliers; asterisks * represent extreme outliers.

Figure 10

The box diagram of GCC distribution in different clusters (A), the proportion of green spaces with different levels of GCC in different clusters (B). Hollow circles ° represent mild outliers; asterisks * represent extreme outliers.