In the context of urbanization, the issue of Urban Heat Islands (UHI) has gradually come into focus. UHI phenomena pose multiple hazards to humans. However, in response to these dangers, developing Urban blue green space (UBGS) is considered an effect...
In the context of urbanization, the issue of Urban Heat Islands (UHI) has gradually come into focus. UHI phenomena pose multiple hazards to humans. However, in response to these dangers, developing Urban blue green space (UBGS) is considered an effective means to mitigate heat island effects. Among these, city parks, as the main green spaces in urban areas, provide a cooling effect (CE) that directly contributes to the reduction of UHI impacts. Most research on urban greenery and UHI mitigation focuses on internal factors within city parks, but factors external to the parks influencing the CE have not been thoroughly studied. To address this gap, we focused on external factors affecting the cooling effect indices of city parks (such as the area and ratio of surrounding small-scale green spaces, the Landscape Pattern Index (LPI) of these areas, and the Normalized Difference Vegetation Index (NDVI) of the surrounding areas). We selected 50 city parks from 256 in Jeonju, South Korea, as our study subjects. Using Landsat 8 satellite imagery, we calculated the city's Land Surface Temperature (LST) map through the Radiative Transfer Equation method (RTE). We then quantified the current thermal landscape distribution of Jeonju based on the LST, studied the correlations between LST and surface parameters (NDVI, NDBI, MNDWI), and analyzed temperature changes within 600m buffer zones around the parks. We quantified the cooling effects of all parks, examined the coupling between park cooling effects and surrounding small-scale green spaces, and established correlations through Pearson or Spearman correlation tests to identify effective external influencing factors. We also developed simple linear regression models to determine the extent of these impacts, yielding the following results:
Firstly, in Jeonju City, the heat island effect is more severe in the Deokjin District than in the Wansan District, with mountainous and water areas being more effective at reducing heat islands than other land use types. Additionally, mountainous areas in Jeonju City have a stronger heat island reduction effect compared to water bodies.
Secondly, both NDVI and NDWI show a linear negative correlation with LST, while NDBI shows a linear positive correlation with LST. This also indicates that increasing blue-green spaces in cities can significantly reduce urban heat island effects, as well as the enhancing effect of buildings on urban heat islands.
Thirdly, among all the city parks in Jeonju City, Kirin Park, Aramgil Park, Dream Village Park, and Dorangil Park have the best cooling effects and should be further studied as model parks for reducing heat island effects. The average cooling effect of Jeonju City's urban parks is as follows: PCD (Park Cooling Distance) = 207.94m, PCI (Park Cooling Intensity) = 1.2436°C, PCG (Park Cooling Gradient) = 6.1204°C/km.
Fourthly, Fourth, we found that large parks are more susceptible to external factors compared to smaller parks. For large park clusters, we discovered that increasing the small-scale green spaces within the 150-180 meter buffer zone can significantly enhance both PCD and PCI. Additionally, the landscape pattern indices of small-scale green patches within the 0-300 meter buffer zone (such as CA, PLAND, and COHESION) also affect PCD to varying degrees, with CA and PLAND being significantly correlated with PCI. Increasing CA and PLAND both lead to an increase in PCI.We also confirmed that the NDVI in the surrounding areas of urban parks affects the cooling effect of the parks. In the 0-60 meter and 90-180 meter buffer zones, increasing NDVI significantly extends PCD. The regions that significantly affect PCI are 0-60 meters and 120-270 meters, with the 210-240 meter region being the most efficient. Each 0.01 increase in NDVI can increase PCI by 0.325°C.For small park clusters, the landscape pattern indices that significantly affect PCD are LSI and AI. Increasing LSI can significantly enhance PCD, while increasing AI can lead to a significant reduction in PCD. The landscape indices that significantly enhance PCI are NP, PD, and LSI. Additionally, within the 0-60 meter buffer zone, NDVI is significantly positively correlated with PCD.
These research results provide valuable directions for analyzing external factors of urban green spaces that reduce urban heat islands.