How do spatial scale-dependent urban and forest areas influence regionally seasonal temperatures in mid-latitude South Korean cities?

Abstract

This study aims to investigate the effects of urban and forest areas measured in three dimensions on seasonal temperature over forty years in South Korean cities. We measure the urban and forest areas at the city, neighborhood, and spatially clustered levels in four periods every ten years. Using Hot Spot Analysis (Getis-Ord Gi*), this study detects the spatially clustered urban and forest areas. We establish a multilevel regression model to explore the relationship between urban and forest areas measured in three dimensions, as well as seasonal temperatures. The study shows that while spatially clustered urban and forest areas have consistent associations with the four seasonal temperatures, urban and forest areas at the city scale have different associations with the seasonal temperature, depending on the season. When spatially clustered urban areas increase by 10 km², four seasonal temperatures increase by about 0.0016-0.0067 Celsius degree (°C); on the other hand, when spatially clustered forest areas increase by 10 km², four seasonal temperatures decrease by about 0.0001-0.0016 °C. At the neighborhood level, urban and forest areas are negatively associated with the four seasonal temperatures. The results of this study can be utilized by urban planners and policymakers to establish land use planning or policy by providing evidence of whether land use plans should be established and at what scales to manage regional thermal environments. To alleviate seasonal warming, we recommend increasing forest areas at the neighborhood and spatially clustered levels and controlling the size of spatially clustered urban areas.

Keywords: Forest areas; Regional climate change; Seasonal temperatures; Spatial clusters; Urban expansion.