File Download

  • Find it @ UNIST can give you direct access to the published full text of this article. (UNISTARs only)
Related Researcher

임정호

Im, Jungho
Intelligent Remote sensing and geospatial Information Science Lab.
Read More

Views & Downloads

Detailed Information

Cited time in webofscience Cited time in scopus
Metadata Downloads

Nationwide Local Climate Zone Mapping and Comparative Assessment of Urban Heat Island Effects in South Korea

Author(s)
Park, SujeongKim, Hui-JungKim, YoungseokLee, YeonsuCho, DongjinYoo, CheolheeBae, DukwonLee, SiwooIm, Jungho
Issued Date
2025-12
DOI
10.7780/kjrs.2025.41.6.20
URI
https://scholarworks.unist.ac.kr/handle/201301/89621
Citation
KOREAN JOURNAL OF REMOTE SENSING, v.41, no.6, pp.1207 - 1223
Abstract
The pervasive expansion of impervious surfaces in urban areas alters surface thermal properties, resulting in the urban heat island (UHI) effect. Elevated urban temperatures constitute a considerable threat to human health by intensifying thermal stress and exacerbating air pollution. The magnitude and characteristics of UHI are highly sensitive to seasonal, diurnal, morphological, and geographic conditions. Although canopy-level temperature provides a more relevant indicator of human thermal exposure than surface-based measures, most previous assessments have relied on satellite-derived land surface temperature (LST). Motivated by these limitations, this study implemented a nationwide assessment of canopy-level UHI variability across 157 cities in South Korea, combining Sentinel-2 imagery, Korea Meteorological Administration (KMA) gridded air temperature products as a canopy level measure complementing LST, and a deep-learning local climate zone (LCZ) mapping framework. Using a transfer-learning strategy applied to satellite imagery, we produced a novel nationwide LCZ map that captures the spatial heterogeneity of urban and natural landscapes, with an overall accuracy of 0.78. Leveraging spatially continuous air temperature datasets, we found that canopy-level UHI exhibits significantly higher values at night (0.05°C in summer and 0.19°C in winter) compared to daytime UHI intensity (0.03°C in summer and 0.02°C in winter). These effects were amplified in areas with denser and taller building morphologies, highlighting the crucial role of the built environment in shaping urban thermal conditions. Geographical analyses revealed a distinct association between summer UHI intensity and coastal proximity in both day and night (Pearson’s r > 0.2 and 0.3, respectively; p < 0.005), while this pattern was attenuated in winter. Our findings provide a valuable scientific basis for evaluating human relevant thermal environments across cities and offer essential insights to inform urban planning and climate-resilient design strategies.
Publisher
KOREAN SOC REMOTE SENSING
ISSN
1225-6161
Keyword
Local climate zones Transfer learning Convolutional neural networks Urban heat island Air temperature Urban climate Remote sensing

qrcode

Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.