| Title |
Urban Microclimate Dynamics Influenced by Urban Morphological and Architectural Characteristics - Evidence from High Pedestrian-Flow Areas in Seoul Using S-DoT Data |
| Authors |
정혜진(Jeong, Hyejin) ; 정윤남(Jeong, Yunnam) |
| DOI |
https://doi.org/10.38195/judik.2025.10.26.6.65 |
| Keywords |
도시 미기후; 도시 및 건축 형태; SUR; 다변량선형혼합효과모형; S-DoT Urban Microclimate; Urban and Architectural Form; Seemingly Unrelated; Regression; Multivariate linear mixed-effects Model; Smart Seoul Data of Things (S-DoT) |
| Abstract |
This study examines associations between urban form and microclimate in Seoul’s high-pedestrian-flow areas using multivariate and hierarchical models. Around ten S-DoT sites, 1km×1km areas were gridded at 100 m (N=1,000) and integrated with Landsat-8 LST, S-DoT near-surface air temperature, relative humidity, and wind speed, plus GIS-derived predictors (road ratio, elevation, slope, building coverage ratio, floor-area ratio, building height, building count, BVG = NDBI ? NDVI). We estimated SUR to account for cross-equation error covariance and an MLMM with site-level random effects. In SUR, LST was positively associated with BVG, building coverage, building count and negatively with elevation, slope, building height, indicating higher LST in low-rise, high-density, built-dominant grids. Other indicators showed weaker form sensitivity, though building height was generally negative for LST/air temperature and positive for humidity/wind. The MLMM yielded ICC?0.76, evidencing strong site-level clustering; random-effects covariances suggested contrasting hot-dry-windy versus cool-humid-calm regimes. Findings support climate-adaptive urban design-balancing urbanization and greening, using height-setback-street geometry for shading and ventilation, and mitigating thermal exposure along major roads. |