Ms Wonsill Hwang (p), prof. Jung Eun Kang

Compact-city strategies are central to climate mitigation, yet densification can simultaneously amplify urban losses from floods and heatwaves by expanding impervious surfaces, intensifying heat storage, and concentrating exposure along mobility networks. This paper reframes urban compactness as a multi-objective optimization problem and identifies two decision-critical thresholds: (i) a structural threshold at which marginal mitigation gains diminish while hazard impacts accelerate, and (ii) a Pareto “knee” that marks a Pareto-efficient stopping point between transport decarbonization and climate-risk reduction.
We assemble a balanced panel of 229 South Korean municipalities (2014–2023) and operationalize a time-comparable Korean Compactness Index (K-CI) based on the “3Ds + accessibility” paradigm—capturing density (population, employment, floor-area ratio), land-use diversity (entropy mix), connectivity (intersection density), and transit accessibility (station-area share; with bus-stop density as a robustness proxy). To avoid dilution by non-urban land, compactness components are measured over urbanized areas. Mitigation performance is proxied by per-capita road-transport CO₂ emissions (cross-validated with per-capita VKT), while adaptation outcomes capture realized impacts: real per-capita flood damages and heat-related illness per 100,000 residents. The model conditions explicitly on hazard triggers—extreme-rainfall days (≥80 mm/day) and heatwave days (≥33°C)—and baseline geographic exposure (elevation/lowland share), while controlling for socioeconomic vulnerability and asset exposure (aging structure, GRDP, land values).
Non-linear effects are estimated using municipality- and year-fixed-effects panel threshold regression with bootstrap tests, allowing for multiple endogenous thresholds. Inference is robust to cross-sectional dependence via Driscoll–Kraay standard errors. As a quasi-experimental mechanism check, we exploit the COVID-19 period (2020–2022) as an exogenous shock to activity flows; the persistence of compactness gradients under suppressed mobility is consistent with built-form “stock” mechanisms rather than transient activity effects.
Finally, we simulate the bivariate response space implied by the estimated functions to recover Pareto frontiers and locate the knee point using curvature-based detection. Comparing the Pareto knee with the structural threshold reveals a policy-relevant gap between social optima and physical limits, enabling an operational zoning logic (safe/warning/danger) for resilient, climate-smart densification strategies.

Dr. Hiroaki Shirayanagi (p), Dr. Yukisada KITAMURA, Mr. Ryo YAMASHITA, Mr. Yamato ASADA, Mr. Kanata IWATA

　This study aims to propose a rapid and efficient evacuation plan against large-scale tsunami inundation caused by a megathrust earthquake along the Nankai Trough. The study area is Kasugade-Kita 1-chome in Konohana Ward, Osaka City. A three-dimensional urban model capable of calculating evacuation time was constructed using the national 3D city model “PLATEAU” and digital road network data.
　First, based on tsunami inundation assumptions, non-wooden buildings with accessible rooftops were selected as evacuation sites. The accommodation capacity of each building was estimated from rooftop area and required space per person. The results revealed that the existing designated tsunami evacuation building alone cannot accommodate the entire daytime population. Therefore, the installation of a tsunami evacuation tower in Konohana Park was proposed to compensate for the shortage.
　Next, evacuation completion time was calculated through network analysis based on the 3D urban model under multiple evacuation behavior patterns, considering both scenarios with and without capacity constraints. The results indicate that distributed evacuation, which actively utilizes vertical evacuation to reinforced concrete buildings, is the most rapid and efficient strategy.
　Furthermore, tsunami propagation was reproduced using fluid simulation in Blender and visualized within the detailed 3D urban model. This approach enables intuitive understanding of tsunami dynamics and enhances disaster preparedness materials.
　These findings demonstrate that integrating 3D urban models with evacuation simulation provides an effective method for quantitatively evaluating evacuation strategies and verifying the effectiveness of distributed evacuation planning.
　Furthermore, tsunami propagation was reproduced using fluid simulation in Blender and visualized within the detailed 3D urban model. This approach enables intuitive understanding of tsunami dynamics and enhances disaster preparedness materials.
　These findings demonstrate that integrating 3D urban models with evacuation simulation provides an effective method for quantitatively evaluating evacuation strategies and verifying the effectiveness of distributed evacuation planning.

David Castells-Quintana, Marcos Sanso-Navarro, Dr. María Vera-Cabello (p)

This paper investigates the long-run determinants of cross-country differences in the equality of city size distributions, with a particular emphasis on the role of climate-related factors. Using a harmonized global definition of cities and a balanced panel dataset covering the period 1975–2025, we estimate country-level Pareto coefficients as our main indicator of the degree of equality in national urban structures. To account for model uncertainty in a context characterized by multiple plausible explanatory dimensions, we employ a Bayesian model averaging framework that jointly considers geographic, socioeconomic, and institutional factors. The results reveal that climatic conditions are systematically and robustly associated with the internal configuration of national urban systems. Specifically, higher mean temperatures are positively correlated with more equal city size distributions, while greater average precipitation is linked to more uneven urban hierarchies, even after controlling for both country- and time-specific effects. Conversely, drought intensity does not emerge as a consistent determinant across model specifications. When focusing on urbanization rates instead of city size distributions, precipitation remains the only climatic variable displaying a stable and statistically significant association. Overall, the findings suggest that persistent climatic patterns shape not only the scale of urbanization but also the spatial concentration of populations within countries.