Japan’s Urban-Environmental Exposures
A Tripartite Analysis of City Shrinkage, SAR-Based Deep Learning Versus Forward Modeling in Inundation Mapping, and Future Flood Schemes
Bibliographic Data
| ID | 13112970 |
|---|---|
| Authors | Mohammadreza Safabakhshpachehkenari (0009-0004-2002-7216, Ibaraki University, corresponding author), Hideki Tsubomatsu (Ibaraki University), Hideyuki Tonooka (0000-0002-4866-5955, Hitachi (Japan)) |
| Year | 2025 |
| Volume | 9 |
| Issue | 3 |
| Pages | 71-71 |
| Publication date | 2025-03-05 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Urban Science (JOURNAL) |
| Journal identifiers | ISSN: 2413-8851 • E-ISSN: 2413-8851 |
| Publisher | MDPI AG (PUBLISHER • IT) |
| DOI | 10.3390/urbansci9030071 |
| OpenAlex | W4408161225 |
| Language | EN |
| References cited | 36 |
This study investigates how urban decline and intensifying flood hazards interact to threaten Japan’s urban environments, focusing on three main dimensions. First, a fine-scale analysis of spatial shrinkage was conducted using transition potential maps generated with a maximum entropy classifier. This approach enabled the identification of neighborhoods at high risk of future abandonment, revealing that peripheral districts, such as Hirakue-cho and Shimoirino-cho, are especially susceptible due to their distance from central amenities. Second, this study analyzed the 2019 Naka River flood induced by Typhoon Hagibis, evaluating water detection performance through both a U-Net-based deep learning model applied to Sentinel-1 SAR imagery in ArcGIS Pro and the DioVISTA Flood Simulator. While the SAR-based approach excelled in achieving high accuracy with a score of 0.81, the simulation-based method demonstrated higher sensitivity, emphasizing its effectiveness in flagging potential flood zones. Third, forward-looking scenarios under Representative Concentration Pathways (RCP) 2.6 and RCP 8.5 climate trajectories were modeled to capture the potential scope of future flood impacts. The primary signal is that flooding impacts 3.2 km2 of buildings and leaves 11 of 82 evacuation sites vulnerable in the worst-case scenario. Japan’s proven disaster expertise can still jolt adaptation toward greater flexibility. Adaptive frameworks utilizing real-time and predictive insights powered by remote sensing, GIS, and machine intelligence form the core of proactive decision-making. By prioritizing the repositioning of decaying suburbs as disaster prevention hubs, steadily advancing hard and soft measures to deployment, supported by the reliability of DioVISTA as a flood simulator, and fueling participatory, citizen-led ties within a community, resilience shifts from a reactive shield to a living ecosystem, aiming for zero victims
Archaeology · Environmental planning · Environmental resource management · Flood myth · Geography · Machine learning · Remote sensing · Shrinkage · Computer Science · Environmental Science · Flood Risk Assessment and Management · Land Use and Ecosystem Services · Urbanization and City Planning
Applied Logistic Regression Analysis
U-Net
Why Do Cities Shrink?
Deep learning
Convolutional Neural Network-Based Deep Learning Approach for Automatic Flood Mapping Using NovaSAR-1 and Sentinel-1 Data
Declining Fertility Rates in Japan
Identifying and quantizing the non-linear correlates of city shrinkage in Japan
The development of urban shrinkage discourse and policy response in Japan
Driving factors of urban shrinkage
Urban shrinkage and resurgence in Germany
Shrinking Cities
| Citation velocity | historical |
|---|---|
| Highly cited | No |