Open-data critical-access screening for compound heat–flood infrastructure risk in low-data cities
Low-data and conflict-affected cities need infrastructure-access screening that uses public data without turning proxies into predictions. We rebuilt a four-city Yemen analysis from archived raw sources using a topology-preserving road graph, WorldPop, SRTM, HydroSHEDS, regional NASA POWER, CHIRPS, UCDP GED and UNOSAT flood observations. Population inside 30-km analytical frames was routed on a 45-km road-and-facility halo to reduce boundary truncation. Network distances increased mapped education remoteness beyond 5 km to 57.81% of population in Taiz and 48.11% in Al Hudaydah. Exact clipping placed 0.59–2.54% of road length within 100 m of mapped waterways. CHIRPS independently indicated higher recent annual rainfall in Taiz (+ 109.0 mm) and Al Hudaydah (+ 78.2 mm), while regional POWER cells exposed substantial within-frame climate variability. A transparent road-link screening priority was evaluated with alternative aggregation, three weight-perturbation families, boundary, snapping and conflict-impedance sensitivities, and 400 link-removal experiments. UNOSAT comparison was restricted to observable cloud-free coverage and did not validate the composite score as an inundation classifier. The framework is therefore an auditable triage tool for field verification, not a flood, closure, travel-time or investment model.
Authors
- Saleem Ahmed Al-Azazi (ORCID: https://orcid.org/0009-0005-1221-1379)
- Farag Qasem
- Aawag Mohsen Alawag
- Abdullateef Ali
Institutions
- Taiz University (YE)
- Universiti Teknologi MARA System (MY)
- Al-Rowad University (YE)
- Universiti Teknologi MARA (MY)
Publication Details
- Journal
- Scientific Reports
- Published
- 2026-11-20
- DOI
- https://doi.org/10.1038/s41598-026-80878-3
- Primary Topic
- Flood Risk Assessment and Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00