A historical data curation (HDC) framework integrated with GIS-based multi-criteria decision analysis to inform flood risk prioritization in data-scarce regions

Data-scarce regions lack reliable long-term gauging records and standardized flood databases which constrains classical flood risk assessments and fail to adequately reflect local contexts. Global initiatives emphasize investments in non-traditional data curation to improve the representation of vulnerability and to transform how disaster risk is addressed, complementing systematic flood risk assessment strategies. In this study, we present the outcomes of past flood information curation from newspapers and archives to build a national flood database and evaluate the basin-scale flood risk. We particularly propose a reproducible Historical Data Curation– Multi-Criteria Decision Analysis (HDC–MCDA) framework that transforms qualitative historical archives into standardized quantitative Curated Historical Flood Descriptors (CFD) and integrates them with conventional GIS-based MCDA flood risk analysis. We curated a database of 700 flood events (1955–2020) in Lebanon with documented description and impacts. Two CFD, Historical Flood Occurrence ( Ne ) and Historical Flood Severity ( Sev ), were developed through text processing and integrated in the MCDA analysis. The study was applied for 16 basins, where hazard and vulnerability are represented by conventional morphological, hydrological and land-use descriptors extracted from available data. Across the studied scenarios, HDC-MCDA led to notable shift in flood risk level, suggesting that HDC captures the realized impact and recalibrates flood risk prioritization by perceived impact information that is often not reflected by scarce hazard and vulnerability data. The study provides a reproducible framework for “soft” documentary data curation and integration into risk assessments, informing prioritization of mitigation and early warning efforts in data-scarce countries.