Bagyong Kristine (TS Trami) in bicol, Philippines: Flood risk forecasting, disaster risk preparedness predictions and lived experiences through machine learning (ML), econometrics, and hermeneutic analysis

This work was conducted just two days after the onslaught of Bagyong Kristine (TS Trami) in October 2024 that flooded the Bicol Region, Philippines. We combined quantitative approaches (machine learning and econometrics) and qualitative techniques (hermeneutic phenomenological, narrative, thematic, and anthropology-at-home) to forecast future flood risks, predict disaster risk preparedness (DRP), and explore the lived experiences of households in Camarines Sur . We utilized both secondary and primary data to offer more robust analysis to support local government, uplift flooded localities, and advance scientific communities. Coastal communities of San Jose are particularly at risk, with varying flood susceptibility levels. Support Vector Machine (SVM) was used to forecast flood risks indicating moderate-to-high risks. The study explores multidimensional factors influencing DRP, floods, and calamity experiences utilizing significant indicators as a priori predictors in ML runs. Improved housing, income, and digital access are associated with higher disaster risk preparedness (DRP). Conversely, living in non-concrete housing, lacking access to basic services, experiencing poverty, and engaging in informal livelihoods elevate risk levels. Experiences with floods are linked to place of residence, water and sanitation, garbage collection, and education. Calamity experiences are associated with housing, access to amenities, informal livelihoods, and preparedness. ML predictions suggest that SVM and Random forests yield the best performance in predicting DRP. Hermeneutic analyses offer valuable and fresh insights for policymaking. It has been revealed that the region is very accustomed to typhoons but not to severe flooding. Geographical vulnerabilities near water bodies underscore the constant threat of floods, emphasizing the mix of resilience, faith, fear, and community solidarity among respondents. By blending scientific methods with indigenous wisdom, a comprehensive analysis was conducted to develop culturally integrated policies. The unexpected challenges faced reveal unpreparedness for extreme rainfall events. Community cooperation, government accountability in disaster management, and environmental conservation efforts are emphasized, advocating for proactive measures, accurate forecasting, and sustainable practices to reduce flooding disasters.