Enhancing precision flood mapping: Pahang's vulnerability unveiled
Tahmina Afrose Keya1,2, Siventhiran S Balakrishnan3, Maheswaran Solayappan4
1Department of Community Medicine, AIMST University, Bedong, Kedah, Malaysia.
This study developed a flood susceptibility map for Pahang, Malaysia, using ensemble machine learning. Rainfall and elevation were identified as key factors, highlighting areas at high risk for improved flood management.
Area of Science:
- Environmental Science
- Geospatial Analysis
- Machine Learning
Background:
- Pahang, Malaysia, faces severe annual floods causing extensive damage.
- Effective flood management requires accurate susceptibility mapping.
Purpose of the Study:
- To create a flood susceptibility map for Pahang using an Ensemble Machine Learning (EML) algorithm.
- To identify and rank factors influencing flood susceptibility in the region.
Main Methods:
- Utilized geographic information system (GIS) and ArcGIS software.
- Employed the Random Forest (RF) model and Feature Selection (FS) for analysis.
- Analyzed nine geospatial factors including rainfall, elevation, and land use/land cover (LULC).
Main Results:
- Rainfall and elevation were identified as the most significant flood-influencing factors.
- 'Very high' and 'high' flood susceptibility classes covered 37.1% and 26.3% of the area, respectively.
- Eastern, Southern, and central Pahang were identified as high-risk flood zones.
Conclusions:
- Flood susceptibility mapping using EML provides a robust tool for risk reduction.
- The findings can inform targeted flood management strategies in vulnerable areas.
- Advanced methods enhance community resilience to flooding events.
Related Concept Videos
Design Example: Analyzing Capacity Contours for Flood Risk Assessment
Applications of GIS: Disaster Management and Emergency Response
Manipulation and Analysis
Common Leveling Mistakes and Errors
Selected Data About Geographic Locations
Design Example: Identifying the Locations of Monuments in the Field Using Global Positioning System Device


