Exploratory regression modeling for flood susceptibility mapping in the GIS environment

Area of Science:

• Hydrology and Environmental Science
• Geographic Information Systems (GIS)

Background:

• The Awash River basin in Ethiopia's Afar region faces significant and recurring flood risks.
• Understanding flood patterns is vital for effective water resource management and disaster preparedness.

Purpose of the Study:

• To assess and map flood susceptibility in the Awash River basin.
• To identify key factors contributing to flood risk.
• To evaluate the role of spatial planning in mitigating flood impacts.

Main Methods:

• Utilized Geographic Information Systems (GIS) for spatial analysis.
• Employed exploratory regression analysis with eight independent variables (elevation, slope, rainfall, drainage density, land use, soil type, wetness index, lineament density).
• Developed and validated 31 flood susceptibility models, identifying the optimal model (Model 31) based on Adjusted R² and Akaike's Information Criterion.

Main Results:

• Land use was identified as the primary driver of flood susceptibility.
• The best-fit model (Model 31) achieved an Adjusted R² of 0.8.
• Spatial autocorrelation analysis indicated random distribution of model residuals, suggesting model validity.

Conclusions:

• Flood susceptibility in the Awash River basin is significantly influenced by land use patterns.
• The developed GIS-based model provides a robust tool for understanding and predicting flood risk.
• Findings underscore the importance of integrated spatial planning for flood risk reduction in the region.