Projecting groundwater flood risk in a lowland karst system under future climates

Area of Science:

• Hydrology and Climate Science
• Environmental Modeling
• Geoscience

Background:

• Lowland karst catchments are sensitive to climate variability.
• Galway Bay faces increasing flood risks influenced by rainfall and tides.
• Understanding future flood dynamics is crucial for coastal adaptation.

Purpose of the Study:

• To assess future flood dynamics in a lowland karst catchment under climate change scenarios.
• To project changes in rainfall, flood volumes, and extreme flood events.
• To evaluate the effectiveness of integrated statistical and machine-learning models for flood risk assessment.

Main Methods:

• Utilized a Bayesian neural network (BNN) model calibrated with historical data (1980-2015).
• Employed CORDEX climate model projections under RCP 4.5 and RCP 8.5 scenarios.
• Applied wavelet coherence and extreme value analysis to assess flood-rainfall relationships and event frequency.

Main Results:

• Projected significant increases in rainfall and flood volumes under both RCP 4.5 and RCP 8.5.
• Observed higher rainfall-flood coupling under RCP 8.5, indicating amplified flood responses.
• Found that a 100-year storm event could become a 1-in-16-year event under RCP 8.5 by 2080-2099.

Conclusions:

• High-emission pathways substantially amplify flood magnitudes and frequencies.
• Integrated statistical and machine-learning frameworks are effective for robust flood-risk assessment.
• Findings underscore the need for proactive climate adaptation planning in vulnerable coastal areas.