Low frequency water level correction in storm surge models using data assimilation.

Area of Science:

• Oceanography and Coastal Engineering
• Numerical Modeling and Data Assimilation

Background:

• Traditional data assimilation (DA) in storm surge modeling shows limited value for rapid surge prediction.
• Existing hydrodynamic models often fail to capture large-scale background water level errors caused by unresolved drivers.

Purpose of the Study:

• To develop and validate a novel, efficient DA scheme to correct for unresolved drivers in storm surge models.
• To improve the accuracy of water level predictions by addressing errors from unmodeled physical processes.

Main Methods:

• Developed an optimal interpolation-based DA scheme utilizing coastal water level gage observations.
• Dynamically corrected for unresolved drivers, including far-field forcing, baroclinic processes, and steric variations.
• Applied the methodology to Hurricane Matthew (2016) for validation.

Main Results:

• The novel DA scheme effectively removed water level residuals, significantly reducing overall surge errors.
• Demonstrated a substantial reduction in surge errors, approximately halving them for Hurricane Matthew.
• The method proved computationally efficient and suitable for both hindcast and forecast applications.

Conclusions:

• The proposed DA method offers a significant advancement in storm surge prediction accuracy.
• It provides a computationally efficient and extensible solution for improving coastal flood modeling.
• The approach effectively addresses limitations in current models by accounting for unresolved physical drivers.