Satellite-derived bathymetry from wave kinematics using radar.

Type of project

M.Sc.

Contact

Nicklas Simonsen (nisi@dhigroup.com)

Topic

Develop a robust machine learning (ML) model that uses satellite-based radar (Sentinel-1) observations of ocean wave kinematics (e.g., wave height, wavelength, phase velocity) to estimate water depth in the intermediate range (10–50m).

This range is particularly challenging due to the limitations of both traditional optical and deep-water wave inversion methods.

Relevance to DHI

• Bathymetric coverage in areas beyond the reach of traditional optical-based methods, but still within the zone of high uncertainty in global models like GEBCO.
• Satellite-based methods allow for frequent updates to bathymetric data, enabling models to reflect changes due to sedimentation, erosion, or human activities.
• Collecting bathymetric data via shipborne surveys or LiDAR is costly and time-consuming. Satellite-based ML methods can reduce reliance on fieldwork, lowering operational costs and enabling rapid, large-scale assessments.
• The method can be applied globally, allowing DHI to scale hydrological modelling services to new regions without prohibitive data collection costs.

What is the student's contribution

The student will explore machine learning solutions for optimizing the parameterization of retrieving bathymetry from wave kinematics. Several works have already explored the potential of radar-based satellite-derived bathymetry but none (that we are aware) have looked at optimizing the parameterization using machine-learning. Solving this would allow for a scalable solution.

References

• Automated High-Resolution Bathymetry from Sentinel-1 SAR Images in Deeper Nearshore Coastal Waters in Eastern Florida
• Bathymetry derived from Sentinel-1 Synthetic Aperture Radar data
• The feasibility of assessing swell-based bathymetry using SAR imagery from orbiting satellites