Researchers Develop New Method for Bathymetric Mapping Using Machine Learning

Strategic Insights -

Researchers from Istanbul Technical University have developed a new approach to bathymetric mapping using machine learning, which enables cost-effective and efficient mapping of underwater terrain. Traditionally, shipborne echosounders are used for bathymetry, but they are costly and inefficient in shallow waters. The new method leverages emerging technologies such as satellite imagery, drones, and spaceborne LiDAR to enhance bathymetric mapping.

Key Takeaways:

  • The research integrates multi-sensor datasets, including Sentinel-2, Gokturk-1, and aerial imagery, to improve spatial coverage, resolution, and accuracy over traditional methods.
  • The study uses data fusion and machine learning algorithms, such as Random Forest and Extreme Gradient Boosting, to enhance bathymetric inference.
  • The research focuses on bathymetric modeling in coastal and inland waters, using the Gulbahce Bay in Izmir, Turkey, as the study area.
  • The method achieved A1 level accuracy for 0-10 m depth intervals, A2/B accuracy for 0-15 m depth intervals, and C level accuracy for 0-20 m depth intervals, as assessed using IHO CATZOC standards.
  • The research aims to improve bathymetric mapping in shallow waters, where traditional methods are limited by spatial coverage and accessibility.

Statistics:

  • The research used multi-sensor datasets with varying resolutions and sensor characteristics to enhance bathymetric mapping.
  • The study achieved A1 level accuracy for 0-10 m depth intervals, which is a significant improvement over traditional methods.
  • The research used data fusion and machine learning algorithms to enhance bathymetric inference, achieving A2/B accuracy for 0-15 m depth intervals and C level accuracy for 0-20 m depth intervals.
  • The study area was the Gulbahce Bay in Izmir, Turkey, which featured varying water depths and bathymetric features.

Sources:

  • "Sensor Synergy in Bathymetric Mapping: Integrating Optical, LiDAR, and Echosounder Data Using Machine Learning." Remote Sensing, 2025, 17(16), 2912. doi: 10.3390/rs17162912
  • "New Machine Learning Study Findings Have Been Reported from Istanbul Technical University (Sensor Synergy in Bathymetric Mapping: Integrating Optical, LiDAR, and Echosounder Data Using Machine Learning)". Information Technology Newsweekly, September 9, 2025.