Mapping marine litter using UAS on a beach-dune system: a multidisciplinary approach

Area of Science:

• Environmental Science
• Remote Sensing
• Coastal Geomorphology

Background:

• Marine litter, primarily plastics, poses a significant global environmental challenge in coastal zones.
• Conventional monitoring methods (in-situ visual census) are labor-intensive and time-consuming, necessitating innovative solutions.

Purpose of the Study:

• To develop and present an automated Unmanned Aerial System (UAS)-based procedure for mapping marine litter in beach-dune systems.
• To demonstrate the integration of photogrammetry, geomorphology, machine learning, and hydrodynamic modeling for enhanced marine litter detection.

Main Methods:

• Processing Unmanned Aerial System (UAS) imagery using a multidisciplinary framework.
• Employing photogrammetry to create very high-resolution orthophotos.
• Utilizing a random forest machine learning algorithm for automated screening and characterization of marine litter on beach and dune areas.

Main Results:

• The automated UAS-based mapping achieved a 75% F-test score compared to manual identification of marine litter.
• Marine litter distribution was correlated with beach slope and water level dynamics.
• The study identified key environmental parameters for optimizing UAS deployment and post-processing for beach litter mapping.

Conclusions:

• The developed UAS-based framework offers an efficient and automated solution for monitoring marine and coastal pollution.
• This technology can support scientists, engineers, and policymakers in environmental management.
• The findings provide a basis for optimizing and automating beach clean-up operations.