Abstract
Single beam sonars can provide valuable acoustic information on the structure of benthic habitats and the contents of the water column. Nominally, acoustic sensors that provide water column data in scientific applications can cost tens or hundreds of thousands of dollars. In contrast, consumer grade fish finders that are mass produced are very inexpensive, costing only tens or hundreds of dollars. Unlocking a fish finder for scientific use could increase access to low-cost sensing methods for coastal communities that are historically underserved. The principal challenge with using a fish finder for benthic habitat classification is that the sonars are generally not interoperable and are often limited to visualization on a display or chart plotter made by the sonar manufacturer. This vendor lock prevents the sonar data, and in particular water column data, from being stored and processed to create mapping products. In this project, “SonarPhony” was developed to provide interoperability software to enable the real-time visualization and logging of water column data from a low-cost fish finder. A machine learning approach was used to demonstrate that the logged data could be used to estimate bottom type and identify the presence of seagrasses. This solution thus provides a low-cost means for both benthic habitat classification and bathymetric mapping.