Wide-Area Seafloor Characterization Using Autonomous Underwater Vehicles

Deep ocean industrial and hazardous waste disposal was a pervasive global practice near coastlines in the 20th century. Uncertainty in the quantity, location, and contents of the discarded materials at historic dump sites pose ongoing risks to marine ecosystems and human health.  This talk will describe observations and analytic techniques developed to characterize a deep ocean chemical dump site using robotic vehicles. Autonomous underwater vehicles (AUVs) and a remotely operated vehicle (ROV) were used to complete a wide area survey in San Pedro Basin, CA, a known dumpsite for the chemical Dichlorodiphenyltrichloroethane (DDT) and derivatives.  The dump site is thought to be a missing link in the budget of DDT which has been found in higher-than-normal concentrations in Southern California marine life. Statistical, spectral, and machine learning techniques applied to sidescan sonar data are used to characterize the seafloor and an object classification algorithm is developed using optical imagery of ground truth targets.  We interpret spatial patterns of both the seabed and debris field to inform chemical and biological studies tracing DDT through the marine food web. Data science techniques combined with the survey capabilities of AUVs provide a framework for efficient mapping and characterization of uncharted deep-water disposal sites.

Speaker: Sophia Merryfield

Attend at Hopkins Boatworks or Mitchell 350/372 on Stanford's Campus