Date of Award
Doctor of Philosophy (PhD)
Systematic and remote imaging techniques capable of detecting fluid density anomalies will allow for effective scientific sampling, improved geologic and biologic spatial understanding and analysis of temporal changes. This work presents algorithms for detection of anomalous fluids using an ROV-mounted high resolution imaging suite, specifically the structured light laser sensor and 1350kHz multibeam sonar system.
As the ROV-mounted structured light laser sensor passes over areas of active the turbulent nature of the density anomaly causes the project laser line, imaged at the seafloor, to blur and distort. Detection of this phenomena was initially presented in 2013 with significant limitations including false positive results for active venting. Advancements to the detection algorithm presented in this work include intensity normalization algorithms and the implementation of a support vector machine classification algorithm. Results showing clear differentiation between areas of plain seafloor, bacteria or biology, and active venting are presented for multiple hydrothermal vent fields.
Survey altitudes and the direction of travel impact laser data gathered over active vent sites. To determine the implications of these survey parameters, data collected over a single hydrothermal vent at three altitudes with four headings per altitude are analyzed. Changing survey geometry will impact the resolution and intensity of the laser line images, therefore, normalization and processing considerations are presented to maintain signal quality. The spatial distribution of the detected density anomaly will also be discussed as it is impacted by survey range and vehicle heading.
While surveying hypersaline brine pools the observed acoustic responses from the 1350kHz high frequency multibeam sonar system indicate sensitivity to changes in acoustic impedance and therefore the density of a fluid. Internal density stratification was detected acoustically, appearing as multiple returns within a single water column image, and confirmed using a reel-mounted CTD. Additional acoustic returns correspond to rising bubbles, the surface of the brine pool, and the seafloor. This work allows for a systematic and complete reconstruction of the internal density structure of a brine pool.
Development of sensors and algorithms capable of efficiently gathering the data necessary to establish a comprehensive understanding of the density variations within an area will improve the geologic understanding of a vent field or brine pool and allow for associations to be made between fluid flow and associated geological and biological activity.
Smart, Clara, "Detection of Fluid Density Anomalies Using Remote Imaging Techniques" (2016). Open Access Dissertations. Paper 530.