Instantaneous continental-shelf scale sensing of cod with Ocean Acoustic Waveguide Remote Sensing (OAWRS)
Author(s)
Jain, Ankita Deepak
DownloadFull printable version (34.04Mb)
Other Contributors
Massachusetts Institute of Technology. Department of Mechanical Engineering.
Advisor
Nicholas C. Makris.
Terms of use
Metadata
Show full item recordAbstract
Reported declines in the population of Atlantic cod have a potential to affect long-term ecological balance and the sustainability of the cod fishery along the US northeast coast. These assessments have led to severe fishing cuts over the past few years, have consequently threatened the centuries-old Atlantic cod fishery along the New England coast and put the livelihood of thousands of fishermen at risk. Amidst this fisheries crisis, calls by elected officials, environmental groups and fishing consortiums were made for an Ocean Acoustic Waveguide Remote Sensing (OAWRS) survey of the Gulf of Maine cod stock. Typically, cod stock assessments incorporate data collected from conventional acoustic and trawl line transect surveys that highly undersample the marine environment in space and time and lead to ambiguities in population estimates. The combination of conventional methods and OAWRS techniques, however, has been demonstrated to provide rapid and accurate fish stock assessments over ecosystem-scale areas for other species. In this thesis, the feasibility of accurately surveying cod stocks with OAWRS is theoretically assessed. These theoretical predictions are then experimentally verified by successfully sensing cod with OAWRS over ecosystem scales in the Nordic Seas. Following direct requests by Massachusetts state officials to determine if OAWRS could be used to detect and survey the reported waning cod populations in coastal New England waters, we obtained measurements of typical aggregation densities and occupancy depths of spawning cod in Ipswich Bay from conventional echosounder surveys conducted in Spring 2011. Cod length distributions were also measured from which we estimated the swimbladder resonance frequencies of local cod via a harmonic oscillator model that includes the effects of damping, the cod's swim bladder air volume at a given neutral buoyancy depth as well as changes to this volume for deviations from neutral buoyancy depth. The optimal frequency for OAWRS detection typically corresponds to that where the resonance peak is found. We showed that our theoretical estimates of cod swimbladder resonance matched very well with independent measurements of caged cod resonance from decades old Norwegian data. Using parabolic equation modeling of ocean waveguide propagation, the scattered level of typical spawning cod aggregations was estimated and compared with that from seafloor scattering, which is a typical limiting factor in long range active sensing. Seafloor scattering was estimated via a Rayleigh-Born approach we developed, where the magnitude squared of seafloor scattering amplitude was empirically determined from thousands of measurements made during major OAWRS experiments along the US Northeast coast. It was found that near cod swimbladder resonance (roughly 150-600 Hz), determined from the New England length and depth distribution data, OAWRS was capable of robustly detecting spawning cod aggregations from many tens of kilometers in range with high signal-to- noise ratios (SNRs) greater than 20 dB for typical spawning cod configurations in New England waters. Above the resonance frequency peak, it is possible to detect cod for typical shoaling densities because cod scattering reaches a plateau due to geometric scattering that is above the seafloor scattering trend for typical OAWRS frequencies. Well below the resonance peak, scattering from cod is expected to fall off rapidly and faster than seafloor scattering, and so provides important information about resonance behavior but can be difficult to probe given the very low frequencies involved. This theoretical feasibility study emphasized the need for a low frequency source that spans cod swimbladder resonance and helped demonstrate the potential for use of OAWRS for cod assessments over ecosystem scales. To confirm our theoretical predictions on the OAWRS detection of cod and other keystone fish species, we designed, prepared and conducted a major oceanographic experiment in the Nordic Seas in the Arctic in the winter (February-March) of 2014 using three major research vessels, the US RV Knorr, the Norwegian RV Johan Hjort and the Norwegian FV Artus. The Nordic Seas 2014 experiment was conducted in difficult gale and hurricane force weather conditions along most of Norway's western and northern coast. MIT's OAWRS Source, obtained through a NSF-Sloan MRI grant, spanned the 800-1600 Hz range, and the receiver was ONR's Five Octave Receiver Array (FORA). Unlike the declining trend of cod population in New England waters, cod population in the Nordic Seas has been thriving for many years and is currently at its healthiest recorded state. The experiment period was chosen such that it coincided with the peak spawning period of cod along the coastal Lofoten region in Norway where they congregate in high densities, as well as other keystone species that migrate from the ice-edge to spawn in some of the world's largest mass migrations. In planning, we determined likely spawning grounds for cod, and other keystone species such as capelin, herring, and haddock using historic survey data collected along the Norwegian coast. With our calibrated model of fish swimbladder resonance and historic length distribution data from Norway, swimbladder resonance frequencies and target strengths of these fish species were estimated. We also determined optimal OAWRS ship tracks for remote detection of these species above seafloor scattering using waveguide propagation modeling. While the OAWRS frequencies were greater than those expected for cod swimbladder resonance, cod shoals over ten kilometers in length were robustly detected and successfully imaged from tens of kilometer ranges during the experiment. This produced the first instantaneous images of a vast cod shoal. It also confirmed our predictions that OAWRS can be used to remotely sense and survey cod populations. Our theoretical predictions suggest that the use of lower OAWRS frequencies near cod swimbladder resonance would lead to greater dynamic range in population density estimates. The Nordic Seas experiment provided the first look revealing the entire horizontal morphology of vast cod, capelin, haddock and Norwegian herring shoals. This was done with instantaneous OAWRS imaging. The presence of multiple shoaling fish species during the Nordic Seas experiment provided us with a unique opportunity to study general shoaling behavior across species over ecosystem scales with OAWRS. For example, many pelagic and demersal fish species are known to undergo distant migrations for feeding, spawning and overwintering year after year. This suggests that migrating populations have an ability to efficiently sense their environment. By combining OAWRS estimates of fish scattering strength and population density obtained from simultaneous depth echo-sounding along line transects, areal population densities over entire shoals were determined. This enabled estimation of total shoal population, shoal aspect ratio, and shoal migration speed via cross correlation of population density over time. It was shown that across several species, as shoal population increased (tens of thousands to hundreds of millions of individuals), shoal aspect ratio also increased (roughly from one to ten). Single-celled organisms with higher aspect ratios have been shown to more efficiently and accurately detect chemical gradients at microscopic scales. The high-aspect ratio or elongated morphology of a large migrating fish shoal is consistent with the entire shoal serving the function of a biological antenna for efficient spatial and temporal sensing of mesoscale processes in the environment. We also studied the evolution of air resonance power efficiency in the violin and its ancestors. We collected historical data, including samples from roughly 500 classical Cremonese violins from the renowned workshops of Amati, Stradivari and Guarneri, to establish historic time series of key design traits. We determined the primary physical mechanisms governing radiated air resonance power in the violin and its ancestors and used this knowledge to explain the evolutionary trends we discovered.
Description
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2015. Cataloged from PDF version of thesis. Includes bibliographical references (pages 259-278).
Date issued
2015Department
Massachusetts Institute of Technology. Department of Mechanical EngineeringPublisher
Massachusetts Institute of Technology
Keywords
Mechanical Engineering.