Preliminary results of field mapping of methane plumes offshore of Coal Oil Point, California with a RESON 7125 multibeam sonar in water-column mode
Abstract
From June 17 - 23 2010, the U. S. Geological Survey (USGS) in collaboration with the Bureau of Ocean Energy Management Regulation and Enforcement(BOEMRE), the Royal Netherlands Institute for Sea Research (NIOZ) , RESON Inc. and the University of California, Santa Barbara(UCSB) conducted a comprehensive marine-seep gas-plume mapping study offshore of Coal Oil Point, California. The ultimate goal of the experiment is to quantify the amount of methane emitted from natural seeps using multibeam sonar, with results calibrated using field measurements of aqueous and atmospheric methane in the seep fields. Success will lead to better estimates of natural marine methane contributions to the global methane budget. We mapped selected seeps, some twice, with a pole-mounted RESON 7125 multibeam with a 10-degree forward rake. Other equipment included a Benthos Stingray ROV equipped with high-definition video cameras and in situ gas sampling apparatus, Niskin bottles for water column sampling of dissolved methane, and a Picarro G1301 cavity ringdown spectrometer for mapping atmospheric methane concentrations. This paper focuses primarily on the data reduction and data visualization strategies employed while processing the more than 1.2 TB of raw water column data collected by the multibeam system over several high-output oil and gas seep areas. Water depths ranged from about 30 to 80m. Turnkey software solutions for processing these data are currently unavailable so most of the processing code was developed in-house by the USGS. The main challenge in processing the sonar water-column data is ray-tracing the large volume of data, with each ping containing more than 4500 times as many samples as a conventional multibeam ping. We employed two strategies to make processing tractable on conventional workstations: (1) decimate the raw data based on desired output resolution before ray-tracing; and (2) design the ray-tracing program to run in parallel on multi-core workstations. Utilizing an 8-core, 3.00GHz Intel Xeon X5365, we achieved a processing throughput rate of 14 MB/s while utilizing only 12% of the available I/O bandwidth (the code is CPU-bound). Processing 1.2 TB of raw water column data to the point of statistical analysis and visualization required about 25 hours of computation, and resulted in an output data set of about 200 GB. Because the data processing is “embarrassingly parallel,” that is, each raw sonar file can be processed independently, a number of divide-and-conquer optimization schemes could make processing water column data even more efficient. Use of trade names are for descriptive purposes only and do not imply an endorsement by the US Government.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2010
- Bibcode:
- 2010AGUFMOS12B..04F
- Keywords:
-
- 3004 MARINE GEOLOGY AND GEOPHYSICS / Gas and hydrate systems;
- 3094 MARINE GEOLOGY AND GEOPHYSICS / Instruments and techniques