Swimming in Data Offshore Alaska

by | 7.31.2011 at 9:22pm
Follow us on Facebook or Twitter

Distraction from data analysis: Rainbow over the snow capped mountains of the Alaska Peninsula

Although we still have ~3 days of data collection aboard the R/V Langseth to go before we pull in our equipment and head for port, we are already drowning in beautiful seismic data. Following each pulse from the air gun array, the two 8-km-long streamers listen for returning sound waves for 22 seconds. This is enough time for the sound waves to travel down through the water, sediments, crust and upper mantle and back again. Arriving sound waves are recorded on a total of 1272 separate pressure sensors along the streamers, producing about 60 Mb of data for each pulse.  Repeat this every 25 seconds for 3 weeks, and you end up with a pile of data!  We have already recorded over 2.5 terabytes (2500 gigabytes!) of raw seismic data.  This does not include other large datasets that we are simultaneously acquiring, such as detailed bathymetric mapping of the seafloor.

Seafloor faults imaged by bathymetric mapping

Once we obtain the raw data, our eager scientific party cannot resist beginning some rudimentary analysis, thereby generating even more large data files that take up yet more disk space.  In search of instant gratification, we use some quick and dirty processing steps to produce preliminary images from our data and get a first peek at the structures beneath the seafloor. This is standard procedure on cruises aboard the Langseth and other seismic ships. Often, such images reveal very little; careful analysis of seismic data to create clear and accurate images of earth structures takes years. But in our case, the data are of such high quality that spectacular features are evident even in these rough first images, including the plate boundary and other faults. This assures us that hard work on the data following the cruise will produce very exciting results.

One of the key shipboard tasks is determining the position of the gear in the water and combining this navigational information with the raw data.  Our streamers are 12 m under the sea surface, so we cannot simply attach tons of GPS sensors to them to figure out where they are at any given time. Instead, the Langseth’s infatigable Chief Navigator, David Martinson, works out the locations of the streamers using GPS’s at the beginnings and ends, a series of compasses spaced along the streamers, and several “acoustic nets,” sets of instruments that give the distances between the streamers at key positions. He can determine the positions of our two unruly 8-km-long cables to within ~5 m or less at any given time – amazing!

We also produce initial images of seafloor topography from bathymetry data. At sea we begin the arduous task of manually editing vast quantities of the data, but the effort pays off. Careful analysis of these high-resolution data can reveal faults that cut through the seafloor, seamounts, and sedimentary features.

Follow us on Facebook or Twitter

Comment Using Social Media

Comment