August 9, 2018

Hacking the ocean’s mysteries

Bottom view of a Big Shark and Sting Ray or Myliobatis aquila, swimming under blue ocean. Underwater blue background. Undersea marine life.

The Regional Cabled Array of the Ocean Observatories Initiative is a vast underwater network of systems and sensors capturing continuous data.

A vast underwater network of systems and sensors are capturing rich, never-before-accessed data from the mysterious world beneath our oceans.

To build a stronger community of scientists using that data to make new discoveries, oceanographers will convene at the UW for Oceanhackweek, August 20-24, 2018, five intensive days of collaborative investigations and tutorials in modern data analysis tools and techniques.

The hackweek aims to close the gap between the massive amounts of data being continuously streamed to shore and the ability of oceanographers to analyze that data, said Rob Fatland, director of research computing in UW Information Technology and an Oceanhackweek co-organizer. “We’ve become increasingly talented at gathering data,” he said, creating what he calls “a deluge of data.”

Without data science methodologies and computational tools, scientists are at a disadvantage when it comes to making sense of so much data.

Wu-Jung Lee, a research associate at the Applied Physics Lab (APL), faced this same challenge two years ago.

“I could see what I wanted to know, but my skill set was limited,” Lee said.

That changed in 2017, when Lee joined a 10-week Incubator program in the eScience Institute that equipped her with data science techniques and computational tools she needed to quickly sift through acoustic data and make scientific discoveries.

She along with colleague Aaron Marburg, a research engineer also at APL, teamed up with Valentina Staneva, a data scientist at the UW eScience Institute, to work with the real-time, continuous data from the Regional Cabled Array of the Ocean Observatories Initiative (OOI), a vast underwater network of systems and sensors.

Lee had seen the successes of other hackweeks — Astrohackweek, Neurohackweek, Geohackweek — all held at the eScience Institute, and an idea was forming. Why not do a hackweek for oceanography?

The overarching goal of hackweeks is to provide opportunities for researchers to develop data science capabilities, enhance collaboration and improve reproducible science in order to accelerate research.

Lee then found an ally in Friedrich Knuth, a Cabled Array data evaluator at OOI who facilitated data access. He wanted to help scientists better navigate and use these unprecedented volumes of data.

Together, the team quickly garnered interest and support among a whole host of interdisciplinary researchers and data scientists to develop Oceanhackweek (and an earlier Cable Array Hackweek). In addition to Fatland, Lee, Staneva, Marburg and Knuth, the organizers include Amanda Tan (UW-IT/eScience Institute) and Deborah Kelley and Don Setiawan (UW School of Oceanography). The August event is supported by more than $100,000 in grant funding from the Consortium for Ocean Leadership, the nonprofit organization that operates the Ocean Observatories Initiative. Funding for the initiative comes from the National Science Foundation.

“If a participant in Oceanhackweek can have that same inspiration I had in the incubator, then we are successful,” she said.

People collaborating at the Cabled Array Hackweek

Participants at the Cabled Array Hackweek in Feb. 2018. Photo courtesy of Valentina Staneva.

 

The organizers received 143 applications from people at universities around the world, in industry, and at governmental organizations. The number was three times as many applicants as they had room for in the hackweek.

The popularity, explained Lee, isn’t surprising.

“It’s not just shaking hands at a conference. It’s a much deeper connection. Half of the hackweek is devoted to project time. We bring all these people with very different interdisciplinary interests and backgrounds together; you know there will be interesting things that result.”