Logo: University of Southern California

ISI Researchers Organize Geoscience Cooperation as Part of NSF Cyberinfrastructure Initiative

The EarthCube project is developing research workflow technology in cooperative, coordinated effort

July 11, 2012 —

The Viterbi School’s Information Sciences Institute (ISI) is playing a leading role in a new National Science Foundation effort to make earth science and geological research more effective, collaborative and efficient. 

The NSF described the hoped-for structure and functionality in a June 2011 "Dear Colleague" letter. “The scientific objective,” said a subsequent NSF mission statement, “is to develop a framework over the next decade to understand and predict the Earth system from the sun to the center of the Earth.

Yolanda Gil
Yolanda Gil
The initiative, known as EarthCube Workflow Community Group officially began April 1, co-directed by ISI Director of Knowledge Technologies Yolanda Gil and ISI Project Leader Ewa Deelman, both of whom also hold faculty appointments in the USC Department of Computer Science.

Although the EarthCube effort is very broad and encompasses issues of data management and access, governance, interoperability and more, researchers at ISI are focusing on the use, promotion, and enhancement of scientific workflows in earth sciences.

"EarthCube is a groundbreaking NSF program," said Tim Killeen, assistant director for NSF's geosciences directorate. "It represents a dynamic new way to access, share and use data of all types to accelerate and transform research for understanding our planet. We are asking experts from all sectors — industry, academia, government and non-U.S. institutions — to form collaborations and tell us what research topics they think are most important. Their enthusiastic and energetic response has resulted in a synergy of exhilarating and novel ideas."

The group, composed of computer scientists, earth scientists and cyberinfrastructure providers met via conference calls, virtual and in-person meetings to discuss the role of scientific workflows technologies in earth sciences now and in the next decade.

A workflow, as described by the group, “is a series of connected steps employed to accomplish some overall goal. Typically each step produces output that the next step needs as input and the steps are repeated every time new input is available at the first step. Almost all geoscientists employ some kind of workflow in their work while processing data from a sensor, cataloging data from the field, visualizing data or analyzing output from a numerical model.”

As research has become more intensive, the group notes, informal, seat-of-the-pants workflows have become increasingly dysfunctional. A formal discipline for Workflows has emerged, with research by Deelman and Gil playing a key role in creating and building the paradigm as well as providing the tools needed to support workflow-based computing (http://pegasus.isi.edu, http://wings.isi.edu/).

Ewa Deelman

At USC, Deelman has been working with the Southern California Earthquake Center (SCEC) on developing tools and techniques to support large-scale earthquake simulations. An example of a SCEC application is CyberShake, which generates seismic hazard maps of the Southern California area. These maps are computed using the Pegasus Workflow Management System and the resources provided by the NSF XSEDE cyberinfrastructure.

The goal of the EarthCube Workflow group is to discuss workflow concepts within the group and more broadly with the GEO community, and using their input and feedback, define a roadmap for scientific workflows. The roadmap was presented and discussed at the June 12-14 charette, an active community meeting hosted physically in Rosslyn, Va. and virtually via Webex and other technologies.

The June charette further refined the draft workflow that has come out of the previous months from six geosciences community groups, each focusing on a different aspect of workflow architecture, including both an overview and Data, Governance, Workflow and Semantics.

The paradigm is emerging, but not immediately. As the NSF statement notes," We anticipate the structure and functionality of EarthCube will emerge over the next few years."