EarthCube is a joint effort of the NSF Geosciences Directorate and Office of Cyberinfrastructure. Its goal is to use technology and knowledge management techniques to increase the productivity and capability of researchers and educators working at the frontiers of Earth system science. (If you're new to EarthCube, you may want to read these previous posts.)
Three members of the Unidata Program Center staff attended the EarthCube charrette held in Washington D.C. on November 1-4, 2011. The following is Unidata netCDF developer Russ Rew's summary of the meeting. Impressions from Unidata Program Center director Mohan Ramamurthy and Development and Outreach Manager Ben Domenico are also available.
By Russ Rew
The EarthCube Charrette was an experiment in people participating in processes to self-organize, reach consensus, and provide feedback to NSF on the EarthCube idea. During three and a half days of expert facilitation in breakouts, at breaks, and over beers, we talked about capabilities we need but don't have, interoperability that seemed either just out of reach or nearly impossible, barriers to achieving needed capabilities, challenges and milestones and next steps, white papers and sticky notes, and collaborations that EarthCube just might make possible.
I traveled to the Charrette somewhat skeptical of getting much out of such a large meeting. Had everyone read all 111 white papers? With the diversity of aims represented in the white papers, would it be possible to achieve agreement on anything specific?
I was surprised at the results, and came back more "enthused" than skeptical about the potential of EarthCube. Working in changing groups of six to ten people, we actually spent most of the meeting fully engaged in completing tasks that resulted in organized lists of capabilities, barriers, and priorities.
In addition to the breakouts, there were several interesting plenary talks on similarly ambitious projects, including
- Steve Wolff (NSF) giving an insider's view on how NSF kick-started the public Internet
- Robert Gurney's (UK/NERC) talk about lessons learned from a pilot project to plan and establish the Environmental Virtual Observatory
At the end of each day, we watched a dozen five-minute presentations on white papers, with the added the excitement of a huge digital clock projected on the back wall counting down the seconds until the Maori dance music volume would rise to drown out the speaker/participant whose time had run out, or who had managed to cram an amazing amount of information and persuasion into exactly 300 seconds.
A goal that wasn't accomplished at the EarthCube Charrette was to narrow the options and requirements to achieve consensus on a vision or conceptual architecture of what EarthCube should be. Goals that were achieved included the generation of some excellent white papers, rough agreement on the most important capabilities needed, engagement of participants in struggling with obstacles to implementing those capabilities, and enthusiasm and energy of participants in tackling problems across disciplinary boundaries.
I left with ideas for some new collaborations and an admiration for the skills of a few expert facilitators who had wrung some tangible recommendations out of academics, scientists, developers, and managers working in parallel in a kind of organized chaos. Whether the results are actually useful remains to be seen, but by the start of the last day, it seemed that the EarthCube Charrette exceeded the expectations of the NSF leadership who had bought into the risky experiment of a new kind of meeting to organize a new kind of collaborative effort.
Some quotes and paraphrasings from the meeting notable enough to write down:
- "To change the world, you need vision, passion, commitment, timing, some money, and friends in high places." -- Steve Wolff
- "We reject kings, presidents, and voting. We believe in rough consensus and running code." -- Dave Clark (IETF)
- The pitfall of "profiles": two organizations could implement two profiles of the same standard resulting in no ability to interoperate
- Build in proper governance from the beginning -- Robert Gurney
- "Reverse the 80/20" -- bumpersticker concept about purported 80% of the time it takes to researchers to locate and transform the data to use in the 20% of the time spent on science
In a follow-up discussion among members of the Boulder Earth and Space Science Informatics Group (BESSIG), there was considerable discussion of whether skepticism is still called for, because other similar efforts to achieve broad infrastructure interoperability across disciplines have consistently failed in the past. What's different now? Answers included:
- new understanding of the power of service-oriented abstractions
- new appreciation for the need for cyberinfrastructure across Earth sciences
- recognition within NSF GEO that sustained cyberinfrastructure could almost fund itself by savings from current efforts in individual research grants to deal with lack of interoperability