Tuesday, January 22, 2013

Surface Process Experiments -- Town Hall Report

It has been about one year since we started this group, and we want to thank everyone who has contributed their time, suggestions, or even just interest and attention. In the next few emails we will report on some of the recent events. The first report is on the Town Hall that was held at the AGU Fall Meeting. If this summary sparks any ideas or comments you’d like to share, contact us!

Town Hall summary

On December 7, 2012, we convened a town hall meeting on “Surface Process Experiments – A Community Discussion” to solicit from the AGU Earth and Planetary Surface Process (EPSP) focus group community opinions, challenges, and future directions on managing Earth-surface experimental data sets. The meeting included six panelists with perspectives from collaborative community initiatives, funding agencies, and journal publishers. Approximately 70 people attended the meeting, which was funded by NCED (A full report and transcript of the meeting is available at: http://tinyurl.com/sedex-townhall2012).

Panelists and audience members expressed universal concern for the current lack of infrastructure and support for data management and sharing for our particular data types. Participants expressed challenges falling into three primary categories:

(1) Data structure and description
: Metadata describing data provenance, structure, and uncertainty are as important if not more important than raw data itself. Templates and guidelines would help, though the diversity of experimental types and institutional directives likely preclude a universal standard.

(2) Long-term archiving: Preserving data in usable and accessible formats over time spans far exceeding typical funding and project cycles (>10 years) is a challenging and expensive prospect but extremely important for our community.

(3) Publication and recognition
: Scientists will only invest their time in data management if they feel their efforts are recognized, and publication of citable “data papers” appears a natural answer to this question. AGU journals support publication of data papers, but fundamental questions about data paper structure, scope, and attribution must be overcome to attract more than a token number of such submissions.

Participants identified limited human, computing, and financial resources as major barriers to proper data management, especially in smaller universities and facilities. Furthermore, data and experiments come in many different forms (and degrees of processing), and it is hard to imagine a generalized way to describe these. Rather than starting from scratch, many participants suggested that we learn from and build on existing resources, such as the NCED data repository and the Consortium of Universities for the Advancement of Hydrologic Science, Inc. (CUAHSI) Hydrologic Information System (HIS). The best approach to managing data will likely combine top-down directives and support (for example, NSF could potentially provide support for long term data archiving and centralized databases) with bottom-up engagement by Earth-surface scientists to define database structures, templates, and guidelines.

Best regards,
Raleigh Martin
Wonsuck Kim
Brandon McElroy
Leslie Hsu