The purpose of this workshop is to address the complexities of cyberinfrastructure and effective use of technology to increase the productivity and capabilities of the geosciences community. Building on a substantial bibliography of reports from workshops, advisory committees, and government agencies, the focus of this workshop will be in three principle areas:
1) Develop and nurture a community of practice. Identifying and encouraging the use standards in cyberinfrastructure, such as use of SOAs for access and use of data, that will support trans-disciplinary work well into the future (e.g. GIS standards that have been adopted by geosciences researchers). It is recognized that there will be a variety of approaches used by different groups within geosciences. However, these standards, and interoperability among them, will improve over time, if technology history provides any guidance. It is important the geosciences community influence the development of standards that are widely used in more diverse communities. Developing effective methods to influence standards and practices will benefit the geosciences. 2) Address the full lifecycle of data. NSF data policy encourages researchers and educators to share data. The recent implementation of a required data management plan for each project proposal submitted to NSF encourages the community to consider the full lifecycle of data that is generated during a grant. There are several approaches to lifecycle paradigms and adoption of community practices, the choice of technology employed, and the development of partnerships all influence the implementation of the full lifecycle of data. 3) Symbiotic partnerships. Partnerships with Federal agencies, international organizations, non-governmental organizations, and industry are essential for the long-term success of the geosciences. Much of the geosciences community supported by NSF depends on data, software, and practices gathered, developed and encouraged by other entities. For example, data provided by USGS, NOAA, and NASA are invaluable in the conduct of many NSF supported research activities. However, the full value of the research community to the agencies and organizations nationally and internationally has yet to be realized because symbiotic interactions have not been developed to their full potential.
The principal deliverable from the workshop will be a state-of-the-field report to NSF and the USGS (a co-convener of the workshop) of the geoinformatics community?s capabilities and needs that could in turn ultimately benefit from an academic-multi-agency community-focused set of development activities. Additional deliverables would include: a) readiness assessment and identification of gaps for both technology and education around geo-data informatics and their priorities;;b) grand challenge opportunities as well as immediate next steps; and c) Identification of additional stakeholders and means to include their inputs.
If this workshop is successful in setting directions for the development of: community cyberinfrastructure practices, a culture of consideration for the lifecycle of data, and paradigms for mutually beneficial partnerships, then the foundation for cross-directorate and trans-disciplinary research will be greatly enhanced.
Finding: Communities A pattern of success seems to be fostering communities that work within a discipline, but across institutional boundaries, to define vocabularies and data models, and build useful tools (seismic community, met/ocean community, biological community). While one would like to believe that mutual benefit would keep these communities working together, the reality is that often progress stops when the funding dries up. Recommendation: Sustain funding for discipline-oriented integration. Agencies should consider data integration by discipline with other agencies and academic communities instead of agency-wide data integration efforts. Encourage more participation in OGC Domain Working Groups. Finding: Cyberinfrastructure While best progress is often made within communities, there are often cyberinfrastructure elements such as scientific feature types (images, grids, time series, point samples), concepts, technologies, and approaches that are common, and therefore building cross-walks between communities and enabling cross-fertilization is essential. Recommendation: Support translation/conversion/ontological tools that link communities. Expand the model of sustained funding for groups that build fundamental infrastructure for the community and are driven by community needs (Unidata model). Continue support for workshops like this and encourage continued partnership between agencies and academia. Expand research on how and where data management and curation approaches can apply for particular data types produced and used across disciplines (Where we can implement general solutions, and where we shouldn't). Finding: Marketing There is an open question of how to effectively market Geoinformatics? The usual focus is to state how geoinformatics will lead to 4th Paradigm breakthroughs (however, see Tim Killeen's challenge to the workshop earlier in this report): The open question was: could we also market geoinformatics on the potential to save $$$ or do more science by making "ordinary science" more efficient? Recommendation: Fund social science studies in different geo-disciplines to more accurately document how much time scientists spend doing jobs that better geoinformatics could eliminate (browsing for data, reformating data, writing custom codes to do routine operations). Finding: Where is funding allocated? It is often very hard to get scientists to provide metadata. Until scientists receive adequate credit for good and useful data (e.g. data citations count towards tenure, etc.) they will struggle to see the value and are too busy paying attention to the matters for which they are rewarded. A culture change will be required. Recommendation: Traveling shows (seminars) to present what metadata can enable so that scientists can see the value. Community colleges could provide training on basic geoinformatics approaches, working together with masters level domain folks, interviewing/assisting scientists to enable quality metadata in their discipline. E.g., partnerships between community colleges and scientific institutions and internships at scientific institutions could be valuable. Try creative approaches: H&R Block helped Farmers supply geodata. Outsourcing, crowdsourcing, support work on dynamic, iterative approaches to metadata. Finding: People! As geoinformatics grows, there is a very clear need for more proficient geoinformaticists to solve problems. Recommendation: Help foster academic programs in geoinformatics, sponsor summer internships (like Google Summer of Code, but for geoinformatics). Overall summary of key points: A shift is needed within agencies to provide longer-term funding support, for communities to come together, remain coherent and to enable integration within their communities and across to other communities (to the extent possible). Standards and practices should build from demonstrated successes, and tools need to be developed to support them. Education is critical to broader adoption, marketing studies need to be conducted to provide the business case for integration, and incentives are needed to encourage everyone to participate in making data integratable. Overarching recommendations: Create an interagency working group to foster data life cycle management practices, support coordination and informatics science initiatives, and to develop a high level shared vision and strategy for data life cycle management Create and fund a coordination office that works with the "working" level of agencies and academic institutions to facilitate working groups and workshops, adoption of standards and tools, and creation of sustainable archives Support and actively participate in existing coordination groups, and create where needed, new communities of practice in data life cycle management across agencies, academic institutions, and the private sector Establish an NSF working group to address the issue of incentives and cultural change needed to facilitate implementation of data life-cycle management