2009 GES DISC User Working Group (UWG) Recommendations and Responses May 10-11, 2011

1. 2009 GES DISC User Working Group (UWG) Recommendations and ResponsesMay 10-11, 2011

2. 1. Expand the UWG The current GES DISC User Working Group should be expanded to include a broader community of data users and providers.  The “hands on” constituency, like graduate students and postdocs, is especially important.  This is a priority before the next UWG meeting in about six months.

3. 1. Expand the UWGResponse Additional UWG members have joined the GES DISC UWG: Hook Hua – JPL/ Information Technology Dalia Kirschbaum – NASA/Precipitation Steve Massie - UCAR/Atmospheric Christian Falkenberg – JPL (unable to attend this meeting) Welcome! - No longer able to serve as member: Andrew Gittleman - Attended previous meeting, but could not make this one: Ken Pickering 

4. 2. Organize data by science discipline NASA’s data sets would benefit from organization by science discipline, as opposed to source instrument or physical condition like atmosphere or ocean.  The NEWS project could be a test bed for such an approach.  The Global Change Master Directory already includes all NASA data.  Could something analogous be developed for NASA data sets only? We recognize this is a broader issue than the GES DISC, but this approach will better the community of science users. That community is expected to expand in the coming years as the community of Earth science data users continues to grow with interest in climate changes.

5. 2. Organize data by science discipline Response • The parameter tree in Mirador has been organized along the lines of NASA's six Earth Science Focus Areas.   The organization details were occasion for some debate… Looking for feedback from the UWG members as to its effectiveness in making data easy to find. http://mirador.gsfc.nasa.gov/cgi-bin/mirador/presentNavigation.pl?tree=scienceArea&CURRENT_CONTEXT=scienceArea • With respect to showing non-GES-DISC data, we initiated a Federated Search interoperability framework amongst ESIP Federation members that would enable Mirador to search other data providers at both dataset and file level (http://wiki.esipfed.org/index.php/Discovery_Cluster ). • Collaboration with an ACCESS-NEWS project to explore a reciprocal OpenSearch based catalog interoperability was initiated that would enable NEWS data and services to be presented as search results within Mirador

6. 3. Keep an archive of user scenarios An example of a user scenario can be downloaded in the documents section at the website mentioned in the Introduction. An archive of such scenarios will help many new data users understand the data sets available at the GES DISC. A well-written scenario effectively summarizes the science questions underlying an analysis, and can serve as an entre into a project. They are usually easier to understand than Algorithm Theoretical Basis Documents and user guides.

7. 3. Keep an archive of user scenarios Response • User Scenario Archive suggestion provided a key part of the inspiration for the Earth Science Collaboratory • In the meantime, we will look at some possible approaches Further discussion this afternoon

8. 4. Increase access to multi-sensor observations The scientific user community is often interested in multi-sensor observations, and NASA Headquarters has committed to furthering this work through the wording of ROSES calls.  The greatest benefit will be achieved with Level 2 data sets. However, the general problem of multiple Level 2 data sets is difficult due to the complexity of the constituent data. Specifically requiring a connection between scientists and multi-sensor data providers in proposal calls will help reduce some of the challenges involved with combining data from multiple instruments.

9. 4. Increase access to multi-sensor observations Response The GES DISC provides the following background information in support of this recommendation: GES DISC projects have been successful in co-registering, providing access, and preliminary provenance to data from multiple sensors of common themes. The GIOVANNI data exploration and visualization tool has been central to the GES DISC’s success in facilitating multi-sensor data analysis. With over 40 single sensor and multi-sensor instances, Giovanni has become extremely popular in quickly displaying and providing data download for further data analysis. GES DISC participated in multi-sensor efforts incorporating GIOVANNI into the Year of Tropical Convection (YOTC), and the international collaboration, Atmospheric Composition Portal. Dr. Leptoukh has convened and chaired several AGU sessions on the use of multi-sensor data, which grew from one session 2 years ago, to multiple sessions last year.

10. 5. Help further NASA’s role in climate assessments We can expect increased interest in using NASA satellite data for weather and climate model assessment.  The users of these data will likely be less familiar with methodology, formats, etc., than users over the past decade.  Conversely, model data size and interpretation is beyond the GES DISC capability.  In many cases the climate model assessment community will be interested in high quality (low bias) observations in Level 3 (gridded) formats. We recommend that the GES DISC strengthen ties to model assessment efforts, such as the Coupled Model Intercomparison Project (CMIP) version 5 being conducted for the next Intergovernmental Panel on Climate Change (IPCC) assessment. In addition, the NASA Center for Computational Sciences (NCCS) is working with Goddard model developers, and the roles and tools should be coordinated between GES DISC and NCCS.

11. 5. Help further NASA’s role in climate assessments Response The GES DISC ties to modeling and data assimilation groups have strengthened with the ingest, archive, distribution of, and services for, the GMAO’s MERRA 30 year reanalysis data, and more recently NASA/GSFC’s GLDAS and NLDAS data. The GES DISC has participated in workshops and associated mail list discussions on how to get data to the CMIP5 folks in ESG. The GES DISC has had discussions with NCCS regarding how to make model data production (at NCCS) and distribution and services (at the GES DISC) closer, more efficient, and cost effective Discussions in progress for potential relevant proposals

12. 6. Enable transition to users in the applications and decision support communities Except for weather forecasting, the applications and decision support communities are currently small. However, the decision support community can be expected to grow significantly in the coming years. Decision support user needs can be expected to be different from those of earlier users (such as weather forecasters who use near real time downlinks). We recommend that the GES DISC become familiar with the applications and decision support communities, both inside and outside NASA, to anticipate their needs in the coming years. These needs may be similar to those of the climate assessment community. Simplification of data set search terms and utilities will help maintain a broad user base. Note that the needs of that broad community will also be addressed by the recommendation “Organize data by science discipline” above.

13. 6. Enable transition to users in the applications and decision support communitiesResponse • Past work included partnerships with USDA FAS, USDA ARS. • Performed data inter-comparisons in collaboration with EPA • Current project (NASA ROSES DSS) provides data for decision support to USDA WAOB • Currently supporting public health applications with information on relevant data and information services • Due to finite resources available to explore the broad range of possibilities that could lead to potential new decision support efforts, the GES DISC has limited focus to funded (awarded) projects • Would including an applications tab be a recommended high priority?

14. 7. Help ensure data quality and provenance The previous three items show how our data sets are becoming increasingly diverse and complex, and applied to a broad range of problems.  How do we ensure their quality and record their development history? The current ACCESS proposal call attempts to answer this question. However, formalisms for ensuring data quality, and especially provenance, are not presently mature. We recommend that NASA Headquarters continue to support GES DISC activities in this area. As with multi-sensor observations mentioned above, formal ties between data producers and data providers (e.g. GES DISC) can be required through NASA proposal calls.

15. 7. Help ensure data quality and provenance Response • The GES DISC submitted several successful proposals to ACCESS and AIST ESTO solicitations: • Multi-sensor Data Synergy Advisor explores comparative provenance for multi-sensor data fusion in Giovanni (ESTO) • The Data Quality Screening Service helps users in applying quality screening to products related to the water cycle • Aerostat provides provenance for aerosol datasets in Giovanni, with an emphasis on the impact of cross-instrument bias. • The GES DISC continued to expand Giovanni provenance, e.g., a limited lineage added to A-Train Giovanni. • Initiated the ESIP Federation Information Quality (IQ) Cluster to address data quality issues and methods • Collaborated with European and international efforts • Worked with Measures Principle investigators on provenance • … more discussion later in the proposed Data Citation Policy section • Are there any other recommendations to pursue?

16. 8. Transition to support for future missions NPP / NPOESS and the first round of Decadal Survey missions will create new data streams in the coming years to decades. The GES DISC can be expected to support scientists using these data, and to integrate them into the existing data system. This process will be simplified for new users if data are organized by scientific discipline (as discussed above).

17. 8. Transition to support for future missionsResponse • Known future Missions which will be served by the GES DISC: • GPM • OCO • Selected MEaSUREs Projects • Mission that we were recommended to service, but no decision has been made • SMAP • Missions lost • Glory (did not make orbit) • DESDynI (cut from NASA budget) – DISC role would be that of a Science Investigator Processing System (SIPS), not a NASA public archive and distribution system

18. 9. Allow subsetting and simplified formats Many users are interested in only a few parameters, preferably in simple formats like NetCDF or ASCII. In contrast, instrument teams provide files with many parameters, usually in HDF.  The problem of abstracting out the most important fields is most pronounced with Level 2 data, whose scientific benefit is still being demonstrated.  As with multi-sensor observations, communication between instrument teams, data users and the DAAC is critical. User scenarios can also help in illustrating why only a small subset of data may be relevant to most users. Similar to the issue of provenance, DISC tools that act on the data need rigorous testing and validation to ensure the subset requests are being fulfilled exactly. This validation activity should be considered in development timelines and resource requests.

19. 9. Allow subsetting and simplified formats Response • Subsetting and reformatting options have been expanded, largely through deploying new OPeNDAP servers that support NetCDF download. • Simple Subset Wizard adds subsetting and reformatting services and makes existing ones easier to use

20. 10. Continue asking GES DISC UWG and science community for clarification on data service enhancements Are existing DISC capabilities sufficient to meet the challenges described above, or should others be developed? Because of data set size and complexity, the general problem of, e.g., matching all A-Train parameters to all others is not feasible. Science guidance can help reduce the many possible enhancements to a much smaller, more useful subset. The UWG also considered it important for the UWG to reach out to national and international activities to be serviced (current DISC NEESPI [Northern Eurasia Earth Science Partnership Initiative] activities provide a good example).

21. 10. Continue asking GES DISC UWG and science community for clarification on data service enhancements Response • Advice sort from the UWG on: • Giovanni recommendation • Would like to engage UWG in a quarterly telecon to more regularly exchange information. Thoughts?