EarthCube Town Hall

Lee Allison introduced EarthCube and went over the agenda (see slides for details)

Eva Zanzerkia virtually presented with the goals of EarthCube from the NSF view point and motivation with moving forward.

Vision at NSF - starting in June 2011, they came up with the term EarthCube.  It is about cyberinfrastructure for the geoscience community and the fundemental earth systems.  Simple and easy access to information and seamlessly connect to the information.  Developed flexibly over time and community driven.

It is part of a larger structure Ecosystems - see star chart on slides, there are 4 components at the bottom and the top expands it to include organizations, scientific instruments and expertise.

In partnership with ACI and Directorate of Geosciences and many others (see slides) from related fields represents the programs within the geoscience directorate with fundamental sciences within the US communities which are supported by this.

Looking forward - they have seen that the data, the teams and the models are now interdisciplinary in order to be on the cutting edge.  Eva gave some examples on the slides that span domains.  And the coupling between many complex systems.  Currently, doing the work is difficult because the infrastructure is not there for the data.

In partnership with NSF and the community - scientists, cyber infrastructure and software engineers, and agencies, people and support need to be a critical part of this dialog.  This is a stepwise process, with a community assessment to determine gaps and goals, which will give the NSF a set of requirements to craft the next years opportunities.  For cyberinfrastructure for the entire community.

Over the past year, and towards solicitations for 2014.  First they held 27 end user workshops.  Each of these slices is a domain brough together with scientists in these domains or related agencies.  The scientists talked about their needs and the cyberinfrastructure challenges that prevented them from doing work.  The end reports are all available on the earthcube webportal.  It developed a set of very broad user needs.

THe second component - the framework for how that dialog would begin for building cyberinfrastructure.  There were different award groupings, each with a different role.  It is meant to connect resources to user communities.  RCN’s - bring groups together ..., Building blocks - initial technology that can be demonstrated, and finally test enterprise governance to connect all these pieces.

Test enterprise - 2 year project with outreach.  It has a planning and demonstration section as well.  Facilitates the communities needs and brings them back to NSF for implementation.

2014 solicitations - funding calls released in amendments - call due March 12th 2014.  Allowing collaborative proposals.  Need to show how each of these pieces will be coordinated and who will lead these parts.  Conflicts of interest are still required, and a completed spreadsheet which is found online.

The solicitation has two pieces

Research coordination network - emphasis broad participation in any domain.  WIth a clear outcome after 24 months - not to create cyberinfrastructure but to articulate, standards, priorities or developments they would require.

Building blocks - opportunities they might have missed in the past, and activities for the future. Should connect to the end user workshops.

Building Blocks objectives - looking for broad collaboration and integration and finally to demonstrate a technology that could be more broadly applicable.  Want a demo within 24 months, and why it is motivated to a broader community.  and how it fits within existing ecosystem.

What is not appropriate  - database or something so specific it is domain specific and not accessible or extensible.  Projects that do not deeply involving academic geosciences and their resources, or end to end solutions.

Went over timelines and where that leads towards the future, which will depend on the given awards, the test governance and the annual events.  One potential outcome would be a demonstration of an organization which can really service the geoscience community and give NSF back what it needs.  And then it might be a long term funding enterprise, and there might be large prototypes funded, and perhaps more RCN calls as well to create effective partners.

THey expect EarthCube and NSF to facilitate bring geoscientists the opportunities they need, within NSF at large.

Agency collaboration is an important part of EarthCube.  Federal memos on open access, NSF sees EarthCube as a way to organize the academic geoscience community, to make it easier for Federal agencies to interact with the academic community.  How can they strengthen that dialog and serve an array of communities.  Also at the agency level and with organizations like ESIP.  Specifically with the test enterprise project.

Ended with a set of guiding principles and guidance for the awards.  Looking for shared solutions but also being flexible as technology changes instead of now, looking to 10 years from now.

Questions

Christine White - in the next session, thinking about solutions to the end user workshops, representative from ESRI - how can they participate in building blocks.  Are there things like matching contributions they need to be aware of or conflicts of interests etc.  Eva said there are special requirements and guidelines within the call - there are no cost sharing this year, and they are allowing partnerships with industry.  And you have to demonstrate how you will integrate in the academic community.  Not saying it can’t be done.

Lee mentioned that they have had a lot of industry connections in the past and they are creating a hub for them to communicate, other federal agencies, other core groups outside so there is a conduit for these other stakeholder groups to get involved.

Ken Casey - They have had interactions with DATANET and DATAone.  Could Eva explain how EarthCube differs from what these groups are already trying to do?  Eva - there is a focus in DATAONE which was part of DATANET and the goal had been to support groups with cutting edge cyberinfrastructure and implement them, it was not focused on any domain, and has a strong core group on ecological data.  EarthCube is very domain focused, on academic geoscientists which might already be working with DataNEt and Dataone.  And understanding the resources available, who is well served and where does this community as a whole want to go in the future and this might have impact on DATAONE as well as providing them with user requirements for the future.  To integrate with the other resources that exist out there.

Christine asked another question - Eva mentioned that we should coordinate with the test governance - we will have some ideas and want to bounce them off a group, how should we do that?  Eva said, we are not requiring you, awards do not have to go through the test governance to coordinate.  They are a resource for you to have access to discussions in the past to see what other groups have done and as a resource for you in the community.  Feel free to email Eva and her group and give her ideas and thoughts and she can help you to understand how these ideas fit in any of the competitions.

Lee elaborated - at AGU town hall.  No you dont have to run it past the test governance, you are submitting it to NSF.  And they are happy to provide guidance and information about what is currently being done, but you are talking to NSF directly.  There are not structures or designs as how to deal with standards, there are no EarthCube standards or system designs - EarthCube is hoping to develop these but we are a long way away from that.

Ken asked another question - he mentioned all the groups which they work with, there is a broad range which make it clear on how to connect their data and he does not have a clear idea of how to do that with EarthCube or even if that is a goal of EarthCube.  He asked what it would be 5 years from now, for federal data centers etc. Eva said - the stage we are at right now is planning for the geoscience community as a whole, to have that to have a system architecture and that is why we started these pieces, to begin to have this dialog.  From the NSF standpoint we support 100 million cyberinfrastructure from different domains, that may or may not be integrated with other domains.  This is to bring the right data providers together and build the architecture from there.  The conceptual design award is to bring together architectures but there are communities which have not thought of specific tasks and are at different stages than more mature communities.  Lee mentioned the end user workshops, and how many of these groups are not ready to integrate, they do not have the standards within their own communities, and how they become cyber enabled.  Gave an example of EPA trying to integrate data but lots of silos exist.  He also discussed all the projects within NSF which might have ended up unconnected or silos if they even still exist.  Next week there is a data facilities workshop coming together in DC, about 60 based on an earlier steering group and hoping there is more of an effort in EarthCube in the future.  NASA and NOAA will be represented there.  

Ken asked a follow up - what about the fact that these data centers do not cover all the domains, so how would earthcube help some area of geosciences that does not have a data center?  Eva said she thought that yes, this is exactly what they would like to see.  It is a goal of the test governance and that it might provide a venue for academic scientists to go out and have a need met by resources or to identify that as a user requirement and well develop might be easier.  Governance would help have that dialog and serve these needs.  Eva talked about registering samples, and where to store them.  And differences in how groups define data - where apparatus might be data as well, and storing these designs are important.  And with new kinds of data and how to identify things and these are going to be ongoing discussions.

Lee mentioned that each of the workshops have been funded and each of the reports go back to that NSF program officer and maybe that will help coordinate these efforts.  and what role can we play to galvanize and facilitate these efforts.

Lee provided an overview of the outreach activities.  Erin Robinson is in charge of this effort.  They have set up a YouTube page and every one of the funded project has posted a 3-5 minute video related to their efforts.

Talking about the governance and how to get people engaged?  EarthCube governance is really a blank canvas and they are looking for community engagement.  Next week is their first outreach effort (the data facilities workshop).  They want the community to paint this picture, and are hoping to get enough feedback from the community this summer to paint a picture for NSF, and hope they will be able to give them an OK for year two.  There are a number of stakeholder groups they are engaging as well as crowdsourcing.  And if NSF thinks they are successful, they will fund a year two and get community consensus on a system and standards for moving forward.  They will do a call for the 3-5 year effort to coordinate the future efforts.

Stakeholder groups - see slides.  Current earthcube awards PI’s, the PI’s from the End User communities, professional societies, computer science, technology developers.  They have developed an advisory council and it will give them feedback on how to move forward.

Crowdsourcing, new website up and running.  It is not only an information site but also a workspace.

Community engagement

Steve from Scripts - gave an overview of the new website.  It has two sides, informational and the other is a workspace for communities.  It replaces the ning site.  EarthCube.org and workspace.EarthCube.org  it is evolving reflection of EarthCube and it is under the test governance.  

Steve went over the structure of EarthCube. See slides.

Steve mentioned what is available on the website for the workspaces.  It is built on open atrium, a drupal installation.  

There is a site demo on January 23rd 8am PST 11am EST.

In discussions there was a need for a list of resources, and once that was created, it was then made dynamic on the website to fit the communities needs.  Ken asked about sharing this metadata with other aggregate creators.

Last month at AGU, they experimented with engaging the community - they had a physical prize given for people to post data sources for posters they viewed at AGU - instead of asking the presenter and not sharing outward.  These are the kinds of activities they are engaging in.  They will do the same at ocean sciences and other meetings they are attending, and coming away with things that have long term value.  Ken asked - if they were doing it again at Ocean Sciences and his group would like to contribute.

Ken ? asked - do you see EarthCube as a community or multiple communities that interact?  Mentioned a NASA project with specialised communities that fit their needs..  Lee repeated the question and mentioned the work the test governance did last year - and there is not large cyberinfrastructure that has one overarching body.  So this is a community, but what might evolve is different bodies to handle different aspects of this and fill a role.  But starting out, this is like how the internet was designed by DARPA and then it was let loose once it was big enough. A single body to get things going but that it will take off and have a life of its own.  Steve added that scientific data infrastructure has diverse views - and why they have to be either?  Domain specific or connected.  He used states vs federal as an example.  In these early stages, this is to kickstart things.

Brian Wee comment - does it make sense to have our domain specific or integrated.  And it depends on the state of sciences.  Thinking back on the history and how systems were not aware that they interact and needed to integrate, and personally he does not understand how heliophysics and other fields of science, we are not there yet understanding the full connection, so full integration is not required.  Some communities that don’t see the value in interoperability, that is fine.  But years from now they might realize it.  This test that earthCube is doing is the science they want to promote, it is not a litmus test, but which groups make sense to have interacting.  Not an either or, but a both situation.

Ken Casey - systems of systems, or communities of communities - is there a view in EarthCube like GEOS?  Lee said Yes, but there is a jump here, they are focused on U.S. Academic geoscience but they need to connect to international collaboration which can not be funded through NSF.  The Belmont forum gives an opportunity for a single proposal across national domains.  But they realized that there are issues with infrastructure and data integration.  They are now working on an 18 month process of determining existing capabilities and gaps to bring these pieces together, and should have a report in may 2015 and then have a global solicitation for integration for global change research. They want to look at architecture and interoperability.  Each country has a 10 member representation, and they will look at best practices and needs.  And they will put nearly 20 million euros for the first project to create a global connection.

GEOS is not a member of Belmont but they are working with them and GEOS has been looking for resources to implement the vision they have been working on for the past 10 years.  Barb Ryan who was the head of Geo.  And someone else is a rep.  They have official observer status.

Bruce Caron - comment on community.  One of the governance issues, a large diverse group of people will have to determine what they have to be.  And it is building a community sense and forum and vocabulary for interaction so they can rapidly work together if they want to and not required to.  Like ESIP it is a federated organization.

Steve made a comment about being a data manager and data analyst and gave an example of two diverse data sets and how they had different names and they have been disintegrated.  And they are integrated in nature.  And we want to reintegrate them. So they don’t have to integrate these things in the future.  This is one of those first steps.