Talk:Data Summit Workspace

Data Summit Announcement -- Cwayland 1 November 2007 (EST)

This message is to inform you of a meeting on air quality data that we are planning for Tuesday and Wednesday, February 12 -13, 2008, at the main EPA campus in Research Triangle Park, NC. The purpose of this meeting is to bring together those organizations and individuals with key roles in retrieving, storing, disseminating and analyzing air quality data in order to explore efficient means of leveraging the numerous operations underway, and to assist EPA-OAQPS in defining its role in the larger air quality data community. The existence of several data organization systems that already provide and/or share information (e.g., Data Fed, ESIP, AirNow and AirNowtech, VIEWS, GIOVANNI, RSIG, and AQS) raises questions regarding their basic capabilities and interoperability. Although these programs have been catalyzed through various air quality management, research and GEOSS-related missions, we believe that some effort in considering a cooperative “data” strategy may yield exciting synergistic benefits for most of these programs. Internally at EPA, we want to benefit from the collective counsel of this community as the OAQPS AIRQUEST air quality data system evolves.

More information on the meeting objectives, agenda, logistics, and hotel arrangements will follow. Please let Rich Scheffe (919-541-4650; scheffe.rich@epa.gov) or Joe Tikvart (919-541-7793; tikvart.joe@epa.gov) know of your availability to attend. Once we have a compilation of planned participation, we will coordinate with you on goals and key topics for discussion. I thank you in advance for your consideration of this meeting.

Data Summit Table-- Tesh Rao 17 January 2008 (EST)

Data Summit Table - Word Doc
I expanded the table a bit, but not too much. I mainly look at it as link, that definitely needs to be maintained from am day one to pm day one to day two. It can be polished a bit as needed.

This could be one of the items handed to the breakout sessions for completion during its individual discussion. The (my kind of half-baked) idea it to use this to "tie" things together. Under each "business topic" category, the groups would identify which raw data systems and which data processing centers are most important via this table. Then, the group would discuss, theoretically, which of the chosen attributes are most important in summation when we begin talking about a "system of systems." {There may need to be more than one of those depending on application...which is what we want to discuss and best to streamline that :-) ). Those thought would go in the colored row at the bottom of each of the table subsections. Talk could even be focused on whether any data processing center that is existent and mature now can handle many of the attributes that we arrive at for each "business topic." Then the discussion can expand, as possible and do-able, to streamlining techniques across all the business topics. Beyond that, a few other points:

• In the 8 minute presentations in the morning of the 1st day, on top of the items we gave to Lou Sweeny as topics for each speaker to cover, I would add something that describes the maturity of the system in question (as well as perceived user community). Only after today's RSIG meeting did I realize that it has advanced so far in terms of implementation (though the number of users is probably small in that case). Also covered should be issue of sustainability for each of the systems. If the presenters have examples/papers/reports/policy related work that has been done using their system (this wouldn't really apply to some of them like AQS), it would be good to identify those briefly.
• We are addressing 4 very broad "business topics"--and the ones chosen are good--but we should think about anything we are missing (as far as future importance for OAQPS) under the umbrella of topics we have. Climate Change, for example, comes to mind.
• I am sure there will be discussion items that come up that aren't directly related to any of the structured discussion items, but may be important to the overall big picture. I would have the note-takers keep track of those in a "parking lot" list.
• When Tsirigotis/Curran were here, we had a discussion about how a data analyst spends 80% of any given project on data assimilation, data reduction, and general data readying. Only 20% of his/her time is spent on actual visualization/interpretation. The idea as we think about approaching a "system of systems" is for being able to reduce the 80% number for a data analyst. One of Airquest's initial goals was to do this. This could be framed as a discussion item for the breakout sessions. Even though we have a proliferation of tools now for accessing (ambient) data now compared to say 6-7 years ago, this 80/20 split hasn't gone down a great deal.
• Again, after today's RSIG meeting, we need to manage time very effectively, otherwise we could get caught in a very narrow discussion of relatively few topics. Need the facilitator to keep track of time well. And, you have it on the agenda, but next steps should be clearly delineated both at the beginning of the first day and especially at the end of the second day.
• To me, this is a very needed discussion and it needs to be done efficiently and well. There just seems to be so much overlap between systems that exist now (AQ, RSIG, etc.) and funds seems to be used for duplicative items. I'll stop here for now.