2008-01-31: ESIP AQ Cluster Telecon

Notes from ESIP AQ Cluster Telecon - 2008-01-31 On call:

• Carol Meyer
• David McCabe
• Mike and ?/EPA NC
• Rich Poirot
• Greg Leptoukh
• Rudy Husar
• Frank Lindsay
• John White
• Stefan Falke
• Terry Keating
• Steve Young?
• Rich Scheffe?

John - Phil/met with David/Terry Bring back to AQ

• What are needs for Forecast/Analyst
• Public/Media/Decision-makers
• Model evaluation (GEOSS provides way to search/diff datasets for integrated analysis) provide model

Big Air Quality Scenario:

• These are sub scenarios
  • Dust-Smoke
  • Data-Model synthesis
  • Forecasting

David - Three scenarios with different actors (end-users)

• Smoke/Dust - not forecast, understand the event/regulatory
• Model-Data synthesis - more science
• Forecast - oriented for public/public health

Felt needed to define first paragraph well - Is EPA completly on board? John - all on board

Number of activities will dovetail nicely with products that NASA is producing

Greg - Who is audience for scenario b? Recipients of activity End users - science community, will use test bed

Stefan - process, refined further once you see participants All same tools - brought together in diff ways for three scenarios

Rudy - Driver is end-user, what kind of interoperability and arch comp would help execute scenario? Techies calling for participation, properly string together components This is because they didn't look at longitudinal aspect (how to make something out of) Component - constellation, not just service, or dataset GEOSS is supposed to be broad - make sure to connect to other aspects of activities (standard interfaces)

Actual responding - how any agency solves a particular problem, BUT what kind of components do you contribute?

Rich Poriot - User Perspective

• certain adrenaline that takes over in some of the events - build own community of interest
• One community would be state/regional aq managers with exceptional event status
• Where is historical perspective - how trends change

Rudy - All three scenarios tie together Which part can you do under diff. circumstances? where things are well-monitored/not monitored.

• European - similar regulations, but agencies don't interact
• what is common between europe/us?

• people always gather for events
• post analysis if group agrees worth continuing.

David - Three communities of end-users will be used Frank - start with enduser of dust/smoke product - event notification, model evaluation (analyst), forecasting element with public health. Work back to middleware with scientist/analysts. Describe reusable generic components

Rudy - Question about smoke/dust - external, not user driven Analyst - data analysis/synthesis - not seperate independent of users. data - analysts - end user

Frank - don't know a priori who will use data/products

Rudy - will take actual REAL event - want to understand system better

David - GEOSS serves science, Science needs GEOSS need global model, satellites well-validated (societal needs) not necessary to peg to more concrete health outcome

Rudy - What is the big picture (observations/models (reality)/Societal Benefits) GEOSS is supposed to bridge two sides.

Science digest - tell what's really going on All societal benefit areas benefit All three parts of scenario a,b,c fit together - can't have one without the other

Stefan - three parts have different temporal analysis output from one is input of another

Rich ?? - different starting points (part of same network?) or user communities - benefit from having some specific examples. Easy to useful, but difficult to see examples

David - three parallel efforts are occuring now and will continue, more comfortable leaving three user communities seperate and marketing strategy that they all use similar things?

Rudy - Model, synthesis etc. are inputs to public Combine respective views -

Stefan - Do we pick particular societal benefit area? is there an issue being fire/smoke

David - Wildfire discussion can get away from air quality

John - Air quality has to be a

David - High global mortality, AQ

Frank - Event driven exceptional event/can also have air quality questions that aren't event driven/ More than just a gee whiz video

David - Benefits don't flow just to public/science too

Rudy - GEOSS hasn't gelled - no outcome yet, tremendous opportunity to get GEOSS what does this event mean to existing air quality Meaningful - it would be helpful to make sure that whatever event is given in context of general air pollution

David - grasping at this, look at some of the ADM highlighted - a lot event driven/not all (e.g. African Biodivesity).

Rudy Asking for generic scenario - user should be from societal benefit area. Takes all of the things you are talking about into consideration

Erin - are three parts seperate?

David - define seperate parts

John - will present users/high level needs/what questions all actors are going to use same tools. Not going to go through scenario a b c...

David - want a long-distance impact, not right on top of users. Long-range transport of dust/smoke then need satellite and models

Rudy - identify actors that have been proposed and at presentation. lay out generic terms, broad features other people will also make case for what else should be included.

David - Improve Science, disperate beneficiaries

Rudy - dual perspecitive - scenario proposed that explicitly includes science components, but all connected (forecasters, data integrators/emisison, post-process) instead of dividing

Terry - call for participation specific not build GEOSS, that we can try to do. Break down into more doable tasks. Model evaluation - if you can do that you can do many other things. Rudy - shooting way to broad, doesn't match examples on OGCNetwork.

Rudy- thinking of same size, but more longitudinal than lateral that you are explaining. doesn't need to cut off. Harvest what is already going on instead of saying that only forecasting going on. Why is it extra work? you make forecast visible/I make data integration visible and then together better explanation/notification downstream. Emphasize forecast/data integration is perfectly meaningful, but cutting off event analysis is going backwards. How do we harvest, be inclusive rather than declare artificial boundaries.

Terry - don't disagree that there is benefit that linking, the three scenarios aren't boxed in. Can approach from multiple perspectives. Instead of describing entire network. Look for way to combine approaches. Come concreteness - can't be all things/all people

Frank - One criticism of first round, dog/pony show. Videos created... If we were going to participate again, leverage to a more concrete product standing up loosely coupled system. Moving to persistent and really there. Not going to sell anyone with one user they connect to this system and there needs are satisfied. Air quality is a good scenario b/c it has broad number/type users.

Rudy - Implicit goal that frank articulated, seek movement forward toward the connections resulting in permanent system of system...