AQ Cluster Meeting Notes

July 20, 2006 10:30 -

• Introduction: R.Husar

• Unidata: Mohan
  • Created to provide data to community (30 streams) - 200 universities in subscription service worldwide (U.S., Latin America)
    • Universities primary audience, but not the sole audience. System is open.
    • Also redistribution of data once reaches the university - other applications like public health.
  • Important aspect to this cluster netCDF - HDF IO in 6-9 months
  • Started as meteorological data, but become increasingly more interdisciplinary e.g. hydrology, AQ data, etc
  • AQ Community also has need for atmospheric datasets so the flow is two ways
  • Unidata is a community of communities (oceanography, GIS, AQ)

• Husar: Thrilled to have Unidata offer to distribute AQ data. Portal idea: move data from provider to user - centralized approach. Distributed approach, move data as close to user as you can get - surface area approach, putting it in as many places as you can - multiple paths to the same place. RSS - provider publishes whoever subscribes can have it in their home setting, Unidata is an RSS system. Bring individual projects into the cluster?
  • Function of ESIP - required to be there and it took a while to set own direction and should be self-motivated.
  • ex. DataFed - We keep pushing certain activities harder than others, will that help or choke cluster activities?
  • Clusters microcosom of federation - group finds internal dynamic that works.