#persist

CCIT and EVA Teleconference:

Steve Kelling
Matt Jones
Dave Vieglais
Bertram Ludaescher
Bob Cook

Toll Free Number:  888 551 7631
Pass code: 7813231


Goal:
Identify and develop data intensive exploration, visualization, and analysis exemplars that can serve as the foundation for the creation of workflows and DataONE analysis and visualization functionality. 

Focus on Carbon Cycle Models

Background:

-a number of Model Intercomparision Projects (MIPS) have occurred over the past 20 years or are occurring now
    VEMAP, LBA-MIP (Brazil), three North American Carbon Program MIPS, C-MIP5, Multi-scale Synthesis and Terrestrial Model Intercomparision Project (MsTMIP), Agri-MIP, etc.
-each group builds their own system, including scoring and visualization
-a DataONE CI system and standard scoring / benchmarking would enable 

-ILAMB is busily conducting scoring for the IPCC (papers due in Summer); EVA will build on ILAMB and apply to MsTMIP


Approach

1.  Model Benchmarking Cyberinfrastructure
       -system to allow modelers to compare model output to observations 
        -system has scoring /benchmarking code, key observations as a function of time / space/ and biome
            http://www.ilamb.org/benchmarks/
            http://www.climatemodeling.org/c-lamp/results/diagnostics/CN_vs_CASA/
        System will take the following as inputs
            -standardized model output
            -observations (perhaps provided in a standardized manner on DataONE)

        -opportunity for Scientific Workflow  and provenance  
        -Dave Koop (Arizona, formerly a Claudio grad student) is involved in Provenance WG
        
2.  Visualizations for Benchmarking
    -tied to Model Benchmarking CI (above)
    http://www.climatemodeling.org/c-lamp/results/diagnostics/CN/ameriflux/FortPeck_model_vs_ob.png


3.  Carbon / Climate Model Output Visualization 
    -build off of BirdVis activity (DataONE eBird plus custom software from Claudio's group)
    -http://www.ornl.gov/filedownload?ftp=e;dir=uP17NktRq768
            
            
Next steps:
April EVA Working Group meeting
Claudio and student coming to ORNL to discuss #3

Resources:
EVA WIKI: http://dataoneeva.pbworks.com/w/page/48275384/FrontPage


What should CCIT or the Workflow Provenance WG tackle?
   - Focus on data access to the tools
     - storing model output data (in Thredds/OPeNDAP/NetCDF)
     - storing observations data (for comparisons)
     - BirdViz requires a specific schema

ProvWG Notes:
Goal: use old and/or new EVA use cases to validate provenance model developed by ProvWG
=> Science side: need scientific questions that provenance information allows to answer 
=> Technology side: what are the right observables to record, how to instrument "the code" (hairball ;-)
=> get Claudio and/or Juliana and/or Dave Koop involved in 
development of joint ProvWG & EVA use case