Project: CCSM pop2 Acceptance Testing 

Goal:
  Make the transition to a pop2-based model in CCSM.
 
Motivation:  
  Now that certain basic CCSM features have been added to LANL pop2,
  we want to assess and understand the differences between the two models.
  This process will involve close scrutiny of differences between the
  models.
  
 
Authority:
  Peter Gent
 
Project Lead:   
  Nancy Norton, software engineering
 
Scientific Lead:
  Gokhan Danabasoglu, science

Plan:          
   I) Side-by-side differences between pop and pop2
 
        * Develop tools to aid the comparison of the model solutions
        * Identify pop2 deficiencies (eg, netCDF
          history files are not CF-compliant and lack unlimited time 
          dimension, scale factors, masking,etc; CCSM naming convention
          requirements need to be satisfied)
        * Remedy pop2 deficiencies
        * Instrument the pop and pop2 codes to help to identify sources 
          of differences between the two models
        * "Base Code" Difference Analysis  (without new CCSM developments)
             o "hunt and destroy" process to identify and eliminate
                errors at startup
             o "hunt and destroy" process to identify and temporarily
                eliminate equivalencies that introduce small differences
                (eg, div, grad, shift computations) at startup
             o certify codes are equivalent at startup
             o document errors and known equivalencies 
             o document the testing cases used and the output from
               the tests
 
        * New Development Difference Analysis
          After base codes are certified to be equivalent, confirm the
          correctness of the new CCSM pop2 developments in the same context
          as the base code was certified:
             o chlorophyll
             o ideal age
             o "partial coupling" forcing option
             o marginal-seas balancing
             o correct handling of ice and snow in forcing coupled
   II) Testing:
        * Compare energy diagnostics 
        * Test hybrid starts  (the *.rh.* problem)
   III) Cleanup:
        * Remove temporary modifications used to eliminate known
          pop/pop2 differences in stage I
   IV)  Final Accpetance:
        * Comparison against CCSM3 pop1.4.3-based solutions
          long-term integration
        
Completion Criteria:
    
    This project will be completed when the pop2 solutions are
    accepted by the Project Authority, the Scientific Lead, and 
    selected OMWG members.
 

Start Date:  June 2005   
                         
 
Status as of 15 August 2005:      
        
   I) Side-by-side differences between pop and pop2
 
        * Tool development -- done
        * Identify pop2 deficiencies -- done
        * Remedy pop2 deficiencies -- 75% done
        * Instrument codes -- done
        * "Base Code" Difference Analysis
             o "hunt and destroy" process to identify and eliminate
                errors at startup -- done
             o "hunt and destroy" process to identify and temporarily
                eliminate equivalencies -- done
             o certify codes are equivalent at startup --done
             o document errors and known equivalencies -- in progress
             o document the testing cases and output -- done
 
        * New Development Difference Analysis
          After base codes are certified to be equivalent, confirm the
          correctness of the new CCSM pop2 developments in the same context
          as the base code was certified:
             o chlorophyll -- done
             o ideal age   -- done
             o "partial coupling" forcing option -- next
             o marginal-seas balancing -- done
             o correct handling of ice and snow in forcing coupled -- done
   II) Testing:
        * Compare energy diagnostics -- done
        * Test hybrid starts  (the *.rh.* problem) -- tbd
        * Comparison against CCSM3 pop1.4.3-based solutions
          long-term integration -- tbd
        
Estimated Completion Date