Transferability Working Group

Transferability Working Group

Presentation Summary

10^th Annual GHP Meeting

Montevideo, Uruguay

14 September 2004

Transferability seeks answers to questions such as

• 
  How portable are our regional climate models?
  
• 
  How much does “tuning” limit the general applicability to a range of climatic regions?
  
• 
  Can we recover some of the generality of “first-principles” models by examining their behavior on a wide range of climates?
  

The overall objective of the TWG is to understand the physical processes underpinning the global energy budget, the global water cycle, and their predictability through systematic intercomparisons of regional climate simulations on several continents and through comparison of these simulated climates with coordinated continental-scale observations and analyses. Transferability experiments are carried out by use of ensembles of regional climate models on multiple domains with all modeling parameters kept fixed during model transfer among domains.

We recognize that

• 
  The water cycle introduces exponential, binary, and other non-linear processes into the climate system
  
• 
  Water cycle processes occur on a wide range of scales, many being far too small to simulate in global or regional models
  
• 
  The water cycle creates spatial heterogeneities that feed back strongly on the energy budget and also the circulation system
  
• 
  Major advances in climate modeling hinge on advances in understanding and being able to simulate processes of the water cycle
  

From this we conclude that transferability is inherently a GHP activity.
The strategy of TWG is to:

• 
  Identify key processes relating to the water cycle and energy budget that express themselves to different degrees in different climatic regions
  
• 
  Create hypotheses that can be tested by use of ensembles of models on more than one climatic domain
  

Expected outcomes include:

• 
  Improved understanding of the water cycle and its feedbacks on the energy budget and circulation system
  
• 
  Improved capability to model climate processes at regional scales
  

The workplan for TWG is:

• 
  Phase 0: Write an article for the Bulletin of the American Meteorological Society summarizing lessons learned from various model intercomparison experiments and how “transferability experiments will provide new insight on the global climate system, particularly the water cycle and energy budget, amd report results of preliminary studies
  
• 
  Phase 1: Conduct pilot studies
  
• 
  Phase 2: Perform sensitivity studies on key processes relating to the water cycle. Create and test hypotheses
  

The timetable for TWG is:

• 
  October 2004: Presentation at WGNE
  
• 
  December 2004: Paper submitted to BAMS
  
• 
  February 2004: Organizational meeting (AMS Annual Meeting)
  

Transferability consolidates lessons from modeling and observations:

• 
  Models: Use experience gained from simulating “home domains”
  
• 
  CEOPS: Use dominant features of the water cycle and energy budget of each CSE to generate testable hypotheses
  
  • 
    Review what has been learned
    
  • 
    Identify unique climate features
    

Further points:

• 
  Transferability contributes to “horizontality” by pulling together modeling groups and CSEs and contributes to GEWEX Phase II objectives.
  

• 
  Transferability experiments can entrain the capacity and expertise of regional modeling communities (e.g., RegCM3 community) to provide the support needed to perform experiments, especially for Phase 2. Large communities like the RegCM3 community are globally distributed, so transferability testing offers the opportunity to engage climate scientists from many countries, especially developing countries.
  

• 
  Simple sensitivity tests, however, emphasize the importance of having well-structured, tightly controlled simulation programs when engaging large groups. Sensitivity to obvious modeling issues, such as choice of convection routine, and not so obvious issues, such as time-step size and computer processor, can yield differences between simulations that are comparable to model-observations differences. There is need to control carefully simulation design so that all simulations by one model use an identical version.
  

Progress report on the pilot study ICTS (Inter-CSE Transferability Study):

After the GHP meeting in Lüneburg (2003) representatives from two centers (John Roads and Insa Meinke from ECPC and Burkhardt Rockel from GKSS) came together at the Scripps Institute in March 2004 for a start up meeting. Later this year William Gutowski, Eugene Takle and Raymond Arritt from the Iowa State University joined the ICTS. Simulations are being carried out with three regional models: the regional spectral model (RSM / ECPC), the climate version of the Lokalmodell (CLM / GKSS), and the Regional Climate Model (RegCM3 / ISU).
With the expertise of regional modelers from each CSE currently five target areas CSEs have been defined: two areas in North America for (GAPP and MAGS), one each in South America (LBA and LaPlata), Europe (BALTEX), Australia (MDB), and Africa (AMMA).
The horizontal resolution of the regional models is about 50 km. Forcing data are from NCEP/NCAR re-analysis. A time period from July 1999 to December 2004 was chosen for the model simulations where the first six month were considered as model spin up. The same model setup (physics, numerics) are to be used for all CSEs. Results of model simulated precipitation compared with data from GPCP and CEOP reference sites for the CEOP1 period (July – September 2001) were presented for those areas analyzed so far.
Gene Takle

Burkhardt Rockel

Bill Gutowski