Report of the CSM Natural Variability Working Group

[an error occurred while processing this directive] [an error occurred while processing this directive]
Report of the CSM Natural Variability Working Group

Jim Hurrell, David Battisti, and Ed Schneider, Co-Chairs

23 and 24 June 1998

The Village at Breckenridge

The purpose of the CSM Natural Variability Working Group is to encourage and facilitate study of climate phenomena on all timescales by the community using CSM. The discussions, therefore, centered on two issues: (1) the utility of the current version of the coupled model for such studies, and (2) the accessibility of the simulated data.

The working group met on Tuesday afternoon and Wednesday morning, 23 and 24 June 1998. A series of short presentations focusing primarily on existing analyses of natural variability in CSM were made. The speakers and their topics included:

Maurice Blackmon (NCAR): Introduction of existing integrations and presentation of existing plans for model changes/integrations;

Grant Branstator (NCAR): Annual cycle of interannual variability in the Community Climate Model version 3 (CCM3) and partitioning of internal versus external variance;

R. Saravanan (NCAR): Impact of convection coding bug on existing simulations, signal-to-noise ratio in seasonal prediction using CCM3 compared to other models, and simulation of the tropical Atlantic in CSM;

Jim Hurrell (NCAR): Comparison of Global Sea Ice and Sea Surface Temperature (GISST) and the National Centers for Environmental Prediction (NCEP) sea surface temperatures (SSTs) and the impact of differences on CCM3 ensembles;

J. Shukla (Center for Ocean-Land-Atmosphere (COLA) Studies): Signal-to-noise ratios in CCM3 versus other models;

Bette Otto-Bliesner (NCAR): ENSO signals in the paleo-CSM;

Hsin-Hsin Syu (NASA/Jet Propulsion Lab): Discussion of new hybrid coupled model with the NCAR Ocean Model (NCOM) and testing of different mixing schemes;

Rong Fu (University of Arizona): CCM3 variation of upper tropospheric humidity in response to the El Nino-Southern Oscillation (ENSO);

Jerry Meehl (NCAR): ENSO response to greenhouse warming in CSM and other coupled models;

Ed Schneider (COLA): CSM simulation of annual cycle and interannual variability of SST in the tropical Pacific;

Following these presentations, a general discussion and a formulation of a statement of priorities ensued. The consensus was that the outstanding priority in CSM should be work aimed at improving biases in the mean tropical SSTs and the simulation of the seasonal cycle in the tropical Pacific. The relatively poor simulation of these aspects adversely affects the ability of the CSM to reproduce ENSO variability and related global phenomena. Diagnoses of tropical/extratropical interactions on all timescales are, consequently, limited with the current version of CSM.

To address this priority, the CSM Natural Variability Working Group recommends that the relationships between surface fluxes, SSTs, distribution of precipitation, and atmospheric circulation be systematically examined and evaluated through a series of process experiments designed to lead to improvement of the atmospheric component of CSM. These process experiments should include short (e.g., 10 year) integrations with fixed SSTs as lower boundary conditions, in addition to a series of short, fully-coupled runs, with the latest version of the CCM and an ocean model with enhanced meridional resolution at low latitudes (e.g., x2'). Studies investigating the sensitivity to planetary boundary layer (PBL) and stratus cloud formulations, solar absorption schemes, component model resolution, and convective cloud and heating parameterizations are believed to be most relevant to the improvement of the tropical simulation in CCM and, thus, are of highest priority.

The CSM Natural Variability Working Group should and will work closely in collaboration with members of the CSM Atmosphere Model Working Group to coordinate, diagnose, and evaluate the sensitivity experiments. A meeting in Boulder (autumn, 1998) was proposed to coordinate a series of sensitivity experiments to address the key issues and weaknesses in the CSM. The meeting will be open, but it is crucial that members of the CSM Atmosphere Model Working Group attend, in particular.

A critical step in the analysis and evaluation of model runs will be the definition of a standard set of diagnostics produced using a standard set of software. In this way, the results of sensitivity experiments can be systematically and objectively quantified. This goal, relevant to all working groups, will require additional human resources to develop and maintain the analysis package.

The working group session concluded with a discussion of infrastructure. In particular, to ensure better community involvement in the CSM effort, major issues include communication and improved access to CSM data. The CSM Natural Variability Working Group recommends that quarterly summaries of significant CSM-related activities (i.e., activities of both CSM project scientists and working group members) be provided to the CSM Scientific Steering Committee and all CSM scientists. In addition, a web-based bulletin board and access to model output via the web are desired objectives.

List of Participants and E-Mail Addresses:

George Lai, lai@dao.gsfc.nasa.gov

S.J. Lin, lin@dao.gsfc.nasa.gov

R. Saravanan, svn@ucar.edu

A. Robertson, andy@atmos.ucla.edu

A. Seth, seth@ucar.edu

J. Bergman, jwb@cdc.noaa.gov

Dailin Wang, wangd@soest.hawaii.edu

Andrey Proshutinsky, prosh@ims.alaska.edu

Andrea Hahmann, hahmann@air.atmo.arizona.edu

Paul Dirmeyer, dirmeyer@cola.iges.org

Bette Otto-Bliesner, ottobli@ucar.edu