[an error occurred while processing this directive] [an error occurred while processing this directive]


Report 4

10 August 1998

 Chairs: John Kutzbach, University of Wisconsin and Bette Otto-Bliesner, NCAR


The CSM Paleoclimate Working Group held a meeting on 24 June 1998 in Breckenridge, Colorado as part of the CSM Workshop 1998. Scientists, programmers, and graduate students representing institutions involved in paleoclimate studies participated.

Presentations were made by staff of the NCAR Paleoclimate Program on the status of the PaleoCSM:

1) Brady and Felzer described the NCOM slab model, developed in conjunction with the Oceanography section. This slab model is the NCAR ocean model collapsed to one 50 meter layer with oceanic heat transport either prescribed from observational data or calculated as a diffusive mechanism based on SST gradient, latitude, and land-sea distribution. Results and a User's Guide have been posted on the Paleoclimate Working Group web page.

2) Brady reported on a series of experiments exploring the sensitivity of the present-day ocean simulation to degraded initial prescriptions of temperature and salinity. After 50 surface/500 deep ocean years, the horizontal streamfunction and ACC transport for the sumulated present-day ocean equilibrate to similar states. The meridional overturning and surface salinities in the North Atlantic are significantly different depending on the initial state. These results have important implications for past climatic periods for which temperature and salinity conditions are poorly known.

3) Shields finished the NCAR reports with a discussion of tools developed at NCAR allowing paleo-researchers to easily convert their boundary conditions (continental configuration, topography, land surface types) to initial condition files needed for CCM3 and LSM. These tools supplement those developed by Brady for the ocean and sea ice models and presented at the Paleoclimate Working Group meeting in February 1998.

4) Matt Huber, a graduate student working with Lisa Sloan at University of California-Santa Cruz, finished the morning's presentations detailing his preliminary results modeling Eocene warm climates with CCM3.

The remainder of the Paleoclimate Working Group meeting was devoted to evaluating priorities and identifying collaborations for simulations to be conducted in the next year. The following CSM simulations (with identified collaborations) are proposed:

1) The Paleoclimate Working Group is ready to start two community CSM simulations, a 500-year simulation to determine the non-externally forced natural variability (the first in a series of simulations to look at the climate of the 17th-18th-19th centuries) and a 500-year mid-Holocene (6000 years before present) simulation. Both coupled simulations will be run at T31,x3' resolution. The NCAR Paleoclimate Program will take the lead in running the former and the University of Wisconsin the latter.

2) The other major effort will be concerned with extending the ocean spinup investigations to past time periods. University researchers will take some lead roles. Planned sensitivity simulations include: warm climates of the Paleocene/Eocene (Sloan/Huber-UC Santa Cruz), geometry/bathymetry (Barron/Peterson/Nong-Penn State, Sloan/Huber-UC Santa Cruz), Last Glacial Maximum (NCAR Paleo, Fawcett-Univ. of New Mexico, Felzer-PALE), and freshwater impulses (NCAR Paleo and Oceanography, Peltier/Gaca-University of Toronto). University researchers will also be pursuing independent paleo-simulations.

The following recommendations were made by the group:

1) Optimization of the low-resolution CSM (T31, x3') on the SGI Origin should be a priority. The ability to run future versions on vector machines such as the Cray J9 series would be useful for university researchers.

2) Proposed new parameterizations to the CCM need to be benchmarked at a range of resolutions before adoption. The inclusion of a spectral option in future versions of the CCM to allow T31 simulations should be considered.

3) A CSM diagnostics' processing package based on NCL should be developed.

4) A "hands-on" CSM User's Workshop should be scheduled.