Synopsis:
Case g006.35 is the experiment that worked acceptably. This case used a data file atmosphere model (data is daily average fields from the uncoupled CCM3) to spinup the active ice and active ocean models. The ice and ocean models are essentially those presented at the May 1996 CSM Workshop in Breckenridge, but do have some improvements. The ocean was run for 50 years in accelerated mode and over 10 years in synchronous mode. This is equivalent to over 500 years of spinup for deep ocean tracer fields.

This g006 ocean & ice spinup run is intended to replace case g005. Unlike case g005, case g006 is very "clean" in the sense that no code changes (eg. physics improvements and bug fixes) were made as the simulation progressed. In case g005, time constraints associated with the CSM Workshop forced us to upgrade the model codes as the simulation progressed.

Our initial impression is that this spinup run was quite successful. In the later half of the simulation, climate drift was either very slight or imperceptible. Also, when used to provided the initial conditions for a fully active coupled run (see case b003 ) there was no shock to the system, adjustments within the ocean and ice models were very small.

Computer Resources:
This simulation required almost 1000 CPU hours on a Cray Y-MP with a 6 nano-second clock. When the ocean code ran in accelerated mode (years 0 thru 49), 9.4 CPU hours per simulated year were required. When the ocean code ran in synchronous mode (years 50 thru 60), 46 CPU hours per simulated year were required. In either case, 32 mega-words of main memory were required.

Table of g006.nn runs: