|
The following was a note written by Roy Jenne (dated 3 Dec 1993): The 100-year GFDL transient runs were made with a dynamic ocean and an atmosphere. The first run (around 1989) was made on the CDC 205. It was published in the first IPCC reports. NCAR has decade averages for each season from this run. GFDL later found that the run would be a little better if the ocean time step was a bit shorter. GFDL ran the model again (around 1991) on the Cray YMP. Year by year, it will differ from the first CDC 205 run because of natural variability. In December 1993, NCAR received the 10-year monthly, seasonal, and annual climatologies from this run from Ron Stouffer at GFDL. Data are in 32-bit IEEE format. There is only one thing that ties the data to a particular date. That is the equivalent CO2 of 300 ppm (valid around 1950) used in the model. More important, the flux corrections essentially make the starting time valid when the SST was valid, probably around 1960. The sea ice dataset has a valid date around 1958, meaning that it also used earlier data. |
