TRANSIENT RESPONSES OF THE CSIRO CLIMATE MODEL TO 2 DIFFERENT RATES OF CO2 INCREASE

The responses of the CSIRO coupled atmosphere-ocean-sea ice model to t wo greenhouse gas induced warming scenarios are described and compared to a control run with the current CO2 level. In one scenario, denoted IS92a, the atmospheric CO2 increases such that it reaches doubling af ter 128 years. In the other, the CO2 increases at 1% per year compound ing (doubling after 70 y). As the CO2 increases in both scenarios, the top-of-atmosphere outgoing longwave radiation increases giving enhanc ed cooling of the coupled system, while the outgoing short wave radiat ion decreases contributing to a warming of the system. The latter over compensates the former leading to a global mean net radiative heat gai n. The distribution of this heat gain produces the well-known interhem ispheric asymmetry in warming, despite a decrease in the sea ice aroun d Antarctica in this model. It is found that the volume mean temperatu re response over the southern ocean is greater than that over the nort hern hemispheric oceans, and a maximum warming takes place at the subs urface rather at the surface of the ocean in the southern mid-to-high latitude region. The enhanced high-latitude freshening associated with the strengthened hydrological cycle significantly affects the latitud inal distribution of warming and other responses. It enhances the warm ing immediately equatorward of the deep water formation regions while produces a reduced warming, even a cooling, in these regions. In both runs, there is a decrease in the large-scale oceanic currents which ha ve a significant thermohaline-driven component. The reduction in these currents reduces the poleward transport of salt out of the tropical a nd subtropical regions of these oceans. This and the enhanced evaporat ion contribute to considerable increases in surface salinity in the tr opical and subtropical regions. In IS92a, the warming rate before doub ling is smaller than that in 1% scenario, but the cumulative effects o f the two experiments at the time of doubling are similar. Nevertheles s, significant contrasts exist. For example, at the time of doubling i n IS92a, the warming of the upper ocean is greater because a more deve loped temperature-albedo feedback occurs. In addition, a longer time i s allowed for heat anomalies to spread downward, and so the effective heat penetration depth is greater than that in the 1% scenario. Thus t he oceanic response is influenced by the CO2 increase scenario used.