Siobhan P. O'Farrell CSIRO Division of Atmospheric Research Private Bag No 1, Aspendale Victoria 3195, Australia Email: spo@dar.csiro.au
and
William Connolley British Antarctic Survey, High Cross, Madingley Road Cambridge CB3 OET, UK Email: wmc@bas.ac.uk
The accompanying paper on the trend in warming simulated over the last century by three climate models, showing that the models do not match the warming seen in observations, raises questions on how accurate the predictions from these models may be in the Antarctic region for future climate. Issues to be addressed concern both the impact of regional drift and the inter-decadal variability in both control and transient simulations. Regions of ocean convection in the Weddell Sea have shifted in the early part of the CSIRO control simulation as evidenced by changed pathways of a passive tracer. The MPI simulation also shows considerable drift in the Bellinghausen sea. Both these effects have biased the air temperature response to increased CO_2 over the Antarctic Peninsula. The ocean noise caused by convective events has lead to localized signals as large as the potential warming signal of the model in the last century. So to improve the signal to noise ratio we will now examine the greenhouse signal in the following century when the rate of CO_2 increase has accelerated to see whether the pattern of warming in the models over the Antarctic region is reinforced and what pattern if any develops over the Antarctic Peninsula. At least two of the models show some evidence of an enhanced warming in the area of sea ice retreat in the Bellinghausen sea adjacent to the peninsula but the signal is not present in all seasons and does not strengthen with higher CO_2 concentrations.
As well as these three simulations which were initiated last century several other simulations were performed by CSIRO which have different initial conditions and different rates of CO_2 increase and higher CO_2 concentrations are imposed. Some of the earlier CSIRO simulations have shown periods when there is an enhanced warming over the Antarctic peninsula for particular decades but the pattern does not always occur. The earlier UKMO and MPI climate change experiments did not have an enhanced warming in the Peninsula region.
The tendency of the climate models to refuse to produce a significent enhancement in warming around the Peninsula, even with enhanced forcing, suggests that either the Peninsula warming is a response to forcings not included in the models or that the Peninsula warming is sensitive to effects poorly modelled. There are many possibilities for the latter - the model resolution may not be high enough, which among other effects may mean that the topographic barrier that the Peninsula presents in the real atmosphere is not well represented in the models. The position of the ice edge may be crucial; and the Antarctic Circumpolar Current, poorly represented in all GCMs, may have important effects.