Challenges for Southern Hemisphere atmospheric and ocean sciences in the 21st Century(Invited Presentation)

. The scanty observational data base over the Southern Hemisphere has meant that substantial uncertainties exist in analyses, and weather forecasts are less accurate than in the Northern Hemisphere. Although the situation has improved with satellite observations, the latter need to be complemented by a strong in situ network. Observations indicate that climate change is underway and the prospect is for even greater changes in future as human effects become more evident relative to the natural variability. Accordingly, the weather and climate of the past will not be as useful a guide as it always has been, and a major challenge is to predict what the future climate will be, on many timescales. This means developing (i) an observing system that encompasses the atmosphere, ocean, land and ice fields, (ii) assimilation techniques to turn those into global fields, and (iii) much improved climate models that can make reliable predictions of interannual and interdecadal variations as well as the longer term trends. Of course the predictions will not be deterministic, but must include likely distributions, sequences and extremes of variables as well as their mean. This means that major advances are needed on several fronts, on observations, on analyses, on processes, on research quality datasets, and on the models as a whole. The Climate Variability and Predictability project (CLIVAR) of the World Climate Research Programme is one framework for these developments. In this talk we will touch on some of the exciting prospects for advance