At the opening of the discussion, Hartmut Grassl, director of the World Climate Research Programme (WCRP) of the World Meteorological Organization (WMO), presented an overview of the recently established Climate Variability and Predictability (CLIVAR) project. At its core this project includes the development, evaluation, and application of coupled numerical models of the atmosphere-ocean system. The goal of the project is the prediction and understanding of seasonal and interannual climate variability, such as the El Niño Southern Oscillation variations, as well as the assessment of climate change induced by forcing factors operating at centennial time scales, in particular the anthropogenic perturbation by the emissions of greenhouse gases and large-scale changes in land use. CLIVAR also supports the design and implementation of global observing systems for the physical state of the atmosphere, ocean, and land surfaces. Considerable progress has already been achieved in the atmospheric domain, spurred primarily by efforts to improve operational weather forecasts. Successes have also been noted in the oceanic domain with the implementation of the space-based TOPEX-POSEIDON system and the continuation of the in situ measurement arrays deployed as part of the previous Tropical Ocean Global Atmosphere (TOGA) project. However, deplorably, the Global Terrestrial Observing System (GTOS) is much less advanced. The main reasons for this failure are linked to a multitude of political difficulties in many parts of the world in establishing local components of a global observational system.