understanding of ENSO has been achieved, theoretical models for the mechanism of ENSO have been proposed, and the beginning of an understanding of the relationship between tropical sea-surface-temperature perturbations and the atmospheric response in the middle latitudes has been attained.
However, TOGA failed to meet its objectives fully. The program did not live up to its name by completing studies throughout the tropical oceans and the global atmosphere, but instead concentrated only on the large signal of seasonal-to-interannual coupled atmosphere-ocean interactions in the tropical Pacific (the ENSO phenomenon). Variability and predictability arising from processes in other ocean basins, and from the interactions of the atmosphere with land and with ice, received scant attention. Understanding of the effects of tropical sea surface temperatures on the higher latitudes developed slowly. The program did not address the possibility of variability and predictability on seasonal-to-interannual time scales arising from interactions within the middle latitudes. Research on seasonal-to-interannual variations of atmospheric circulation conducted under other auspices was not well integrated with the TOGA Program. Implementation of both the COARE field program, the largest process study in TOGA, and the Tropical Atmosphere/Ocean (TAO) array, the heart of the TOGA Observing System in the tropical Pacific region, were both accomplished only near the end of TOGA, so that their usefulness for improving prediction could not be demonstrated within the time frame of the TOGA Program itself. The expected satellite-based observation system, especially a scatterometer for estimating surface winds, did not materialize. Observational strategies for understanding seasonal-to-interannual variability were not developed for anywhere but the tropical Pacific. These failures, combined with the unexpected difficulties in coupling complex atmosphere and ocean models, meant that while much progress was made, the full accomplishment of the TOGA objectives will be realized only by future programs.
By developing short-term climate predictions (predicting tropical Pacific sea surface temperatures months to a year or so in advance), TOGA has transcended the confines of a research program. Climate prediction (and the applications of climate prediction) has become a motivation for limited-duration research programs, and will continue to demand attention far into the future. Tension developed during the second half of TOGA between research motivated by the ideal of operational prediction and actual operational prediction efforts. Incipient applications of climate predictions were initiated under TOGA and a prototype International Research Institute for Climate Prediction was designed. Such an institute would support the needs of individual regions, through Regional Applications Centers. Peru, Brazil, and Australia are already finding applications of short-term climate predictions useful to their own economies and