links between the climate research community and the communities of applied climatologists, social scientists, and users of climate information. TOGA has brought us to a time for the establishment of a prototype international institute for making predictions and demonstrating the applicability of these predictions, as previously recommended in NRC 1995b.

TOGA concentrated on only the strongest climate variation on seasonal-to-interannual time scales, ENSO, and concentrated on ENSO only in a limited geographic region where the signal is strongest, the tropical Pacific. Much research is still required to develop skill for predicting short-term climate variations caused by other processes or in other places. Unless they coordinate their efforts, researchers are unlikely to develop this skill efficiently. National and international programs are needed for research on, and development of, the exploitation of predictability and the making of predictions of seasonal-to-interannual climate variations throughout the world. In particular, the World Climate Research Programme study CLIVAR (Study of Climate Variability and Predictability)/GOALS (WCRP 1995) and its U.S. contribution GOALS, proposed in reports from the NRC (1994b, 1995a), provide a path for furthering the accomplishments and building on the legacy of TOGA.