As discussed in section 1, the implications of drought vary among the island locations. Agricultural and other water consumption needs are unique to each island. An island may or may not have one or more useful reservoirs. The residents of some islands may use diluted salt water that is bathable but not drinkable. The effects of ENSO may interact with the water storage/usage situation in such a way as to threaten the welfare of some islands much more than that of others. Thus, successful forecasting of the ENSO situation may be highly beneficial for certain locations. Seasonal forecasts of ENSO are already routinely issued by several independent organizations (e.g. from NCEP using the statistical canonical correlation analysis [CCA; Barnston and Ropelewski 1992] and a two-tiered comprehensive dynamical model system [Ji et al. 1994]; from Lamont-Doherty Earth Observatory using a simple coupled model [Cane et al. 1986, Chen et al. 1995]; and dynamical models from innumerable others such as Scripps Institution of Oceanography [Barnett et al. 1993], the Center for Ocean-Land-Atmosphere Studies [Kirtman et al. 1997], the Australian Bureau of Meteorology Re-search Center [Kleeman 1993], the European Center for Medium Range Weather Forecasts [Balmaseda et al. 1994], and other European dynamical ENSO-related modeling efforts [Harrison, et al. 1997, Carson 1998]). Some of these modeling efforts are still in a semi-experimental phase. These forecasts can be used cautiously as forewarning devices, as warranted by their estimated skill. Pacific island rainfall may also be forecast directly, using but not explicitly showing the ENSO state, as in the statistical forecast systems described in Barnston and He (1996) and He and Barnston (1996).

Use of the basic rainfall data presented in this atlas is anticipated in the context of a more detailed study by the present authors. Such a study would address not only the rainfall anomaly history, but also the associated drought consequences to the islands. Among other considerations, the study would lead to more refined quantifications of the rainfall deficit/duration thresholds associated with drought on an individual island basis. The addition of rainfall data covering the very strong El Niño of 1997-98, still in progress at the time of this writing, will likely contribute further to our knowledge and understanding of the Pacific basin rainfall responses.