MERGING AND ERROR ANALYSIS OF REGIONAL HYDROMETEOROLOGIC ANOMALY FORECASTS CONDITIONED ON CLIMATE PRECURSORS

Forecasts of hydroclimatic variables and incorporation of their error bounds are invaluable in water resources planning and operations under uncertainty; In this study, regional long-term operational hydrologic forecast models conditioned on climatic precursor are presented. The forecasts also include uncertainty intervals and confidence limits. Th e forecasts are based on the temporal and spatial variability of hydro meteorologic anomalies and their relationships with climatic interannu al and intraseasonal El Nino-Southern Oscillation (ENSO). The forecast skills of the proposed model, which incorporates ENSO forecasts on tr opical rainfall and streamflow, are compared with those that are uncon ditional and do not incorporate ENSO. Significantly improved skills ar e achieved by incorporating forecasted ENSO indices and their errors. The seasonal variability of the forecast model skills are also evaluat ed. These ENSO-based forecasts of regional and seasonal-to-interannual hydrometeorologic variables consistently merged with systematic error analysis can provide outputs for direct use in water resources planni ng and operation under uncertainty.