Influence of Climate Variability on Streamflow Variability: Implications in Streamflow Prediction and Forecasting

摘要

An improved understanding of climate variability can lead to a better description of hydrologic variability with many applications. An investigation of the association of surface climate and streamflow with a global scale climate phenomenon, the El Nino - Southern Oscillation (ENSO), showed statistically significant relationships in two regions of the western U.S.; the Pacific Northwest (PNW) and the desert Southwest (SW). Within these regions, the strongest associations were noted in the mountainous areas and streamflow exhibited a stronger relationship than precipitation. During ENSO events, drier conditions were observed in the PNW while wetter conditions occurred in the SW. The strongest associations were observed between annual or winter season precipitation and an index of the ENSO averaged over the previous summer. Several different methods for water supply forecasting were also evaluated including approaches which consider associations with large scale climate features. Statistical regression equations and a water balance model were applied to the problem of forecasting the annual runoff volume for the Wilson River in western Oregon. The statistical approaches were shown to be better than the water balance approach. Of the alternative statistical approaches, the equations based only on data available up to the time of the forecast produced the best results, particularly early in the forecast season. An index of the ENSO was an independent variable in several of these equations.