| Summary: | The development of a long range streamflow forecast model using oceanic-atmospheric oscillations becomes useful when planning for future water supplies. A data-driven model, i.e. M5P, uses proxy reconstructions for streamflow forecasting with 1–5 year lead-times. The proxy reconstructions include annual oceanic-atmospheric indices including El Niño Southern Oscillation (ENSO), Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), and Atlantic Multidecadal Oscillation (AMO) ranging from 1661–2007. The analysis focuses on naturalized streamflow for Lee's Ferry located in the Upper Colorado River Basin (UCRB). A10-fold cross-validation technique is used to test the efficiency of the model. The best results are obtained for 2-year lead-time. The proposed methodology outperforms the traditional linear regression modeling approach. The use of proxy reconstructions provides a robust model that is trained on a larger dataset than models trained only on the instrumental record. The proposed modeling technique is expected to be useful for water managers for long range water resources management within the Colorado River Basin.
|
|---|