Increasing streamflow forecast lead time for snowmelt-driven catchment based on large-scale climate patterns

Abstract

This study focuses on improving the spring–summer streamflow forecast lead time using large scale climate patterns. An artificial intelligence type data-driven model, Support Vector Machine (SVM), was developed incorporating oceanic–atmospheric oscillations to increase the forecast lead time. The application of SVM model is tested on three unimpaired gages in the North Platte River Basin. Seasonal averages of oceanic–atmospheric indices for the period of 1940–2007 are used to generate spring–summer streamflow volumes with 3-, 6- and 9-month lead times. The results reveal a strong association between coupled indices compared to their individual effects. The best streamflow estimates are obtained at 6-month compared to 3-month and 9-month lead times. The proposed modeling technique is expected to provide useful information to water managers and help in better managing the water resources and the operation of water systems.