Regression models in water supply forecasting

TitleRegression models in water supply forecasting
Publication TypeConference Proceedings
Year of Conference1982
AuthorsHuber, A. L., and Robertson D. C.
Conference Name50th Annual Western Snow Conference
Series TitleProceedings of the 50th Annual Western Snow Conference
Date PublishedApril 1982
PublisherWestern Snow Conference
Conference LocationReno, Nevada
KeywordsForecasting, Regression Models, Snowmelt, Water supply

This paper presents a review of some of the most common index or regression methods for making seasonal water supply forecasts. It outlines a strategy for adapting the multiple regression technique to the problem of real or near real-time forecasting and concludes that the regression technique will remain a viable tool of the water supply forecaster because: 1. It is conceptually simple and widely accepted; 2. Confidence intervals about the estimate may be easily calculated; 3. It requires a minimum of computational facilities; 4. The forecast estimates are available instantaneously; 5. The forecast errors are acceptably small.