Effects of Data Extension on Drought Probabilities

by Dennis R. Horn, Univ of Idaho, Kimberly, ID, USA,

Document Type: Proceeding Paper

Part of: Engineering Hydrology


A condensed-parameter disaggregation model was applied to the monthly streamflow records on two rivers in Idaho, to investigate the occurrence and probabilities of drought periods. Using a 40,000 year sequence of generated flows, the cumulative probability density functions of maximum negative run-sums and run-lengths were obtained and the return periods of historical negative runs were estimated. A multi-variate stochastic model was then applied to the original streamflow sequences, to extend the historical data based on the longer term records of nearby streamflow gages. From these extended records, new model parameters were determined and the generation process repeated with the revised parameters. A comparison of the results, based on the extended and unextended station records, indicated significant differences in the probabilities assigned to historical drought periods. Study results are discussed.

Subject Headings: Hydrologic data | Droughts | Probability | Parameters (statistics) | Stochastic processes | Water resources | Mathematical models | Streamflow | Idaho | United States

Services: Buy this book/Buy this article


Return to search