Premium
A condensed disaggregation model for incorporating parameter uncertainty into monthly reservoir simulations
Author(s) -
Stedinger Jery R.,
Pei Daniel,
Cohn Timothy A.
Publication year - 1985
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/wr021i005p00665
Subject(s) - streamflow , univariate , multivariate statistics , reliability (semiconductor) , environmental science , series (stratigraphy) , hydrology (agriculture) , statistics , econometrics , climatology , mathematics , geology , drainage basin , geography , geotechnical engineering , paleontology , power (physics) , physics , cartography , quantum mechanics
A condensed version of the Valencia‐Schaake disaggregation model is developed which describes the distribution of monthly streamflow sequences using a set of coupled univariate regression models rather than a multivariate time series formulation. The condensed model has fewer parameters and is convenient for generating flow sequences which incorporate the intrinsic variability of streamflows and the uncertainty in the parameters of the annual and monthly streamflow models. The impact of parameter uncertainty on derived relationships between reservoir capacity and reservoir performance statistics is illustrated using required reservoir capacity (calculated with the sequent peak algorithm), system reliability, and the average total shortfall. Modeled sequences describe flows in the Rappahannock River in Virginia and the Boise River in Idaho. For high‐reliability systems the results show that streamflow generation procedures which ignore model parameter uncertainty can grossly underestimate reservoir system failure rates and the severity of likely shortages, even if based on a 50‐year record.