ISSN:
1573-1650
Keywords:
Deficient annual inflows
;
univariate model selection
;
bootstrap
;
confidence intervals
Source:
Springer Online Journal Archives 1860-2000
Topics:
Architecture, Civil Engineering, Surveying
,
Geography
Notes:
Abstract The paper considers the problem of univariate model selection in order to assess the risk of sequences of deficient annual inflow sums to a reservoir. A selection criterion is proposed which emphasises the fit of a model to the lower tail of the empirical distribution function. The expectation of the discrepancy between the operating and approximating models is estimated using the Bootstrap algorithm. The Bootstrap is also used to estimate confidence intervals about an estimated percentile and these are compared to those found using conventional asymptotic estimators. Applications are given for three South African reservoirs.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1007/BF00424657
