Premium
Calibration of a conceptual rainfall‐runoff model for flood frequency estimation by continuous simulation
Author(s) -
Lamb Robert
Publication year - 1999
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 - 10.1029/1999wr900119
Subject(s) - calibration , flood myth , flow (mathematics) , sensitivity (control systems) , continuous simulation , surface runoff , goodness of fit , probability distribution , estimation theory , frequency distribution , statistics , sampling (signal processing) , environmental science , mathematics , hydrology (agriculture) , computer science , simulation , geology , engineering , geography , geometry , ecology , archaeology , filter (signal processing) , electronic engineering , computer vision , biology , geotechnical engineering
An approach is described to the calibration of a conceptual rainfall‐runoff model, the Probability Distributed Model (PDM), for estimating flood frequencies at gauged sites by continuous flow simulation. A first step was the estimation of routing store parameters by recession curve analysis. Uniform random sampling was then used to search for parameter sets that produced simulations achieving the best fit to observed, hourly flow data over a 2‐year period. Goodness of fit was expressed in terms of four objective functions designed to give different degrees of weight to peaks in flow. Flood frequency results were improved, if necessary, by manual adjustment of parameters, with reference to peaks extracted from the entire hourly flow record. Although the primary aim was to reproduce observed peaks, consideration was also given to finding parameter sets capable of generating a realistic overall characterization of the flow regime. Examples are shown where the calibrated model generated simulations that reproduced well the magnitude and frequency distribution of peak flows. Factors affecting the acceptability of these simulations are discussed. For an example catchment, a sensitivity analysis shows that there may be more than one set of parameter values well suited to the simulation of peak flows.