Investigating uniqueness and identifiability in auto-calibration of the ARNO daily rainfall-runoff model using the PSO algorithm

ABSTRACT

The so-called problem of identifiability and uniqueness of both the parameter set and the simulated hydrograph (as the output of the rainfall-runoff model), while applying an automatic calibration approach was discussed in many studies in the past decades. In this study, the ARNO daily conceptual rainfall-runoff model was auto-calibrated using the PSO algorithm. It was discussed that by modifying the structure of the model, it is possible to obtain all optimum parameter sets in the search space. It was also shown that although there is not a unique parameter set to describe the rainfall-runoff process, the simulated hydrographs of all optimum parameter sets converge almost to a unique solution. This implies that by modifying the structure of the model, even applying a blind search using a meta-heuristic algorithm (such as PSO) is sufficient to converge to the best and also unique simulation hydrograph almost precisely.

KEYWORDS:

• PSO
• identifiability
• rainfall-runoff model
• auto-calibration

Disclosure statement

No potential conflict of interest was reported by the author(s).