Data-based mechanistic modeling of rainfall-runoff process, case study: upper Karoun subbasin data analysis

A major part of hydrological researches focused on complex and non-linear rainfall-runoff process. Mathematical models were presented to describe this process including a wide range from simple black-box representation to complex physically-based models. Considering inherent uncertainty associated with the process as a result of uncertain input variables and uncertain calibrated parameters, stochastic modeling seemed preferable to deterministic approaches. In this study, data-based mechanistic modeling (DBM) was selected to identify non-linearities of the process. The method is categorized as a stochastic approach relying upon recursive parameter estimation using Kalman filtering algorithm in state space system of equations. In addition, it is capable to reflect a physical interpretation of rainfall-runoff conversion to describe the behavior of the system. The later capability differs it from other black-box modeling approaches. In this research, a parallel structure of flow routes was identified in upper-Karoun subbasin of the great Karoun catchment. Sensitivity analysis was also carried out based on Monte Carlo simulation (MCS) method and the reliability of the presented model were quantified.