A new modified nonlinear Muskingum model and its parameter estimation using the adaptive genetic algorithm

Abstract

First, a novel nonlinear Muskingum flood routing model with a variable exponent parameter and simultaneously considering the lateral flow along the river reach (named VEP-NLMM-L) was developed in this research. Then, an improved real-coded adaptive genetic algorithm (RAGA) with elite strategy was applied for precise parameter estimation of the proposed model. The problem was formulated as a mathematical optimization procedure to minimize the sum of the squared deviations (SSQ) between the observed and the estimated outflows. Finally, the VEP-NLMM-L was validated on three watersheds with different characteristics (Case 1 to 3). Comparisons of the optimal results for the three case studies by traditional Muskingum models and the VEP-NLMM-L show that the modified Muskingum model can produce the most accurate fit to outflow data. Application results in Case 3 also indicate that the VEP-NLMM-L may be suitable for solving river flood routing problems in both model calibration and prediction stages.