Enhancing MIKE11 Updating Kernel and Evaluating Its Performance Using Numerical Experiments

A drainage basin is a flood-prone area in many parts of the world, especially in areas in which major urban development is underway, creating numerous social, economical, and psychological problems in both developed and developing countries. As a result, the use of flood forecasting and early warning systems is mandatory to reduce the economic losses and the risk incurred on people. The MIKE11 Flood Forecasting (FF) updating procedure can identify two different types of errors between measured and simulated data, namely amplitude and phase error, ignoring the nonuniform amplitude error around the flood peak region tentatively called shape error. In this paper, after making a clear distinction of such terms as simulation with and without an updating kernel (UK) and coupled versus decoupled calibration, the conventional UK incorporated into the MIKE11 FF hydrodynamic model is synthesized in some detail regarding the nature of amplitude and phase errors. In light of the nonuniform nature of amplitude error around the peak region, a three-parameter UK either coupled or decoupled is proposed for possible incorporation into MIKE11 FF. A highly controlled numerical experiment is devised to assess and evaluate the effectiveness of the proposed methodology compared with the conventional two-parameter UK implemented in MIKE11. Altogether, four different UKs are constructed and the forecasting results are compared and contrasted with plain simulation without any updating. Computed performance error indices clearly indicate the superiority of three-parameter UK to the two-parameter one. In summary, the proposed three-parameter UK could be regarded as a suitable alternative for adaptive forecasting of flood in drainage basins using MIKE11 or any other hydraulic and/or hydrologic simulation model.