Conclusion
In this paper, a hybrid model is proposed for flood prediction and tested in the case study area. The model contains three mathematical concepts, namely, Gumbel distribution function, drainage density and Muskingum method. Each of these concepts has their own characteristic and represents a difference view to water data. Each concept was calibrated to optimize parameters individually and together using past water data. The evaluation shows that the hybrid model givesfairly accurate prediction. The model could be suitable for real time flood prediction system due to its less complexity and small amount of data required.
One of the possible future works would be to investigate the correlation between the percent chance of flood and the damage level.