Developing Flood Inundation Map Using RRI and SOBEK Models: A Case Study of the Bago River Basin, Myanmar

抄録

The Bago River Basin in Myanmar is highly flood-prone. To develop a flood forecasting system, an inundation map of the Bago River Basin is required. This study applied the Rainfall-Runoff-Inundation (RRI) model and SOBEK model to simulate flood discharges and inundation to determine the model most suitable for analysis of the study basin in terms of user friendliness, cost, type of output, and correlation between simulated and observed data. In this study, five flood events were selected to calibrate and validate the models, using discharge data measured at Bago station. The Nash–Sutcliffe efficiency (E_NS) and coefficient of determination (R²) were used to evaluate the performance of the models. The simulated flood inundation area was validated with satellite images. According to the comparison, the SOBEK model is more accurate than the RRI model, and the simulated and observed discharges are closely related. However, when the calculation time and cost are included in the consideration, the RRI model is preferable, as it is faster and freely available. For the Bago River Basin, the RRI model is efficient in predicting the potential flood duration and areas of inundation in near-real time, whereas the SOBEK model is useful for floodplain management. This study shows that the RRI and SOBEK models are applicable to any basin in Myanmar that is similar to the Bago River Basin.