Two-Dimensional Hydrodynamic Modelling of Flood Inundation for a Part of the Mekong River with TELEMAC-2D

Aims: This paper presents a study on the development of a 2-dimensional (2D) hydrodynamic model based on TELEMAC-2D for the flood simulation of the river from Kratie to Kampong Cham in Cambodia, a part of the Mekong River. The motivation behind the research was to study the feasibility of TELEMAC-2D in flood forecasting, and specifically to determine its adequacy in flood simulations with a focus on the reduction in model run-time through parallelization.
Place and Duration of Study: DHI-NTU Centre, Nanyang Environment and Water Research Institute (NEWRI), Nanyang Technological University, between November 2013 and March 2014.
Methodology: We simulated an actual flood event which occurred between June to November in 2001 for the stretch of the Mekong River from Kratie to Kampong Cham and compared the model simulations with MODIS satellite Images for specific days in the pre-, peak- and post-flood period.
Results: It was found that during the peak-flood period, there was high percentage (> 90%) match between the simulation results and observation obtained from satellite images while the match was below 50% for the pre- and post- flood periods.
Conclusion: The 2D simulation results were consistent with observations from satellite imaging. The discrepancy at pre- and post-flood may be due to the fact that (i) the model takes into account only hydrodynamic processes of flows in the river and flood plain, it does not consider other hydrological processes such as infiltration or evaporation which may be important during the pre- and post- flood periods, and (ii) the resolution of MODIS satellite image at 500m x 500m may be too coarse and therefore not sufficient to identify flooded areas when the area is small or water depth low. Finally, it was found that the computing time can be reduced significantly with parallelization using multi-core processors, albeit with lesser advantage in speedup when the number of cores increased beyond 4.