Abstract
The largest-ever flood event was recorded at the Kumano River in the Kii peninsula of Japan from August 31 to September 4, 2011. Water level data exceeded the observation range in many places except in the Ouga water stage station at downstream. This study aimed to estimate the peak discharge of the event by using the estimation method, which is a 2D dynamic wave model combined with particle filters, and by considering the water level observed at the Hitari and Ouga stations and a river discharge estimated by a hydrological model. The estimation method was applied to the three historic flood events for quantifying the uncertainties of the Manning roughness coefficient(Manning n) of the subject river channel. Based on the optimized range of Manning n, the filtering method was applied to estimate the largest peak discharge of the 2011 flood using the discharge estimated by a hydrological model. The possible range of the largest peak discharge was successfully evaluated through the comparison of the observed flood marks. Finally, a rating curve established by the estimation results at the Ouga station is examined.
