Geoscience Reference
In-Depth Information
HEAVY FLOOD DISCHARGE PREDICTION FOR 2004
FUKUI RAINFALL DISASTER IN JAPAN AND
PREDICTIONS IN UNGAUGED BASINS
Y. TACHIKAWA ∗,§ ,R.TAKUBO ,T.SAYAMA and K. TAKARA
Disaster Prevention Research Institute, Kyoto University, Japan 611-0011
§ tachikawa@flood.dpri.kyoto-u.ac.jp
Department of Urban and Environmental Engineering
Kyoto University, Japan 605-8501
On July 18, 2004, the largest-ever flood since the hydrologic observation began
at a catchment occurred at the Asuwa River basin in Fukui Prefecture, Japan.
The severe rainfall front in the middle of July brought heavy rainfall with
265 mm in 6 hs. The city area of Fukui was heavily inundated due to dyke
breaks along the Asuwa River; and the upper parts of the Asuwa River basin
were severely damaged by flood and sedimentation disasters. When designing
flood control planning in Japan, to obtain a design flood estimated by using
a rainfall-runoff model with a design rainfall having some exceedance proba-
bility is fundamental. While the accumulations of hydrologic data to estimate
a design flood is insu cient especially in small scale basins. In particular,
information of large floods near or above the magnitude of a design level is
less available. In this study, we examine how well/bad the largest-ever 2004
flood in Fukui is predicted using a state of the art physically-based distributed
rainfall-runoff model; then discuss the source of flood prediction uncertainty
and a direction to reduce the uncertainty and enhance the reliability of flood
discharge predictions.
1. Introduction
In 2004, severe rainfall fronts and 10 typhoons hit Japan caused heavy rain-
fall disasters with 232 casualties. These rainfall disasters mainly occurred at
tributary catchments with several hundreds square km. River managements
of these river catchments are organized by prefectural governments; and in
most situations, the river improvements still do not attain a designed safety
level. In the future it is still not easy to achieve high river improvements.
In these catchments, to assess the safety level of river improvements at
present time is the basis to design a future river development program. Also,
to develop a real-time flood runoff prediction system to issue a flood warning
is an urgent issue to save lives. To achieve these purposes, hydrologic pre-
diction by a reliable rainfall-runoff model is fundamental. While especially
89
Search WWH ::




Custom Search