Geoscience Reference
In-Depth Information
characteristics in the rainfall and discharge data to form three groups.
One of clear difference is the number of rainfall observatories. All floods
in group 3 were observed with more than 10 rainfall stations. It is inferred
that the accuracy of rainfall observations affects the value of the tuned
model parameters and makes large prediction uncertainty.
Another possible reason that makes the large prediction uncertainty is
the setting of initial condition for prediction simulations. Groups 1 and 2
have a flood observed with 10 rainfall stations. The difference among the
floods with more than 10 rainfall observations is the initial discharge.
The 1993 flood has the smallest initial discharge; while the 1985 flood has
the largest initial discharge in these floods. For the 1985 flood, the recession
of river discharge is clearly observed. This implies it was not correct for the
1985 flood to assume the steady state condition at the beginning of simula-
tion. Rainfall is spatially distributed and the distributions are memorized
in the spatial distribution of soil moisture, therefore if the initial condition
setting is inappropriate the resultant model parameter values are obtained
wrongly.
For the 1993 flood, a sudden rising up of river discharge followed dry
condition. To simulate the flood with the runoff model used here, the value
of parameter β is needed to be set in a large value to keep water in soil layer
for lasting small discharge at the outlet. If the low flow observation in 1993
is correct, the improvement of model structure including the refinement of
the discharge stage relationship and the initial condition setting are the key
to improve the flood runoff prediction for the rainfall-runoff model.
References
1. M. Shiiba, Y. Ichikawa, T. Sakakibara and Y. Tachikawa, J. Hydraulic Coastal
and Environmental Engineering JSCE
/II-47 (1999) 1-9.
2. Y. Ichikawa, M. Murakami, Y. Tachikawa and M. Shiiba, J. Hydraulic Coastal
and Environmental Engineering ,JSCE 691 /II-57 (2001) 43-52.
3. Y. Tachikawa, G. Nagatani and K. Takara, Annual Journal of Hydraulic
Engineering JSCE 48 (2004) 7-12.
621
Search WWH ::




Custom Search