Geoscience Reference
In-Depth Information
The model is based on a synthesis of all the best functions from many other
existing methods consisting of the Genetic Algorithm (GA), simplex method and
controlled random search of Nelder and Mead ( 1965 ). This method is also capable
for non-linear optimization problems. This method has been found to be robust,
effective and efficient ( Duan et al. 2003 ). There are two objective functions which
can be used in the model calibration using SCE-UA. They are (1) the sum of the
squares of the residuals (SSQ) and (2) the sum of the squares of the difference of
the measured and simulated values after ranking (SQQR). In this study the SSQ
objective function is used. The SSQ, used to target at matching the simulated with
the observed data, is expressed as follows:
SSQ D X
i D 1;n
.x i; sim / 2
Œ
.x i; obs / TF
TF
(17.1)
where, n is the number of pairs of observed and simulated variable and 'TF' is a
user defined transformation function. Detailed description of ParaSol method can
be found in van Griensven and Meixner ( 2004 ).
17.2.2.4
Model Results
Using the above methodology, the SWAT model performance is established.
A period of five years from 2000 to 2005 was used to calibrate the SWAT model
for stream flow with the first year 2000 as a warm up period. The validation for the
model was done for another period 1996-2000 to check if the calibrated model holds
good for a different period. Station discharge data from the station Kon Tum (on the
Dakbla river) and the station Trung Nghia (on the Poko river) were used to calibrate
the model separately. The precipitation data on a daily scale were used from 1996-
2005 from the stations inside and outside catchment (Fig. 17.4 ) for the calibration
and validation processes of the SWAT model. The coefficient of determination (R 2 /
and the Nash-Sutcliffe Index (NE) were used as statistical indices to assess the
goodness of fit of the model.
The NE index is defined by:
P
i D 1 .O i S i / 2
P
NE D 1
(17.2)
2
O i O
i D 1
and the coefficient of determination R 2 is defined by
 
Search WWH ::




Custom Search