Effects of Land Use Changes on Water Balance
__________________________________
1165
Table 3
. Mean monthly discharge values during the calibrat
ion and validation periods at Galinak and
Joestan Stations
.
Station
Period
Observed (m
3
s
-
1
)
Predicted (m
3
s
-
1
)
Deviation (%)
Joestan
Calibration
7.5
8.7
16.1
Validation
8.1
10.4
27.3
Galinak
Calibration
11.5
11.7
1.2
Validation
12.1
11.4
-5.3
Table 4.
Results of the statistical evaluation of model perf
ormance on the monthly discharge in the
calibration and validation periods at Joestan and G
alinak Stream Gauging Stations.
Gauging
station
Period
MARE
R
2
E
NS
Results
Joestan
Calibration
0.43
0.76
0.75
Acceptable
Validation
0.61
0.83
0.73
Acceptable
Galinak
Calibration
0.33
0.84
0.84
Good
Validation
0.34
0.90
0.89
Good
calibrated by SWAT during the model
calibration period for Galinak Station were
in good agreement. The observed discharge
was equal to 11.5 m
3
s
-1
compared with the
calibrated discharge of 11.7 m
3
s
-1
.
Similarly, for the validation period, the
above values were 12.1 and 11.4 m
3
s
-1
,
respectively. Notably, the comparative
evaluation of the average monthly discharge
values at Joestan Station showed relatively
good fit between the estimates during the
calibration and validation periods. As far as
the calibration period is concerned, the
average monthly observed and predicted
discharges assumed the values of 7.5 and 8.7
m
3
s
-1
, respectively. Similarly for the
validation period, the observed average
monthly discharge was 8.1 m
3
s
-1
, whereas
the predicted value was equal to 10.4 m
3
s
-1
.
Meanwhile, the average deviation of the
predicted discharges at Joestan Station from
the observed ones were 16.1 and 27.3% in
the calibration and validation stages,
respectively (Table 3).
The values of MARE calculated for the
two stations are generally low and close to
zero (Table 4). The
R
2
and
E
NS
coefficient
are two important statistical indicators for
evaluation of the results. In the case of
Joestan Station, the
R
2
values corresponding
to the relationships between the observed
and predicted average monthly discharges
were found to be 0.76 and 0.83 during the
calibration
and
validation
periods,
respectively. However, the corresponding
values for Galinak Station were 0.84 and
0.90. Therefore, all of results in both stations
and both periods (calibration and validation)
for mean monthly flow showed the goodness
fit of the simulation in the study area.
Therefore, in general, SWAT model was
reasonably capable to reproduce mean
monthly discharge in Taleghan area.
Consequently, based on statistical analysis,
the results show: (i) the model can predicate
the runoff accurately; (ii) the model is
suitable and recommended for the study
area.
Runoff Components
The runoff components encompass surface
runoff, lateral flow, and groundwater flow.
According to Linsley
et al.
(1949); Linsley
et al.
(1982), and Klemes (1986), drawing
distinction between the components of
runoff is arbitrary and the source(s) of the
water passing a gauging station cannot be
identified,
therefore,
comparing
the
predicted fluxes against any observation at
the two gauging stations within the Taleghan
basin is not possible. Figures 3 and 4 show
the monthly surface runoff, lateral flow, and