Geology Reference
In-Depth Information
Table 6.5 Comparison of some basic performance indices of different models employed in the study for daily rainfall-runoff modelling at the Brue catchment
Models
used
Training data(1056 data POINTS)
Validation data
RMSE
(m
3
/s
a
and
%)
CORR
Slope
MBE
(m
3
/s)
MAPE
(m
3
/s)
E
S
d
(m
3
/
s)
2
RMSE *
(m
3
/s and
%)
CORR
Slope
MBE
(m
3
/s)
MAPE
(m
3
/s)
E
S
d
(m
3
/
s)
2
Na
ï
ve
2.03
(101.6)
2.70
(112.5)
0.51
0.81
0.078
0.432
0.471
4.11
0.39
0.78
−
0.20
0.299
0.313
7.30
Trend
3.33
(155.6)
4.24
(176.8)
0.39
0.92
0.22
0.379
-0.265
9.83
0.30
0.90
0.15
0.494
-0.693
17.96
LLR
0.414
20.7 %
0.922
(37.7)
0.92
0.93
.028
0.0723
0.923
0.152
0.70
0.80
0.171
0.271
0.72
0.301
−
−
NNARX
0.558
(27.9)
0.877
(36.4)
0.83
0.90
−
0.007
0.335
0.843
0.468
0.68
0.80
−
0.085
0.297
0.717
1.152
ANFIS
0.470
(23.4)
0.796
(33.2)
0.88
0.93
−
0.038
0.276
0.889
0.331
0.75
0.85
−
0.039
0.253
0.773
0.635
SVM
0.415
(20.75)
0.692
(28.3)
0.89
0.91
−
0.062
0.790
0.91
0.223
0.79
0.82
−
0.012
0.199
0.831
0.552
W-SVM
0.370
(18.8)
0.670
(27.2)
0.90
0.90
0.051
0.097
0.90
0.137
0.75
0.83
0.112
0.167
0.770
0.497
−
−
W-
ANFIS
0.39
(19.5)
0.702
(28.6)
0.90
0.89
−
0.059
0.120
0.905
0.152
0.80
0.89
−
0.103
0.123
0.802
0.489
NW
0.274
(13.6)
0.699
(28.6)
0.96
0.97
−
0.00025
0.188
0.962
0.075
0.81
0.92
−
0.0068
0.205
0.805
0.490
a
Root Mean Square Error is also shown in percentage of the mean value of observed runoff