Information Technology Reference
In-Depth Information
ln
2
dT
T
T
max
A
(
x
˜
,
y
,
ln
T
)
R
A
(
x
˜
,
y
,
T
)
−
A
(
x
,
y
,
T
)
=
dydx
.
(10
.
107)
A
(
x
,
y
,
ln
T
)
X
Y
T
min
These misfits are readily written in the discrete form. As an example, consider
the apparent-resistivity misfit
ln
2
K
L
1
2
A
(
x
k
,
˜
y
l
,
ln
T
1
)
T
2
T
1
R
A
(
x
˜
,
y
,
T
)
−
A
(
x
,
y
,
T
)
=
ln
A
(
x
k
,
y
l
,
ln
T
1
)
k
=
1
l
=
1
ln
2
K
L
M
−
1
1
2
˜
A
(
x
k
,
y
l
,
ln
T
m
)
T
m
+
1
T
m
−
1
+
ln
A
(
x
k
,
y
l
,
ln
T
m
)
k
=
l
=
m
=
1
1
2
ln
2
K
L
1
2
A
(
x
k
,
˜
y
l
,
ln
T
M
)
T
M
T
M
−
1
,
(10
+
ln
A
(
x
k
,
y
l
,
ln
T
M
)
k
=
1
l
=
1
.
108)
where
k
T
max
.
It is self-evident that since the misfits of magnetotelluric and magnetovaria-
tional response functions are expressed in the metric of the function space
R
,the
inaccuracies
∈
[1
,
K
]
,
l
∈
[1
,
L
]
,
m
∈
[2
,
M
−
1]
,
T
1
=
T
min
,
T
M
=
of the initial data (measurement and model errors) must be expressed
in the same metric. Take, for instance, the apparent-resistivity misfit specified in
the space
R
by (10.107), (10.108). We will compare this misfit with an error
of
the same structure. Following (10.102), we determine
by relative deviations of
apparent resistivities:
2
dT
T
T
max
A
(
x
,
y
,
ln
T
)
2
=
dydx
A
(
x
,
y
,
ln
T
)
X
Y
T
min
A
(
x
k
,
2
K
L
1
2
y
l
,
ln
T
1
)
T
2
T
1
=
ln
A
(
x
k
,
y
l
,
ln
T
1
)
(10
.
109)
k
=
1
l
=
1
2
K
L
M
−
1
1
2
A
(
x
k
,
y
l
,
ln
T
m
)
T
m
+
1
T
m
−
1
+
ln
A
(
x
k
,
y
l
,
ln
T
m
)
k
=
1
l
=
1
m
=
2
2
K
L
1
2
A
(
x
k
,
y
l
,
ln
T
M
)
T
M
T
M
−
1
,
+
ln
A
(
x
k
,
y
l
,
ln
T
M
)
k
=
1
l
=
1
where
k
T
max
.
Next we turn to comparison of geoelectric media. These criteria must stress the
influence of those characteristics of the medium that reflect best of all the objective
structures and contribute most significantly to the response functions. In this respect,
the electrical conductivity is at premium. Magnetotellurics has enhanced sensitivity
∈
[1
,
K
]
,
l
∈
[1
,
L
]
,
m
∈
[2
,
M
−
1]
,
T
1
=
T
min
,
T
M
=
