Models of the Near-Surface Distortions - Models and Methods of Magnetotellurics - page 234

01 d the entire central seg-
ment is embraced with strong S -effect (no visible current leakage from the upper
layer). With widening the central segment, the S -effect weakens due to current
leakage through the upper layer bottom. At

calculations were done by (7.79). We see that at w < 0 .

5 d the S -effect in the
middle of the central segment virtually vanishes. It is a question of distances that
may number in the hundreds and thousands of km.
Let us consider the magnetotelluric and magnetovariational curves calculated for
the three-segment model using analytical solutions (7.75), (7.76) and with the finite
element method (Wannamaker et al.,1987).
Figsures 7.18 and 7.19 display the transverse and longitudinal apparent-
resistivity and impedance-phase curves obtained in the model with the resistive
central segment 20 km wide.
Once again we see that within the S 1 - and h -intervals the

w ∼ 5 ÷ 7 .

⊥ -curves plotted from
the analytical and finite-element solutions agree fairly well. Ascending branches
of the ⊥ -curves are not distorted. They coincide with ascending branches of the

locally normal ˙ ⊥ -curves are dis-

torted by the S -effect. Over the side segments they are shifted down reflecting the
current leakage (the current penetrates into the intermediate layer and flows under
the resistive central segment). These distortions are noticeable even at great dis-
tances from the central segment ( y

n -curves. However, descending branches of the

=− , = | | − v =

1010 km y y 1000 km,

adjustment distance d

≈

1000 km). When coming closer to the central segment,
the distortion increases. In the immediate neighborhood of the central segment
( y ⊥ =− 11 km , y = 1 km), the transverse apparent resistivity is 4 times less

⊥ -curves elongate
and their descending branches shift up reflecting the drop in S 1 .Here

than ˙ n . Over the central segment, the ascending branches of the

⊥

is 2000

times greater than ˙ n .

It is quite another matter with the -curves. Over the side segments the lon-

-curves are slightly distorted. Induction effect caused by the resistive
insertion is noticeable only in the immediate neighborhood of the central segment
( y

gitudinal

-curve shifts to
the left and becomes less sloping. In going to the central segment, we observe a
remarkable strong distortion reflecting the inductive influence of excess currents
concentrated at both sides of the resistive central segment. Here the

=− 11 km , y = 1 km). Here the ascending branch of the

-curves
acquire a well-defined false minimum that could be mistaken as an evidence of a
deep conductive layer underlying the central segment. This effect is known as the
effect of false conductive layer .
Now we turn to the phase curves. Note that the descending branches of the
transverse ⊥ -curves plotted from the analytical and finite-element solutions merge

together. Over the side segments the ascending and descending branches of the
⊥ -curves come close to the normal ˙ n -curves (slight phase distortions). But in

⊥ -curves we observe a considerable discrep-

the region of the maximum of the

⊥ and ˙ n . At the same time the longitudinal -curves are almost

ancy between

n -curves. Only in the immediate

everywhere in close agreement with the normal ˙

neighborhood of the central segment ( -curve, whose

ascending branch is strongly shifted to the left. Somewhat different relations are

y = 1km) we have the