Geuphys. J . K. astr. Soc.

( 1984) 78,439 -- 4 5 2

Audiomagnetotelluric measurements across the Lake

Ladoga-Bothnian Bay zone in Central Finland

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

P. Karkkonen and K. Pajunpaa Departrnerit uf'GeuphJsics, 1/,ziiJersit,v
of O u h , SF-90570 Oulu 5 7 , Finland

Reccivcd 1984 January 3 ; in original form 1 9 8 3 January 5

Summary. A group of 6 4 stations among 140 AMT tneasurenlent sites in

Central Finland forms a profile about 220 km long. The profile starts from
the Svecokarelidic orogeiiic belt (1 800- 1900 Ma) and traverses the Karelidic
schist belt to the Presvecokarelidic basement (2600-2800 Ma).
The data obtained are analysed and are presented as apparent resistivity
profiles, pseudosections and residual profiles a i discrete frequcncies between
8 and 3700 Hz. The distribution of the measured apparent resistivities in the
research area shows that there are observable differences between different
geological units. A I - D interpretation is carried out at every station, after
which a 2-D model is constructed for the whole profile and its AMT
response is calculated numerically with the finite element technique.

1 Introduction
Audiomagnetotelluric (AMT) measurements have been carried out during the summer of
1980 in Central Finland. These AMT measurements are part of a larger electromagnetic (EM)
project started at the beginning of 1980 and includes research of the deep electric structure
of the Baltic Shield.
The main aim of these AMT measurements, when they were started, was to serve as
preliminary measurements for longer-period magnetotelluric (MT) soundings, to guide the
choice of MT stations with minimal near-surface electric distortion effects and t o help in the
interpretation of the deeper MT soundings (see Addm er al. 1982).
The present paper describes and analyses mainly the 6 4 AMT measurements among a
total of 140, which constitute a profile whose length is about 220 kni.

2 Geology
The Baltic Shield is the broadest shield area on the European continent. It is an uplifted part
of the wide basement area on which thc sediments of the North and East European platform
have been deposited (Simonen 1980). In the east and south the surface o f t h e Baltic Shield
dips gently under the sedimentary cover of the East European platform.
 440 P. Kaikkoneri arid K . Pajunpaa
2v 25O
I I

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

LEGEND
Granite
Granodiorite and quartz diorite
n Gabbro. anorthosite and peridotite
Serpentinite and smpstone
c]Metabasalt greenstone and ampfibolite
Fyllite. mica schist and mica gneiss
Quartz-feldspar schist and gneiss
Quartzite
n Granodioritic basement gneiss

Figure 1. (a) Schematic map showing the location of the AMT profile A H on the Baltic Shield.
I , Svecokarelidic orogenic belt. 11, Karelidic schist belt. 111, I'resvecokarelidic basement. (b) Detailed
geological map showing thc AMT profile AB.
 AuLlioMagnetoteliuric nieasurenzents in Central Finland 44 1
The AMT profile (AB) starts from the Svecokarelidic orogenic belt (1800-1900 Ma) in
the SW and traverses the Karelidic schist belt to the Presvecokarelidic basement (2000--
2800 Ma) in the N E (Fig. la). The zone running between Lake Ladoga and the Bothnian
Bay (the Lake Ladoga-Bothnian Bay rone) is coniplex area marked by geological processes
of different ages. If we consider it a s a whole, i t can be regarded as representing a radioactive
age limit and as a boundary between extensive blocks. It constitutes a milieu referred
to as the Savo schist situated between the Svecofennidic and Karelidic schist belts
(Talvitie 1977). The iiiajority of the known sulphide ores in Finland are situated in the Lake
Ladoga-Bothnian Bay zone on the northern side of's deep minimuni o f the gravimctric niap

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

(Mikkola & Niini 1968; Kahrna 1973). Fig. I (b) presents a more detailed geological map of
the region investigated by the AMT measurements.

3 Instrumentation and field procedure

A French scalar audiomagnetotelluric equipment (ECA 541 -0) manufactured by Societe
Eca, Paris, and originally developed at the Centre de Recherches Geophysiques, Garchy
(Benderitter, Ngock Thach & Jolivet I 973), was used for these measurements.
The amplitudes of the variations of the telluric and magnetic fields were measured at
nine fixed frequencies: 8, 17, 37, 80, 170, 370, 800, 1700 and 3700 Hz. Two induction
coils were used, one for the lower (from 8 t o 3 7 0 Hz) and one for the higher (from
170 to 3700 Hz) frequency band. The telluric field was measured galvanically with steel
electrodes separated by 15.8 m only because of the high telluric energy level. Otherwise it
would have been more appropriate, in this kind of regional investigation, to use a longer
telluric line with its greater smoothing etfect.
The scalar value of the apparent resistivity ( p a ) at each ftequency is read directly from
the scale of the resistivity meter. The final p a value at each discrete frequency at every
measuring point was determined as a geometric mean of five successive readings. The scatter
between these successive individual pa values can vary considerably. Generally it was around
10 per cent, depending, e.g. on geological conditions. For example, it was generally larger
at the more resistive sites. The width of the scatter also varied with the frequency, being
dependent on the behaviour of the source and on the noise field during the measurement
at a particular frequency.
The measurement directions chosen for the telluric line and the magnetic coil, and vice
versa, were 317" and 227" respectively. The former is the approximate direction of the
Lake Ladoga-Bothnian Bay zone and the latter is perpendicular to the zone. In addition
227" is also the direction of the AMT profile. In the following discussion the direction of
3 17" therefore refers regionally t o the L'-polarization (the electric field then being parallel
to the general strike of the Lake Ladoga-Bothnian Bay zone) and 227" refers to the
H-polarization (the E-field being parallel to the profile).

4 Results and discussion

Fig. 2 shows apparent resistivity pseudosections for the AMT profile. Pseudosections allow
' the simultaneous investigation of lateral and vertical variations of the apparent resistivity
(KoLiar & Strangway 1978). The ordinate axis (logf') presents the nine discrete lrequencies
used starting from 8 Hz a t the bottom to 3700 Hz at the top and the abscissa gives a distance
along the measured profile with all the stations. The values of l o g p , are contoured at
intervals of 0.2. There are clearly no uniform horirontal crustal layers. Large horizontal
gradients in log pa depict the existence 01' vertical geoelectric structures, e.g. contacts
 P
317 a
R

8:
Q:
227"

Figure 2. Apparent resistivity pseudosections for stations 1-64 along the profile AB. Directions of the telluric line are (a) 3 17" and (b) 227". Contour
values are logp, in a m and at intervals of 0.2.
 Audiomagneto telluric measuremerits in Central bi'nland 443
LOG W%

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

3

10 15 20 30 40 50 60 10 15 20 30 40 50 60

Figure 3. (a) Apparent resistivity profiles measured in the field with five discrete frequencies. (b) Apparent
resistivity profiles measured in the field but smoothed by using the three points moving average (log
mean).

between different kinds of rocks or geological formations, dykes, fractures, faults and
mineral, graphite o r schist zones, which all are typical for this kind of ancient shield area.
In Fig. 3(a) the measured apparent resistivity profiles at five frequencies are plotted for
both directions (3 17" and 227"). In order t o smooth these raw data and t o emphasize the
main features of t h e distribution of apparent resistivities, the siniple procedure of the three
points moving average was applied to the measured profiles (Fig. 3b). After this smoothing
the logp,(f') data show more clearly qualitative features of the profiles. At the lowest
frequencies the character of both pa profiles as a whole between about sites 1 5 and 32 is
distinctly different from the rest of the profile.
The profiles of residual apparent resistivity (see, e.g. Strangway & Koziar 1979) in Fig. 4
are constructed by normalizing the apparent resistivity values (ppp) at every individual
measuring point by the average regional (the whole profile) apparent resistivity value (p:)
for each discrete frequency and by taking the logarithm of that normalized value, i.e. the
 Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018
S
L
O
m

N
P. Kaikkonen and K. Pajunpaa

I
444
 Audionurgnetotelluric. measurements in Central Finland 445

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

24.

a
:
*
+
,6.

12-
m
Tclluric ltn 311'
- 7

8--5
6.

I--) , ' '

0, ' k? '
" " ' ' " T ' l l ' "'
1 1 1 1 ' , ' ;oL ' ;05 ' ' L
'
446 P. Kaikkotzerz and K. Pujurzpaa
ordinate is proportional to log ( p F p / p y ) . By doing so one can see which parts of the profile
are electrically different from the regional average value o f pa. To make it easier t o locate
t h e more conductive areas of tlie protile tlie negative values of the parameter plotted have
been shaded.
The niore conductive region (11) is an obvious feature of the whole profile. Comparing
t h e residual profiles of Fig. 4 with geology (Fig. I ) one can see that this region I1 corres-
ponds quite exactly with the Karelidic schist belt. Lithologically it correlates with the
change o f rock type and tectonically with the Lake Ladoga- Bothnian Bay zone.
Fig. 5 shows the apparent resistivity histogram for the whole AMT data including 140

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

sites. The AMT frequencies used have been grouped into three frequency bands: low (L),
medium ( M ) and high (21). or 8, 17, 37Hz: 80. 170, 370Hz, and 800, 1700 and
3700 Hz, respectively. The pavalues at every station are geometric mean values of the three
frequencies within each frequency band. The ordinate axis is proportional to tlie number of
stations ( N ) and to the percentage from all stations. On the abscissa axis, pa. each decade
has been divided into five intervals. Fig. 5 confirms that the AMT apparent resistivity values
for the Precambrian in Finland vary between a few times 10' and lo5 Rni. being most
often around lo4 to a few times lo4: but even pa values of several l o 5 Qin are fairly
coininon. This fact makes it possible, a t least theoretically. to penetrate the crust into the
uppermost part of the inantle using the lowest AMT frequcncies.
The whole AMT research area (the profile and its surrounding) has been divided into
three regions (I, 11, 111) o n the basis of Figs 1 and 4. In Fig. 6 the geometric mean values of
pa for these different geological regions along the profile (stations 1 - 64) have been plotted

Figure 6. Apparent resistivity values with standard deviations for three frequency bands (see text)
averaged separately for each of the three geological units: I . Svecokarclidic orogcnic belt (Central b'inland
granitoid block); 11. Karelidic schist belt (Lake Ladoga- Bothnian Bay zonc), and 111, Presvccokarelidic
bascnient (Karclian megablock). Symbols for apparent resistivity rangcs (standard dcviation): .--. region
I , - - - - region 11. - - rcgion 111.
 Audio magnetotelluric measurein en t s in Central b-itzlatid 447
for the three frequency bands L, M and H mentioned before. The standard deviations of pa
are also shown for these bands. Filially Fig. 6 shows that the AMT profile can be divided
into three electrically different geological regions: I . the resistive C:entrai Finland graditoid
complex; 111, the more resistive old archean basement (tlie Karelian megablock, Yanskin
1980). and 11, the conducting schist belt between blocks I and 111 (the Lake Ladoga--
Bothnian Bay zone).
In order to obtain a inore quantitative view of the geoelectric structure below the profile
one has to attempt a geophysical interpretation. In the first stage we have assumed that the
structure below evei-y measuring point is onc-diinensional ( 1 -D). A non-linear interpretation

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

with an interactive least-squares fitting and a hyperparabolic optimization (Pelkonen et al.
1979) has been carried out with the measured data at the points 1-64 along the profile.
Fig. 7 depicts the results of that I - D interpretation. The vertical axis is proportional to the
depth in kilometres and the liorimntal one gives the position of the measuring sites along
the profile. The layer parameters ~~ resistivity (log p ) and thickness
~ can be seen below
each measui-ing point. I f the S-value (conductance = conductivity times thickness) for the
first layer (overburden) is 0.1 S or greater it has been marked just above the surface of the
Earth. Otherwise there is no marked value. The numbers with the negative sign below
the surface stand for the S-value of the obtained thin layers. The layers with resistivities
less than 1000 Rm are hatched l.or the sake ofclarity.
The structure is not 1-0 as already seen earlier. The region between points 15 and 3 2 ,
i.e. region I1 being about 50 km broad, is distinctly more conductive than its surroundings
according to the interpretations in both polarizations. The N E side of region I1 seeiiis to
have conducting material fairly near t h e surface. The same thing can also be seen on the
higher frequency residual profiles, particularly in Fig. 4(a). This more conducting material is
going deeper towards the SW side of region 11. In addition here and there along the profile
there are some separate inore conducting ‘layers’ (dykes, fracture zones), e.g. points 7 and 8,
which are located o n the opposite shores of the fairly narrow lake, and points 50, 51
and 5 2 , which are on the Karelidic schist zone in Kainuu according to the map of the
surface geology.
Next, attempts have been n u d e to construct a 2-D geolectrical model for the profile,
using especially the results contained in Fig. 7. It is not possible to find a perfect model,
i.e. a geoelectric structure, which would generate a theoretical profile of apparent resistivities
identical to the measured ones; (1) at every discrete frequency used and ( 2 ) in both
polarizations. The 2-D study was therefore consciously restricted to the direction of 3 17”,
i.e. to the I:’-polat-iLation data, taking into account the assumed geological and tectonic
character of the research area and only the five lowest frequencies measured. For this kind
of regional study the higher frequencies are not very important as they most often reflect
only local effects.
Fig. 8 shows one of the constructed possible 2-D geoelectric models, in which tlie
resistivity varies between 100 and 200000 S 2 m (the seven different p values are shown
in Fig. 8). The finite element technique (Kaikkonen 1980) has been used in modelling.
The agreement between the measured and calculated apparent resistivity profiles is fairly
good in general, as can be seen in Fig. 9. There are some details (depending o n the frequency)
in the profiles, where consistency with the measured and calculated values is poor, particu-
larly at sites between 100 and 160 km. One explanation may lie in the insufficient depth
extent of the inhomogeneous parts (the zones with a symbol different from 7 , which
stands for ‘the host rock’). The depth extent of 3 5 kin is smaller than the skin depth
F = ( 2 / w p ~ ) ”in~ the material labelled as 8 (about 80 kin at the lowest frequency of 8 HL).
The energy at the lowest frequencies (8. 17 and 37 Hz) can thererore penetrate material
448 P. Kaikkonen and K. Pajunpai

YI 3
19
3 1 %
9
9 71 3
as I 3
w ?I 0
il 3
m

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

.93 I
4 I S
m
3 .\\ y
1\ 3\\\\\\\\\\'\' \\\\\\\\\\\\\',\\\\\\\\\\\,
r.j
\

I\ E ~\\\\\\";\\\\\'.'\\\~\\\\\\: \
.\\\\\\\\\\\\\'\\\
0
I, z \\\\\\\\\\\\\\\\\\'
3 14
 Audiomagnetotelluric measurements in Central Finland 449

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

S
h'%33,\\\\\\
3
c

R h8 A\\\\\\\\\'

a 19
B 1. a A\\\\\\
x 13
9 13

3
"5)l slu
3 13

?\\\\\\\Y\2
9 6 %
Y hiah\\\\\\\

13 I9
,. R3
3 19
I PI MA\\\\\\\\
I R
h'a\\\\\\\\\\
p bx\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\*

Y ?:: 12

i I S
3 kc\\\\\\\\\:
3 I:!
8 13
, 1 1
n
1 1 1 1
P
1 1 1 1 1
P
1 1 1 ~
R ~ , 1 , /
Pi8
, -
450 P,Kaikkonen and K. Pajunpaa
0 M I00 200 km

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

[kml
Figure 8. A possible 2-D geoelectric model along the measured AMT profile. The various numbers are
labels for the resistivities. 2 = 100, 3 = 5 0 0 , 4 = 1000, 5 = 5000,6 = 15 000, 7 = 20 000, 8 = 200 000 am.

. ..... -
I- -
Vi

J"
105 2EL .....

M 100 I50 200 km

0 M 150 260 krn

______ CALCULATED
- MEASURED

Figure 9. Calculated and measured resistivity profiles. The model is presented in I:ig. 8.

number 8 into material 7, which is 1 0 times more conducting than its cover. This yields
a decreasing effect on the apparent resistivity values calculated in the regions from 105 t o
120 km and from 185 to 200 k m . Removing the uppermost part of material number 4
between 135 and 1 4 0 km would increase the pa value at the highest frequencies and would
improve the agreement of the paprofiles in that region. These kinds of minor improvements
t o the geoelectric model could of course yield increasingly better agreement in detail
between the theoretical and field profiles.
The poor consistency in some parts of the profiles must be sought in the obvious three-
dimensionality of the geology in the area studied. This means that the field data used for
 Audiomagnetotelluric measurements in Central Finland 45 1
this comparison are actually more or less H-polarization data (Ting & Hohmann 1981)
instead of referring to the E-polarization. It is therefore quite likely that our data are biased
by 3-D effects and cannot be accounted for with a 2-D geoelectric model only.
It is obvious that the proposed model is too complicated taking into account the possible
resolution power of the measured data. Particularly the deeper - let us say deeper than
10 kin - structure of the model cannot be resolved so in detail as is proposed in Fig. 8.
However, the resolution is very likely sufficient for the shallower structure to resolve some
significant features of the model, e.g. zones around 100 and 180 km. These features have
real connections with geology being most likely the NE edge of the Lake Ladoga-Bothnian

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

Bay zone (and at the same time the SW edge of the very resistive old archean basement)
and the Karelidic schist zone surrounded by the archean basement, respectively.
Though it is difficult and almost impossible to justify the proposed 2-D geoelectric
structure geologically we can, however, find it to have typical non-horizontal features
concluded to be characteristic for the Precambrian crust (e.g. Zhanialetdinov 1976; Adim
e t a l . 1982).

5 Conclusions
On the basis of 140 AMT stations including the profile of 220 km long measured in
Central Finland it has been possible to distinguish three main geological units, electrically
different from each other. One of them (region 11) is distinctly more conductive than the
blocks on both sides of it and its boundaries can be located fairly precisely. Whether the
origin of this electrical ‘anomaly’ is due only to the change of the type of the rocks (which
is suggested by the map of surface geology), or to the Lake Ladoga-Bothnian Bay zone
(a fracture zone after Talvitie 1977) or is due to both of them (which is most likely) is
not entirely sure. The analysis and conclusions in this paper have been done mainly on
the basis of one profile, which is of course not sufficient to reach definitive conclusions
concerning the general geological and tectonic structure, and in particular to give a detailed
geoelectric structure of region I1 (the Lake Ladoga-Bothnian Bay zone).
Some additional AMT profiles have already been measured across the Lake Ladoga-
Bothnian Bay zone in order to complete and improve the ideas on this interesting and
important geological region. Analysis and interpretation of these new AMT data has been
started and will be reported in the future.

Acknowledgments
This study is a part of the project ‘The deep EM research’, which has been financed by the
Council of Natural Sciences of the Academy of Finland. We wish also to acknowledge the
useful suggestions and comments of the two referees.

References
Addm, A., Kaikkonen, P., Hjelt, S. E., Pajunpaa, K . , Szarka, L., Vero, J. & Wallner, A,, 1982. Magneto-
telluric and audiomagnetotelluric measurements in Finland, Tectonophys., 90, 77-90.
Benderitter, Y., Ngock Thach, H. & Jolivet, A., 1973. Magnetotelluric method for mining exploration,
paper read at the 3 5 t h EAEG meeting, Brighton, 1973 June, 22 pp.
Kahma, A., 1973. The main metallogenic features of Finland, Bull. geol. Surv. Finland, 265, 29 pp.
Kaikkonen, P., 1980. Numerical finite element modeling in geophysical applications of electromagnetic
fields, PhD thesis. Acta Univ. Oulu A . 93, Ph.ys.. 18, 135 pp.
Koziar, A. & Strangway, D. W., 1978. Shallow crustal sounding in the Superior Province by audio
frequency magnctotellurics, Can. J. Earth Sci., 15, 1701-1711.
452 P.Kaikkonen and K . Pajunpaa
Mikkola, A. & Niini, H., 1968. Structural position of the ore bearing areas in Finland, Bull. geol. SOC.
Finland, 40, 17-33.
Pclkoncn, R., Hjelt, S . E., Kaikkonen, P., Pernu, T. & Ruotsalainen, A., 1979. On the applicability of the
audiomagnctotelluric (AMT) method for ore prospecting in Finland, Contr. Dept Geophys.. Uniu.
Oulu, no. 94, 25 pp.
Simonen, A., 1980. The Precambrian in Finland, Bull. geol. Surv. Finland, 304, 5 8 pp.
Strangway, D. W. & Koziar, A , , 1979. Audio-frequency magnetotelluric sounding a case history at the
~

Cavendish geophysical test range, Geophysics, 44, 1429- 1446.

Talvitie, J., 1977. The Ladoga-Bothnian Bay fracture zone, in Fault Tectonics in the Eastern Part o f t h e
Baltic Shield, ed. Kortman, C., Proc. Fznnish-Soviet Symp., Finland. 1976 September 20-24.
pp. 113-120, The Committee for Scientific and Technical Co-operation between 1:inland and

Downloaded from https://academic.oup.com/gji/article-abstract/78/2/439/576850 by guest on 25 September 2018

the Soviet Union, Helsinki.
Ting, S. C. & Hohmann, G. W., 1981. Integral equation modeling of three-dimensional magnetotelluric
response, Geophysics, 46, 182-197.
Yanskin, A. L. (ed.), 1980. Tectonicsof Northern Eurasia, Nauka. Moscow.
Zhamaletdinov. A. A . , 1976. Electric conductive formations in NW of Kola-peninsula and their effect on
results of deep soundings, thesis, University of Leningrad, 21 pp. (in Russian).