Sie sind auf Seite 1von 9

Downloaded 12/18/14 to 180.254.108.125. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.

org/

GEOPHYSICS, VOL. 57, NO. I (JANUARY 1992); P. 97-105, 12 FIGS.

Effect of temporal and spatial variations of the primary


signal on VLF total-field surveys

Marc A. Vallee*, Michel Chouteau+, and G. J. Palacky

Spatial and temporal variations have been studied using


field monitoring of the transmitted signal. The results of
fieldexperiments indicate that the nature of the received
VLF fields changes significantly even over moderate
distances (20-30 km) and that data cannot be reliably
corrected over larger distances. This observation has a
significant implication for VLF total-field surveys, particularly airborne, in which base stations have been
routinely used to monitor the primary field strength and
to correct the survey data. The results of primary signal
monitoring are also used to demonstrate the effect of
solar flares on VLF surveys. Because of the large
intensity and complex electromagnetic character of solar
flares, survey data recorded during such events cannot
be used for map compilation and interpretation.

ABSTRACT

Most of the airborne and ground VLF instruments


presently used measure the total-field response in
addition to field ratios. Results of surveys using these
instruments are adversely affected by spatial and temporal variations in the VLF primary field. Until now,
the nature of such variations has not been studied from
the point of view of geophysical surveying practice.
Spatial variations are analyzed using radio propagation
models. The most important result is the identification
of primary field minima where surveys would be
unreliable. Their dependence on the transmitter location is rather complex, and modeling should be carried
out before specifying VLF stations for a survey area.

geological mapping in nonconductive environments. As ratio


measurements are sensitive to orientation of magnetic sensors, which is unknown with some airborne platforms, some
airborne instruments evolved that measure the total-field
amplitude in addition to field ratios. Most VLF instruments
presently in operation in North America are of such type,
including Herz Totem, Sander VLF-EM II, and Scintrex
SE-90 (Collett, 1986, Herz, 1986). Some commercial ground
systems also have this option. Unfortunately, some of the
problems that plagued early radio field measurements have
still not been properly addressed.
An important consideration when using total-field VLF
data in geological mapping and exploration is how to compensate for changes in the primary field that are not related
to geology. A correction becomes possible only if the origin
of the primary field variations is fully understood. Although
the nature of the VLF primary field has been studied in depth
by radio engineers and several studies have been written on
the subject (Watt, 1967), not enough attention has been paid

INTRODUCTION

Radio field intensity measurements were among the first


electromagnetic (EM) methods to be considered for possible
use in mineral exploration (Cloos, 1934). However, early
tests indicated that field intensities measured during surveys
were influenced by a number of factors not related to
geology, such as fluctuation of the transmitting power,
interference patterns between ground and sky waves
(Hollingworth, 1926), topography (Howell, 1943), and solar
flares (Dellinger, 1937). For this reason, attempts to use
radio transmissions as a source of the primary EM field
focused on measurements of the tilt angle or the ratio of
orthogonal magnetic fields, which are insensitive to the
primary field variations (Paterson and Ronka, 1971).
With worldwide availability of VLF signals in the 15 to 25
kHz range, which are emitted for communications with
submarines, VLF field ratio measurements have become a
well-established geophysical technique used primarily for

Manuscript received by the Editor September 11, 1990; revised manuscript received June 5, 1991.
*Formerly Departernent de Genie Mineral, Ecole Polytechnique; presently Centre de Technologie Noranda, 240 Boulevard Hymus,
Pointe-Claire, Quebec, Canada H9R IG5.
:j:Departementde Genie Mineral, Ecole Polytechnique, C.P. 6079, Succ. A, Montreal, Quebec, Canada H3C 3A7.
Geological Survey of Canada, Mineral Resources Division, 601 Booth Street, Ottawa, Ontario, Canada KIA OE8.
1992 Society of Exploration Geophysicists. All rights reserved.
97

Downloaded 12/18/14 to 180.254.108.125. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

98

Vallee et a!.

to the problem by the geophysical community carrying out


and interpretating VLF surveys.
From the perspective of exploration geophysics, both
temporal and spatial variations of the primary VLF field are
of great importance. In principle, temporal variations can be
recorded with a fixed receiver (base station) and survey data
subsequently corrected. This approach has become routine
in airborne and ground VLF total-field surveys. However,
very few practitioners realize that the use of a base station
may be ineffective in many situations. As demonstrated in
this study, temporal variations cannot be correlated over a
large distance. The problem of spatial variations is even less
understood by field geophysicists. In regional airborne VLF
surveys flown by the Geological Survey of Canada, significant variations in the intensity of the primary total field have
been observed. This phenomenon was believed by Dr. R. L.
Grasty (1990, Pers. Comm.) to be due to interference between the ground and reflected sky waves.
In this paper, we review the temporal and spatial variations affecting total-field surveys. We present a classification
of temporal variations illustrated with an example. We then
advocate the use of radio propagation modeling to predict
the intensity of the VLF primary field and, in particular,
minima in the interference patterns. Near these minima, the
primary field varies rapidly with distance. Their location
depends on the phase relation between ground and sky
waves. Near these minima, the field intensity is also highly
sensitive to the variations of the ionosphere, and temporal
variations observed with distant receivers may not be well
related. This study is supported by the results of an experiment on correlation of temporal variations with distants
receivers. This experiment shows limitations in the use of a
base station for correction of temporal variations.

- fluctuations and interruption of the transmitter


power,
- sunrise and sunset fading,
- slow drift during the day, and
- rapid fluctuations during the night.
Natural VLF signal from atmospheric noise (sferics) and
whistlers also causes temporal variations of the VLF field,
but it mainly appears as high-frequency noise on a VLF
diurnal record. The magnitude of this natural VLF perturbation can be estimated from world maps of signal-to-noise
ratios prepared for selected transmitter locations and seasons (Watt, 1967, Hauser and Rhoads, 1974).
Temporal variations in the primary field produce a lowfrequency drift of the VLF signal. According to their origin,
these variations can be separated into two groups: (a)
transmitter power variations (man-made, and hence, in
principle, avoidable), (b) changes which depend on the
transmitter-receiver geometry and on the physical properties
of the propagating medium. At a given location, significant

a)
15,-

-r

NAA(Cutler)

E
<,
>

...,>.....

..

10

UI
C

...,
.....c

u
....

.....
....

Interruption

.....

--l>-

...,c,
U
....
lJ.J
0
0

Diurnal variations that affect geophysical surveys can be


recorded with a fixed receiver (base station). To identify and
analyze these variations, an experiment was set up, in which
intensities of electric fields were continuously measured
during defined periods. In the spring of 1988, signals from
VLF transmitters code-named NAA located at Cutler,
Maine, and NSS at Annapolis, Maryland, were monitored at
Saint-Remi-de-Napierville, 30 km south of Montreal, Quebec. This facility, the Spectrum Control Centre, is maintained by the Department of Communications of the Canadian Federal Government.
A rhombic electric antenna was used as a receiver. It was
connected to an HP 8568B sweep spectral analyzer, which
was controlled by an HP 9122 microcomputer. An electric
field intensity spectrum was swept over a second with a
resolution bandwidth of 300 Hz. Intensities at 21.4 (NSS)
and 24 kHz (NAA) were averaged and recorded every 10 s
for periods of two to three days. Similar variations were
observed each day and only examples over a 24 hour period
are presented.
Figure 1 displays examples ofVLF diurnal variations over
a 24 hour period. The following types of temporal variations
have been identified:

Sunset

Sunrise

Ql

TEMPORAL VARIATIONS

12

16

20

24

Eastern daylight-saving time (hours)

b)
15

NSS (Annapo 1is)

E
<,
>

.........>-

10

UI
C

Sunset

Sunrise

Ql

....

C
.....

....Ql

.....

....

.....u

...,c,
u
....
lJ.J

..

12

16

20

24

Eastern daylight-saving time (hours)

FIG. 1.. Diurnal variations recorded on April 23, 1989, for


transmitters NAA, Cutler and NSS, Annapolis (b). Times of
sunrise, sunset, and transmitting power interruption are
indicated.

99

Downloaded 12/18/14 to 180.254.108.125. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

VLF Primary Field Variation

variations in the character of signals originating from different transmitters will be observed due to the fact that
orientation and length of the propagation paths are different.
To explain variations of the second type, a brief outline is
given of VLF signal transmission. EM fields generated by an
electric antenna are proportional to the square root of the
transmitting power. In the VLF frequency range (15 to 25
kHz), waves propagate to the receiver location in an electric
waveguide formed by the earth's surface and the ionosphere.
The part of the ionosphere affecting the VLF propagation is
called the D-layer located at a height of 60 to 80 km above
the earth's surface. This region of the upper atmosphere is
ionized by solar radiation, especially Lyman-a and soft X-rays
(Davies, 1%6). Therefore, the propagation is strongly affected
by the presence of the sun over the propagation path. This
explains the large differences in intensities observed between
night and day and the rapid changes at sunrise and sunset.
Operators who would like to make the most of the day should
note that sunrise and sunset periods are not suitable for VLF
surveying. This has been noticed already by Thiel and Chant
(1982) for wavetilt measurements.
As field surveys are carried out during the day, they are
affected by slow drift and rapid oscillations in transmitter
power (Figure 1). The assumption is normally made that data
obtained at base stations can be used to correct field measurements. Our study shows that this technique can only be
applied in limited cases.
SPATIAL VARIATION MODEL

In routine VLF surveys, measurements are carried out


along lines in a given survey area. To predict accurately
primary field variations at the mobile receiver using base
station data, the patterns of change in the primary field
intensity with distance must be established. Models developed for radio propagation can be used in such studies.

The field intensity of the VLF signal transmitted by an


electric antenna can be calculated from a number of models.
Solutions for a conductive sphere embedded in anisotropic
ionosphere have been proposed since the beginning of the
20th century. An historical overview has been given by
Johler and Berry (1962). Some methods were compared by
Jones and Mowforth (1982). For our study, an approach
based on the summation of zonal harmonics has been chosen
(Johler, 1970).
The contributions of the ground wave and reflected sky
waves to the total field intensity are computed separately. A
cartoon depicting various contributions is shown in Figure 2.
The method assumes a uniformly conducting earth and a
layered ionosphere. Reflection coefficients for a horizontal
anisotropic ionosphere are computed using the Johler and
Harper (1962) algorithm. The ionization distribution of the
D-Iayeris represented by the Wait and Spies (1964) exponential
model. Reflectionof a radio wave by the ionosphere is affected
by the intensity and orientation of the earth's magnetic field.
These parameters are obtained through the IGRF 1985 model
(IAGA Division I, Working Group 1, 1986).
Total horizontal magnetic field intensity has been computed along a south-north propagation path for distances
from the transmitter of 200to 2000km, and parameters of the
earth's magnetic field that are typical of VLF surveys in
eastern North America. The results are presented for transmitters NAA and NSS in Figures 3a and 3b, respectively. In
the calculation, in which four sky hops were used, the
ground conductivity was assumed to be 2 mS/m and its
relative dielectric permittivity 20 (ITT, 1975).
In the same figure, the horizontal magnetic field intensities
contributed by the ground wave and the first sky hop are
separately depicted. At a distance of about 550 km, the
contribution of the first sky hop exceeds that of the ground
wave, which prevails near the transmitter. The contribution

EXPONENTIAL IONOSPHERE
FROM WAIT AND SPIES (1964)

TX-RX~

1;>.,.

NAA (Cutler)
POWER: 1 MW
FREQUENCY: 24 kHz
NSS (Annapolis)
POWER: 400 kW

GROUND CONDUCTIVITY : 2 mS/m


GROUND RELATIVE DIELECTRIC PERMITTIVITY: 20

FREQUENCY: 21.4 kHz

FIG. 2. Characteristics of the VLF radio propagation model. Values of the parameters depicted were used to
produce results of Figure 3.

Downloaded 12/18/14 to 180.254.108.125. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

100

Vallee et al.

of other sky hops is minor at the distances depicted and


hence has not been plotted. However, their effect becomes
significant at greater distances from the transmitter and
should not be neglected in calculations. Minima in the total
field were observed and attributed by Hollingworth (1926) to
interference patterns between the ground wave and the sky
waves. The minima locations are relatively insensitive to the
direction of propagation, but depend on the frequency of the
transmitter. (Notice the slight shift in minima between NAA
and NSS propagation graphs.)
Only the intensity of the magnetic field perpendicular to
the propagation path have been presented. However, the
model selected can also predict other components of the
magnetic field intensity. Intensities of these components, for
a vertical electric dipole source, depend on the conversion
between transverse electric polarization and transverse magnetic polarization of waves that are reflected by the ionosphere. Bracewell et al. (1951) observed that minima in
conversion coefficients occur during the day and in summer.

a)
E
<, 125,-

--.,.

NAA

c-,

.....

.~

---- Total field

100

----- Ground wave

Ql

.....
C
.,..,
"tJ

.....
.,..,Ql
.....

First sky hop


75

.~ 50

.....

Ql
C

01
(Q

.....
(Q
.....

25

.~

......

--.

..::..::..:=:r-

0 +----.---,.----.---,.----.---,.--.....
250
500
750
1000
1250
1500
1750
0

2000

Tx-Rx distance (km)

b)
E

'- 1 2 5 . , - - - - -

NSS

>.....

---- Total field


----- Ground wave
.. First sky hop

.;;; 100

Ql

.....

C
.,..,

"tJ

.....

75

.,..,Ql
.....

.~

.....

50

Ql

01
(Q

.....
(Q
.....

25

.:::

0 +-_--.-__,.-"-_--.-__,.-_--.-__,.-_--.-'-=-"=+
500
750
1000
1250
1500
1750
2000
250
0

Tx-Ax Distance

~m)

FIG. 3. Predicted primary field variations for transmitters (a)


NAA and (b) NSS along a south-north propagation path.
Amplitudes of the ground wave, the first sky-wave contribution, and the complex sum of all contributions (ground
wave plus four sky hops) are presented.

They also decrease with frequency. For example, the ionospheric model selected in this paper does not produce
transverse magnetic propagation. However, variations of the
wavetilt observed at sunrise and sunset by Thiel and Chant
(1982) are attributed to mode conversions.
Near the minima, horizontal magnetic amplitudes vary
strongly with distance. As daytime variations of ionization
affect the intensity and phase of the sky-wave contributions,
they also affect the minima locations. Because of the strong
variations of the fields near the minima, studies of the
temporal and spatial variations of the primary field at these
locations are important.
SPATIAL VARIATIONS EXPERIMENTS

Local variations in the primary field strength will obviously affect the quality ofVLF surveys. The problem can be
studied theoretically or experimentally. Correlations between
VLF signal-phase variations with distance have been studied
by Pressey et al. (1961) and Sobczak and Taylor (1970) among
others. However, as the purpose of their investigation was a
better design of navigation systems, their conclusions which
focused on phase variations are not applicable to amplitude
measurements. To study the dependence of amplitude variations on distance, we opted for field observations followed by
correlation with theoretically obtained results.
In December 1989 and May 1990, an experiment was set
up south of Montreal, Quebec, in which two receivers measured variations in the primary VLF field during several
half-day periods. The location of the study area and VLF
transmitters used in the experiment is shown in Figure 4a. Also
indicated are the distances between the area and the transmitters. Commercial VLF instruments of two types were used to
record the magnetic field intensity continuously at locations
depicted in Figure 4b. In December 1989, two Scintrex VLF-3
units, which measure the horizontal magnetic field intensity,
were used. In May 1990, the VLF total field strength was
measured with two EDA Omniplus receivers. Measurements
were recorded every lOs for periods up to 6 hours long. With
each instrument, several receiver separations were used and
measurements were carried out over a period of several days.
The radio propagation model described above has been
used to predict spatial variations in the area. The total field
and the contribution of the ground wave and the first sky hop
are displayed for transmitter distances relevant to the survey
area. For the NAA transmitter (Figure 5a), the ground wave
contribution is more important than that of the first sky hop.
At this location, a decrease of 30 percent in the NAA
primary total-field intensity has been predicted for a distance
increase of 100 km (from 400 to 500 km). For the more
distant NSS transmitter (Figure 5b), the first sky hop contribution is more important, and the total field remains
relatively constant over the distance depicted.
Figure 6 displays diurnal variations observed with a Scintrex VLF-3 receiver on December 5,1989, at location B. The
instrument has a resolution of one unit, which corresponds
to 156 nA/m at 21.4 kHz (NSS) and 139 nAim at 24 kHz
(NAA). The signal varies relatively slowly during the 6.5
hour interval around a mean of 25 /LA/m.
Spectral analysis has been used to characterize the frequency content of the data. For short series, Thomson (1982)

101

VLF Primary Field Variation

a)

Downloaded 12/18/14 to 180.254.108.125. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

75

150 km
!

NSS
(Annapolis)

b)

Mont@
St-Gregoire

NAA
o

(Cutler)
- -.....
~ 455 km

Sle-Brigide

Iberville

~
B

St-Alexandre

NSS
(Annapolis)
750 km

Farnham

Ste-Sabine

o
!

Ferndon

5 km
!

FIG. 4. (a) Location of the survey area relative to NAA and NSS transmitters and (b) detailed map showing the
location of the receiver sites (A, B, C, D).

Downloaded 12/18/14 to 180.254.108.125. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

102

Vallee et al.

and Walden (1990) recommended multitaper spectral analysis because of higher resolution and low leakage of the
spheroidal windows. Figure 7 depicts the spectrum of the
series of Figure 6, that was estimated with a weighted
average of eigenspectrums computed with 41T prolate eigentapers. The energy maximum is concentrated near DC,
below 10 cycles/hour, which corresponds to a period of 6
minutes. At higher frequencies, the signal is characterized
by a constant noise level at -47 dB, which reflects the
instrument error of 140 nAlm at 24 kHz.
In all surveys, coherence of the transfer function relating
signals recorded at separated receivers is at low frequency,
where the energy is concentrated. However, in this frequency band, the gain and phase of the transfer function
varies with frequency. Figure 8 presents results that were
recorded on December 9, 1990.Two receivers located 18km

<t

Ul

c:

NAA

c-,

QJ

.....

......
..,

.c:
.., 20

- - Total field

Ul

75

~
rl

----- Ground wave

.....

.c:..,

.QJ
.., 15

....

First sky hop

.u..,

rl

.....

50

QJ

c:

'"

.,
.....

QJ

rl

c:

25

.N..,

.....
'"
L
0

L
0

c:

...,

c:

rl

0
N

10

Cl

.....

'"

NAA

......
.., 25

2-

Cl

30

>-

:t 100

.QJ
..,
....
.u..,

In addition to temporal variations identified in Figure 1,


there are sudden ionospheric disturbances, particularly
those associated with solar flares, that affect VLF measurements. Figure 10 presents measurements recorded on May
10, 1990, with two EDA receivers 27 km apart. At about

2-

QJ

SUDDEN IONOSPHERIC DISTURBANCES

E
<,

a)

c:

apart (at location A and C of Figure 4b) recorded the


intensity of the horizontal magnetic field. Most events can be
visually correlated on the two traces. However, the lowfrequency drift is different at the two receivers. In Figure 9
coherence, gain, and phase of the transfer function relating
signal D to signal A depicted in Figure 8 are compared with
estimates for signals received with zero separation. The
coherence deteriorates for distant receivers. Also the estimated gain and phase become noisier.

0
400

425

450

475

10

11

12

13

14

15

16

Eastern standard time (hours)

500

FIG. 6. NAA variations recorded at site B on December 5,


1989, with a Scintrex VLF-3 receiver, with readings once
every 10 s.

Tx-Rx distance (km)

b)
E
<,
<t

30

2c-,

......
.., 25
Ul

c:
QJ

.....
.c:
.., 20

--------

0-,--

l'lSS

<D -10
Eo
............

>~ -20
Ul

c:

~
rl

.QJ
..,
....
.u..,
.....

QJ

c:

QJ
~

'"
.....
'"
c:

15

10

u
~ -40

- - Total field
-----

Ul

Ground wave
First sky hop

QJ

3: -50

a.

noise level

.N..,
L
0

-30

::l
L

.....

Cl

rl

----.

0
700

725

750

Tx-Rx distance

775

800

(km)

FIG. 5. Predicted primary field intensity over the survey


area, for (a) NAA signal along an east-west line, and (b) NSS
signal along a south-north line.

-60 +-_-,-_--,,--_,-_-,--_--,_ _,-_-,--_--,_ _


40
60
80
100
120
140
160
180
20
o

Frequency (cycles/hour)

FIG. 7. Spectrum of field data depicted in Figure 6. The


spectrum was computed with 41T eigentapers. The instrument noise level is also indicated.

Downloaded 12/18/14 to 180.254.108.125. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

VLF Primary Field Variation


11:30,a significant decrease in the NAA signal was observed
at both receivers (A and D), which was followed by a
temporary return to the original value and an even sharper
drop at 15:30. Figure 11 compares the records of signals
received from the Cutler and Annapolis transmitters. At the
time of the drop in the NAA signal at 11 :30, the NSS signal
increased by 80 percent. Unfortunately, no NSS data were
available for the latter event because the Annapolis station
was off the air. These two events can be correlated to solar
flares of M3.9 and X3.4 X-Ray Class, respectively (National
Oceanic and Atmospheric Administration, 1990).
As mentioned in the Introduction, the effects of solar
flares on radio propagation were first observed by Dellinger
(1937). Davies (1966) explained the mechanism in the following way: X-ray emissions associated with solar flares increase the ionization of the D-layer. The effective height of
reflection of the VLF waves decreases for several kilometers
over a short period of time, followed by a slower recovery.
Variations in the reflection height modify, accordingly, the
intensity and phase of the sky hop contributions to the total
field intensities. As solar flares become more frequent during
years close to the peak of the l l-year sunspot cycle, sudden
ionospheric disturbances are expected to increase and reach
a maximum in 1992.
A comparison is made in Figure 12of spectrum and parameters of the transfer function for calm and disturbed ionosphere
(inthis case the transfer function relating signalD to signalA of
Figure 11).The power spectrum density increases by 10dB for
frequencies smaller than 5.0 cycles/hour. The coherence is
near one over this frequency band. However, the gain increases linearly from 0.6 to approximately 2 in the frequency
range (}-5 cycles/hour. Phase decreases linearly over the same
frequency band, which represents an advance of signalA over
signal D. This reflects the observation that the sudden amplitude variation started at location A before D.
Primary-field variations at a moving receiver can be predicted from the base station records only if the transfer
function relating the receivers is known. However, during

103

sudden ionospheric disturbances, the two parameters


change with frequency (or time) and with the receiver
locations. In such situations, it is impossible to use the
base-station data to correct for the variations observed at the
moving receiver. As in the case of magnetic storms and their
treatment in magnetic field surveys, VLF data acquired
during solar flares should not be used for interpretation or
map compilation.
a)

O.B

Q)

iJ

0.6

Q)

L
Q)

s:

00.4

Separation
0.2

o 0 km

1B km

0.0+-_ _- ,_ _--,,--_ _.---_ _---,_ _--,,--_ _


0.0

2.5

5.0

7.5

10.0

12.5

15.0

Frequency (cycles/hour)

b)
2.0,--_-----------------,

Separation
o

0 km

1B km

1.5

~ 1. 0 JPlIIlIlll!!,~..t::-:~~--;;!~Z'li!iii~rIif~~~lt?"'\l~__t
C!l

0.5

<,

30

....,>.....

25

NAA. 18 km separat ion

0.0

.....c 20

5.0

,-__---,__---,
7.5

10.0

12.5

+
15.0

c)
30r---.--------

u
....

.....

2.5

Frequence (eye les/hour)

....,Q)

Q)
.....

+-__-,__--,

0.0

UJ
C

---,.

15

.....
....,
Q)
c 10

Cl
ttl

~ -15

....ttl
....,

ttl

s:

a.
-30 + - - - - - - , - - - - - , - - - , - - - - , - - - - - - , - - - - t
0.0
2.5
5.0
7.5
10.0
12.5
15.0

C
0
N

.....
L

0
I

Frequency (cycles/hour)
9

10

11

12

13

14

15

Eastern standard time (hours)

FIG. 8. NAA intensity recorded with two Scintrex VLF-3


receivers located 18 km apart (site A and C in Figure 4) on
December 9, 1989.

FIG. 9. Estimates of (a) coherence, (b) gain, and (c) phase of


the transfer function relating signals measured at two receivers 0 km apart (site B' related to B) and 18 km apart (site C
related to A). Data at sites (B, B') and sites (A, C) were
recorded, respectively, on December 5 and 9, 1989.

104

Vallee et al.

30

E
<,

3-

25

c-,

.....
.....

Downloaded 12/18/14 to 180.254.108.125. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

3-

Location 0

1Il
C
Ql

"C

~NAA

Ql

.....
.c....

18

"C

15

.-<
Ql

24

;>,

......
....
1Il
c

20

.....
c
.....

30

E
<,

NAA

.-<
Ql

.....
.....

.....
.....
u

12

.u....

10

.....

.....

.....

Ql

Ql

Solar flare

el
lC

el
lC

Solar flare

.-<
lC

~NSS

.-<
lC

.....

0
t-

.....

0
10

11

12

13

14

0
t-

15

Eastern daylight-saving time

0
10

16

11

12

13

14

15

Eastern daylight-saving time

(hours)

10. NAA intensities measured with two EDA Omniplus


receivers located 27 km apart on May 10, 1990. Occurrences
of solar flares are indicated according to solar observations.
FIG.

a)

16

(hours)

FIG. 11. NAA and NSS intensities recorded on May 10, 1990
at location D with an EDA Omniplus receiver. Occurrence
of a solar flare is indicated according to solar observations.

b)
20 , - - - - - - - - - - - - - - - - - - - - - - T

iD

10

Disturbed.

27 km

o Undisturbed.

0.8

0 km

;>,

1Il
C
Ql

Ql

0.6

(J

"C

Ql

E -10

:J
L

.c

Ql

.....

00.4
u

~ -20
1Il

L
Ql
~

0.2

-30

a.

Disturbed.

27 km

o Undisturbed.

0.0

-40 +-----,-----,-----,-------,------"'---"'t4
o
2
3
5

Frequency

(cycles/hour)

Frequency

c)

0 km

(eye les/hour)

d)
3.0 -r-r-

.,.

2.5

Disturbed.

90

27 km

o Undisturbed.

0 km

Disturbed.

27 km

o Undisturbed.

0 km

45
2.0

!ii'Ql
Ql

en

Ql

..... 1.5

lC
(')

Ql
1Il
lC

s:

a.

-45

0.5

0.0

+o

---,

,-2

Frequency

-.,.3

(cycles/hour)

---.
4

+
5

-90

+-

-,-

--,,-2

-r-t

-t-r-

Frequency (cycles/hour)

12. Estimates of (a) spectrum, (b) coherence, (c) gain, and (d) phase of the transfer function for receiver
separation of 0 km (site A' relative to A, quiet day) and 27 km (site D relative to A, with a disturbed ionosphere).
Data were recorded, respectively, on May 5 and May 10, 1990.
FIG.

+
5

VLF Primary Field Variation

Downloaded 12/18/14 to 180.254.108.125. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

CONCLUSIONS

Airborne VLF surveys have been successful in detecting


local conductivity anomalies, particularly those due to clayfilled shear zones in resistive Precambrian terrains (Soonawala and Hayles, 1986; Sinha, 1990). Such information has
been of considerable value in structural geologic mapping in
Canada, and was extensively used in mineral exploration.
However, interpretation of VLF data has been mostly qualitative so far. While numerous attempts have been made to
interpret VLF data in a more quantitative way, the inconsistency in measured amplitudes always posed a serious obstacle.
Some sources of measurement errors have been identified in
this paper and their prediction have been explored.
Variations in VLF total-field intensities that can affect
results of geophysical surveys can be divided into two
categories: spatial and temporal variations. Spatial variations have been investigated by modeling of the primary field
intensity as a function of distance. The results clearly
indicate the presence of Hollingworth interference patterns.
As there is no clear remedy for the disturbance, total-field
VLF surveys should not be carried out using a transmitter
which, for the given survey area, is likely to be affected by
this phenomenon. Before commencing field surveys in a
given area, modeling should be carried out to identify the
regions likely to be affected by primary field minima.
Base stations have been used routinely in VLF total-field
surveys to correct for temporal variations. In a series offield
experiments, the correctness of this practice has been tested.
The results indicate that the parameters of the transfer function
relating the signals recorded at two receivers varies with their
separation and with frequency. Serious errors would be introduced for greater distances between the base station and the
sensor. Even for a distance of 27 km, which was the greatest
receiver separation in the experiment, the two signalscould not
be well correlated. In airborne surveys, the aircraft is often
much further from the base camp, where the primary field
strength is monitored, and, in such cases, inappropriate corrections will distort the field data. In view of our findings, it is
recommended that the existing surveying practice be reviewed.
For instance, it may be desirable to set up several base stations
in large survey areas.
Solar flares, whose intensity will reach an l l-year peak in
1992, cause serious disturbances in VLF signals. Data acquired during such events cannot be corrected and hence
should not be used for map compilation or data interpretation. Information on occurrence and intensity of solar flares
can be obtained from research organizations, such as the
National Geophysical Data Center of the National Oceanic
and Atmospheric Administration, which continuously monitors solar and ionospheric activity.
ACKNOWLEDGMENTS

The staff of the Spectrum Control Center at St-Remi-deNapierville, Quebec, which is operated by the Department of
Communications, allowed us the use of their instrumentation.
Scintrex Ltd. of Toronto, Ontario, provided their VLF-3
receivers. Sagax Geophysics Inc. and Geophysique Sigma Inc.
of Montreal, Quebec, lent us EDA Omniplus instruments. Dr.
R. L. Grasty and R. Shives ofthe GeologicalSurvey of Canada
discussed with us various problems of VLF surveying and

105

critically read the manuscript. Dr. R. Coles of the Geological


Survey of Canada provided information on solar flares observed in May 1990. Part of the project was supported through
Quebec FCAR grant EQ-2632F. GSC Contribution no. 16991.
REFERENCES
Bracewell, R. N., Budden, K. G., Ratcliffe, J. A., Straker, T. W.,
and Weekes, K., 1951, The ionospheric propagation of low- and
very-low-frequency radio waves over distances less than 1000 km:
Proc. Inst. Electr, Eng., 98 III, 221-236.
Collett, L. S., 1986, Development of the airborne electromagnetic
technique, in Palacky, G. J., Ed., Airborne resistivity mapping:
Geol. Surv. of Can., Paper 86-22,9-18.
Cloos, E., 1934, Auto radio, an aid in geological mapping: Amer. J.
ScL,28, 166.
Davies, K., 1966, Ionospheric radio propagation: Dover Publ. Inc.
Dellinger, J. H., 1937, Sudden ionospheric disturbances: Terr. Mag.
and Atmos. Electr., 42, 49.
Hauser, J. P., and Rhoads, F. J., 1974, Coverage predictions for the
Navy's fixed VLF transmitters: Nat. Res. Lab. Memo. Rep. 2884.
Herz, A., 1986, Airborne EM instruments operating at VLF and
higher frequencies, in Palacky , G. J., Ed., Airborne resistivity
mapping: Geol. Surv. of Can., Paper 86-22,55--61.
Hollingworth, J., 1926, The propagation of radio waves: J. Inst.
Electr. Eng., 64, 579-595.
Howell, B. F., Jr., 1943, Some effects of geologic structures on radio
reception: Geophysics, 8, 165-176.
IAGA Division I, Working Group I, 1986, International geomagnetic reference field revision 1985: EOS, 67, 523.
Johler, J. R., 1970, Spherical wave theory for MF, LF, and VLF
propagation: Radio Sci., 5, 1429-1443.
Johler, J. R., and Berry, L. A., 1962, Propagation of terrestrial radio
waves of long wavelength-Theory of zonal harmonics with
improved summation techniques: J. Res. of Nat. Bur. Ltd., 66D,
737-773.
Johler, J. R., and Harper, J. D., 1962, Reflection and transmission of
radio waves at a continuously stratified plasma with arbitrary
magnetic induction: J. Res. of Nat. Bur. Ltd., 66D, 81-99.
Jones, T. B., and Mowforth, K., 1982, A review of the analytical
techniques for determining the phase and amplitude of a VLF
radio wave propagating in the earth-ionosphere waveguide, in
Belrose, J. S., Ed., Medium, long and very long wave propagation
(at frequencies less than 3000 kHz), AGARD Conf. Proc. 305,
NATO.
ITT, 1975, Reference data for radio engineers: Howard W. Sams
and Co. Inc.
National Oceanic and Atmospheric Administration, 1990, Sudden
ionospheric disturbances, May 1990: Solar-Geophysical Data
prompt reports, 551, part I, 113-117.
Paterson, N. R., and Ronka, V., 1971, Five years of surveying with
the Very Low Frequency electromagnetic method: Geoexploration, 9, 7-26.
Pressey, B. G., Ashwell, G. E., and Hargreaves, J., 1961, The phase
variation of very-low-frequency waves. propagated over long
distances: Proc. Inst. Electr. Eng., 108, Part B, No. 38, 214-226.
Sinha, A. K., 1990, Application of airborne very low frequency
electromagnetic surveys in geophysical exploration: in Fitterman,
D. V., Ed., Developments and applications of modern airborne
electromagnetic surveys: U.S. Geol. Surv. Bull. 1975, 117-131.
Sobczak, L. W., and Taylor, G. J., 1970, Results of a differential
Omega test in the Mackenzie River Delta: Geophysics, 35,
514-520.
Soonawala, N. M., and Hayles, J. G., 1986, Airborne electromagnetic methods in the concept assessment phase of the Canadian
Nuclear Fuel Waste Management Program: in Palacky, G. J., Ed.,
Airborne resistivity mapping: Geol. Surv. of Can., Paper 86-22,
153-158.
Thiel, D. V., and Chant, I. J., 1982, Ionospheric induced very
low-frequency electric field wavetilt changes: Geophysics, 47,
6Q-62.
Thomson, D. J., 1982, Spectrum estimation and harmonic analysis:
Proc. IEEE, 70, 1055-1096.
Wait, J. R., and Spies, K. P., 1964, Characteristics of the earthionosphere waveguide for VLF radio waves: Nat. Bur. Std. Tech.
Note 300.
Walden, A. T., 1990, Improved low-frequency decay estimation
using the multitaper spectral analysis method: Geophys. Prosp.,
38,61-86.
Watt, A. D., 1967, VLF radio engineering: Pergamon Press, Inc.

Das könnte Ihnen auch gefallen