Groundwater Geophysics Phần 5 ppsx

210 Anders Vest Christiansen, Esben Auken, Kurt Sørensen

• At early times, the offset configuration is extremely sensitive to small

variations in the resistivity in the near surface. Extensive 3D modelling

of such variations shows a pronounced influence on the measured fields

before the current system passes beneath the receiver coil. In many

cases these data are not interpretable with a 1D model, even if the sec￾tion is predominantly 1D. At later times, after the current system has

passed, the distorting influence has decayed. The central-loop configura￾tion is much less affected by near-surface resistivity variations.

• The offset configuration is sensitive to small deviations in the array ge￾ometry. For a simple 60 m half-space model, a 30% error in the re￾sponse is apparent near the sign change if the receiver coil is located 71

m instead of 70 m from the transmitter. In a routine field situation, it is

next to impossible to work with such accuracy. After the sign change,

the offset configuration is essentially equivalent to a central-loop con￾figuration; the central-loop configuration is insensitive to the placement

of the receiver inside the transmitter loop.

• A compromise is to use a high-power system where early times are

measured in the central loop configuration with a small current of 1 – 3

A. Late times is, in turn, measured in the offset configuration with

maximum output current. In this way the four issues are addressed, and

the field production can still be kept high.

6.9 Airborne TEM

In this chapter we will give an overview of the airborne TEM system

and discuss the specific topics where the airborne and the ground based

techniques differ. We will focus on the relatively new helicopter systems

as they have the sufficient accuracy necessary for groundwater investiga￾tions.

6.9.1 Historical background and present airborne TEM

systems.

Airborne electromagnetic systems (AEM) have been used for more than

50 years. The development was driven by the exploration for minerals with

its needs for surveying large areas within reasonable cost. The first at￾tempts with airborne TEM systems in the 1950s were quite successful in

base-metal exploration in Canada, and in that decade over 10 systems were

in the air. The most successful system resulting from the 1950s was the In￾duced PUlse Transient (INPUT). Canada and the Nordic countries led the

6 The transient electromagnetic method 211

development and use of AEM systems, and by the 1970s the methodology

was seen used worldwide.

With the decline in exploration for base metals, the use of AEM meth￾ods turned from anomaly detection to conductivity mapping, and fre￾quency-domain helicopter EM (HEM) systems appeared. By the 1990s

base-metal exploration was concerned with deep targets, and AEM sys￾tems began to follow two paths: fixed-wing time-domain systems designed

for detection of deep conductive targets, and frequency-domain HEM sys￾tems intended for high-resolution, near-surface, conductivity mapping.

Of the more than 30 systems appeared since the inception of the AEM

method, few are currently in routine use. The GEOTEM and the

MEGATEM systems are digital enhancements of the INPUT system,

which uses a half-sine transmitter waveform. The TEMPEST system uses

a square transmitter waveform as is common for ground-based TEM sys￾tems.

Table 1. Key parameters of different airborne transient systems.

Name of

equipment

Moment

in

kAm2

Transmitted

waveform

Configuration

and measured

components.

Type of

Calibration

Carrier type

GEOTEM 450 Half-sine Offset-loop,

Z and X

Relative Fixed-wing

MEGATEM 1500 Half-sine Offset-loop,

Z and X

Relative Fixed-wing

TEMPEST 55 Trapezoid Offset-loop,

Z and X

Relative Fixed-wing

AeroTEM 40 Triangular Central-loop,

coplanar, Z

and X

Absolute Helicopter

Sling-load

HoisTEM 120 Trapezoid central-loop

coplanar, Z

Relative Helicopter

Sling-load

VTEM 400 Trapezoid central–loop

coplanar, Z

Relative Helicopter

Sling-load

SkyTEM 120 Trapezoid central-loop

coplanar, Z

and X

Absolute Helicopter,

Sling-load

The TEM systems mentioned above are fixed-wing systems, i.e. systems

with the current-loop strung around an airplane from the nose, tail and

wing tips. Only recently has the concept of a transient helicopter system

come of age, and new systems are emerging making broadband measure￾ments with a small footprint possible. Transient helicopter systems carry

the transmitter loop as a sling load beneath the helicopter. Recently devel￾oped helicopter TEM systems are the AeroTEM, NEWTEM, Hoistem,

212 Anders Vest Christiansen, Esben Auken, Kurt Sørensen

VTEM and SkyTEM systems. The AeroTEM, NEWTEM, Hoistem, and

VTEM systems are designed primarily for mineral exploration. The Sky￾TEM system is designed for mapping of geological structures in the near

surface for groundwater and environmental investigations and was devel￾oped as a rapid alternative to ground-based TEM surveying. Table 1 sum￾marizes the key parameters of the airborne TEM systems currently in op￾eration.

6.9.2 Special considerations for airborne measurements

In groundwater exploration, data with precision and quality are required

as the decisive data changes can be as low as 10 – 15 %. When operating

in the air a number of key issues need to be addressed to achieve the re￾quired data quality. The issues are all related to the calibration, the altitude

and the flight speed of the system.

6.9.2.1 Calibration

In the context of high data quality, the calibration of the transmit￾ter/receiver system plays a central role.

When airborne systems operate in the frequency domain, the strong

primary field has to be compensated in order to measure the Earth re￾sponse. Because of drift in the system the compensation changes in time,

and its size has to be determined successively during the survey by high￾altitude measurements. Furthermore, it is necessary to perform measure￾ments along tie lines perpendicular to the flight lines and by post￾processing to provide concordance between adjacent lines. This process is

called levelling, and because of this a frequency system is said to be rela￾tively calibrated.

When airborne systems are operating in the time domain, it is possible

to reduce the interaction between the transmitter and the receiver system to

a level, at which the distortion of the measured off-time signals is negligi￾ble. In this case, a calibration of the instruments can be performed in the

laboratory and/or at a test site before the equipment is used in surveys.

Neither high altitude measurements nor performing tie lines for levelling

are then necessary during the survey. Such a system is said to be absolute

calibrated.

The relatively calibrated systems have a lower S/N ratio and a lower

data accuracy because of the levelling and the filtering of data compared to

the absolutely calibrated systems.

6 The transient electromagnetic method 213

6.9.2.2 Altitude

The Earth response decays with increasing altitude. This is illustrated in

Fig. 6.17a). The model resembles a conducting clay cap above a resistive

aquifer layer situated on a good conducting clay basement.

The random noise contribution from natural and man-made sources has

no significant change within the operating range. Therefore, the decay in

the Earth response solely causes a lower S/N ratio at late times resulting in

a poorer resolution of the deeper part of the Earth.

The determination of the near surface layers also decreases with higher

altitude because the fields have weakened. Fig. 6.17c) shows the standard

deviation as a factor for the model parameters of the model in Fig. 6.17b).

The determination of the resistivity of the first and the second layer and the

thickness of the first layer decrease when the system moves from the

ground to an altitude of 100 m. The thickness of the second layer remains

well determined because it is very thick. In general, increasing altitude

means a lower resolution of the upper layers. Related to groundwater in￾vestigations, the above figures show that high resolution of near-surface

protecting clay layers requires operation at low altitudes.

Another implication of the decaying Earth response with altitudes is in￾creased distortions of the Earth response due to coupling to man-made in￾stallations. As mentioned in chapter 6.7, a safety distance to installations

of at least 100 m, depending on the model, has to be maintained in order to

avoid distorted data sets. Airborne electromagnetic measurements intro￾duce larger safety distances to installations compared to ground based

equipment because of the lower Earth responses. The larger the flying alti￾tude is the larger are the safety distances. If the signal at late times has de￾creased by a factor of X, the safety distance must be increased by a factor

of X (assuming the coupling is caused by an infinite wire with field de￾cay proportional to 1/r

2

, r being the distance to the wire). For the model in

Fig. 6.1, the safety distance at an altitude of 50 m is approximately 1.4

times larger than at the surface. At an altitude of 100 m it has increased to

approximately 1.7.

6.9.2.3 Flight speed

An important tool for increasing the S/N ratio in electromagnetic meas￾urements is to perform stacking and filtering of measurements (see chapter

6.6).

In TEM measurements the noise is reduced by stacking the individual

transient decays. To achieve a certain S/N ratio, a certain number of tran￾sient decay curves are necessary in the stack.

214 Anders Vest Christiansen, Esben Auken, Kurt Sørensen

Fig. 6.17. Altitude and resolution. a) shows the Earth response as a function of al￾titude for the model in b). The transmitter moment is 22,500 Am2

and responses

are measured in the central-loop configuration. The transmitter height is varied

from 0 m to 100 m in steps of 10 m. The response decreases more at early times

than at late times. Data above the noise indicated by the dashed line are obtained

until 1.8 ms at an altitude of 100 m and 3.2 ms at the surface (dotted lines). Plot c)

shows the standard deviation as a factor (a factor of 1 means 0% uncertainty) for

the parameters of the model in b) assuming the noise model indicated with the

dashed line in a). Resistivities are solid lines, thicknesses dotted lines.

If the noise affecting the data sets is uncorrelated Gaussian noise, we

have for the standard deviation, STD, of the average of the stack that