WO2004031808A1

WO2004031808A1 - Vector magnetic data processing

Info

Publication number: WO2004031808A1
Application number: PCT/AU2003/001297
Authority: WO
Inventors: Mark Dransfield; Guimin Liu; Asbjorn Norlund Christensen
Original assignee: Bhp Billiton Innovation Pty Ltd
Priority date: 2002-10-04
Filing date: 2003-10-03
Publication date: 2004-04-15
Also published as: AU2002951919A0; CA2501121C; CA2501121A1; AU2003266819B2; AU2003266819A1

Abstract

This invention concerns airborne geophysical surveys and the processing of vector magnetic data from those surveys to remove the aircraft effect. Coefficients representing the permanent and induced magnetic effects are calculated from data obtained during a calibration flight, with reference to the aircraft coordinate system. Then an airborne geophysical survey flight is conducted, and the permanent and induced magnetic effects are removed from the survey flight data in the aircraft coordinate system. The corrected survey flight data is rotated into the NED coordinate system, and the eddy-current effect is removed from the corrected survey data in the NED coordinate system.

Description

Title

Vector Magnetic Data Processing

Technical Field

This invention concerns airborne geophysical surveys and the processing of vector magnetic data from those surveys to remove the aircraft effect,

Background Art

Magnetic total field surveys have been used in airborne mineral exploration for many decades. Only a single component of the total magnetic field is measured at any survey point. The measurement has to be corrected to remove the aircraft effect. It is known that the aircraft effect includes three types of contribution: the permanent magnetic effect due to permanent magnets (hard-iron materials) on the aircraft, the induced magnetic field due to soft-iron material^'s of the aircraft induced in the earth's field, and the eddy-current magnetic field due to electro agnetically induced electric currents in the metal aircraft frame. There are a number of techniques (Leliak, 1961; Tolles, 1954; Tolles, 1955; Passier, 1970) that remove these effects from the data.

Although it is considered desirable to measure three-component magnetic data in airborne surveys_. nd they are routinely recorded in total field magnetic surveys in the aircraft coordinate system, they are only used to calculate the aircraft attitude for the compensation of the total field. It had not been possible to rotate the Ihree components of the magnetic field from the aircraft frame to the earth's coordinate system without independent measurements of the aircraft attitude angles. In magnetic surveys for mineral exploration, the aircraft attitude is in general not measured directly.

However, a technique (Assous and Petilion, 1997) was developed mat uses gyroscopic aircraft attitude measurements to correct the magnetic field measurements for the purpose of correcting heading angle drift in aircraft navigation.

Summary of the Invention The invention is a method of airborne geophysical surveys and the processing of vector magnetic data from those surveys to remove the aircraft effect, comprising the steps of: conducting a calibration flight in the survey area, comprising four straight line segments at constant attitude in different directions, and measuring the aircraft attitude 5 and the total magnetic field relative to the heading angle, elevation angle and bank angle of the aircraft in each segment; calculating coefficients representing the permanent magnetic effect in the heading angle direction, elevation angle direction and bank angle direction from the calibration flight data assuming the earth's main field is constant; 10 calculating coefficients representing the induced magnetic effect in the heading angle direction, elevation angle direction and bank angle direction from the calibration flight data assuming the earth's main field is constant; conducting an airborne geophysical survey, and measuring the total magnetic field relative to the heading angle, elevation angle and bank angle of we aircraft in each 15 segment; then, removing the permanent magnetic effect, which is independent of aircraft attitude, from the survey flight data, by subtracting the magnetic field components calculated using the coefficients representing the permanent magnetic effect from each of the components of the total magnetic field vector relative to the heading angle,

20 elevation angle and bank angle of the aircraft; removing the induced magnetic effect, which is dependent on the dimension, shape and susceptibility of the parts of the aircraft body, but independent of the orientation of the aircraft, from the survey flight data, by subtracting the magnetic field components calculated using the coefficients representing the induced magnetic effect 25. from each of the components of the total magnetic field vector relative to the heading angle, elevation angle and bank angle of the aircraft; rotating the corrected survey flight data into the NED coordinate system; computing a coefficient tensor for the eddy-current magnetic effect by data regression applied to each line of the survey flight, computing the eddy-current

3 magnetic effect at each point by multiplying the tensor by the inducing time-derivative field, and then subtracting the component of the eddy-current effect from the survey data in the NED coordinate system.

This method has the advantage that it is dynamic in time since eddy-current correction 35 coefficients are obtained for every survey line in the earth's coordinate system rather than using constant coefficients in the aircraft's coordinate system derived from a calibration flight

The formulas for removing the permanent magnet effect, induced magnetic effect and eddy-current magnetic effect of the aircraft may be derived from the model of Leliak (1961), which is similar in concept to the techniques presented in Assous and Petilion (1997, US Patent 5,654,635), and Rice, Jr. (1993, US Patent 5,182,514).

A number of additional steps may be performed to minimise noise in the survey data. These steps include:

Removal of residual aircraft attitude dependence by applying a further regression step to the corrected data on the aircraft attitude for each survey line, to further reduce data noise on each survey line,

Further removal of residual aircraft attitude dependence by applying a moving- window regression of the data on the aircraft attitude angles and their time derivatives along each survey line. The window is in the range 10 to 100 seconds.

The use of International Geomagnetic Reference Field (IGRF) for reference in the calculation of correction coefficients to overcome errors in the calibration constants of magnetometers. Line levelling using total magnetic intensity (TMI) data rather than data from magnetic component sensors to correct the mean levels of the North, East and Down components on each line. And,

Amplitude correction using TMI data rather than data from magnetic component sensors to correct the North, East and Down components so that the magnitude of the vector is the same as the TMI value at every station.

The final corrected NED components of the vector residual magnetic data and its magnitude (the vector residual magnetic intensity or VR I) are all included in the output data that may be written to the survey database.

Use of the invention provides a significant reduction of the noise in the vector magnetic data. Preliminary data processing results show excellent performance of the new technique in noise reduction. This indicates that delivery of the vector magnetic data as a standard product can be achieved in the near future. The time-derivatives may be calculated using finite-difference of the magnetic data in the aircraft frame, instead of angle changes used by Assous et al (1997).

Brief Description of the Drawings

An example of the invention will now be described with reference to the accompanying drawings, in which:

Fig. 1 is a diagram illustrating the relationship between the aircraft-based LTN coordinate system, the world-based ED coordinate system, and the aircraft attitude variables (heading angle, elevation angle and bank angle).

Fig. 2(a) is graph comparing the North component of raw data using the new and old techniques for removing the aircraft effect; Fig. 2(b) is graph comparing the

Down component of raw data using the new and old techniques for removing the aircraft effect and Fig. 2(c) is graph comparing the East component of raw data using the new and old techniques for removing the aircraft effect.

Fig. 3(a) and Fig. 3(b) are plots of Vector Mag Residual Intensity (VMRI) data using the new and old techniques, for removing the aircraft effect.

Best Modes of the Invention

The method first requires a calibration flight at high altitude in the survey area. The flight path is roughly a square with a side length of a few kilometres. The correction coefficients for the permanent and induced magnetic effects are calculated from the calibration flight data. These coefficients are then subsequently used in the correction of vector magnetic survey data.

The measured magnetic field M is composed of the earth's field H, including any ore- body effect, the permanent magnet field of the aircraft A, the induced magnetic field of the aircraft I, and the eddy-current magnetic field E. Hence

M = H + A+ I + E

In the aircraft reference frame, there are three equations at each observation point for the three magnetic field components,

M_L ^ H_L + A_L + I_L + EI. (1) Mτ~ H_τ + Aτ+ Iτ + E (2)

Mv = Hv + Av + I_v + Ev (3)

As illustrated in Fig, 1, here the subscript L denotes the component in the longitudinal direction of the aircraft, T the transverse direction, and V the vertical direction. For the calibration flight data, the earth's main field is assumed to be a constant and is the known International Geomagnetic Reference Field (IGRF) field in the earth's North, East, Down (NED) reference frame. Thus, the LTV components Ht, H_T, and Hy can be calculated by rotation with the known aircraft attitude data (pitch, roll and yaw).

The permanent magnet field components A , A_T. and Ay are constants that are independent of aircraft attitude.

The L component of the induced magnetic field of the aircraft at the sensor is

I_L = HL LL+ HT'TL + HvNL (4)

where LL is the magnetic field in L direction due to induced magnetic dipoles in the L direction for an unit inducing field, TL is the magnetic field in L direction due to mduced magnetic dipoles in the T direction for an unit inducing field, and VL is the magnetic field in L direction due to induced magnetic dipoles in the V direction for an unit inducing field. LL, TL and VL are essentially geometrical factors that are independent of the aircraft attitude variations.

Similarly, the T component of the induced magnetic field of the aircraft at the sensor is

Iτ = H 'LT+HτTT + H_vNT (5)

and the N component of the induced magnetic field of the aircraft at tiie sensor is

Iv = H_L'LN+ H_T'TV + HvW (6)

Here, (LL, TL, VL, LT, XT, VT, LV, TV, W) are only dependent on the dimension, shape, and susceptibility of the parts of the aircraft body, but independent of the orientation of the aircraft. The eddy-current magnetic field is produced by eddy currents in the aircraft body. A change of magnetic flux through a conducting loop will generate a current proportional to the time derivative of the flux in the loop. This current will produce a secondary magnetic field opposing the change in the magnetic flux. As the aircraft hull effectively consists of conducting loops of aluminium, these loops will experience a change in magnetic flux as the aircraft changes direction in the earth's magnetic field. These current loops will generate a secondary magnetic field measurable as the eddy-current field at the sensor. The L component of the eddy-current field can be written as

^J,,_{U +}^J _{t t} ∑.._{vl {7)}

^L dt dt dt

where 11 is the magnetic field in L direction due to eddy-current magnetic dipoles in the L direction for an unit inducing field, ti is the magnetic field in L direction due to eddy- current magnetic dipoles in the T direction for an unit inducing field, and VL is the magnetic field in L direction due to eddy-current magnetic dipoles in the V direction for an unit inducing field.

Similarly,

dt dt dt ^w

E_v = — - < iv+ 2- -. + — Ϊ-- VW (9) dt dt dt

Here, (11, ti, vl, It, tt, vt, lv_s tv, w) arc only dependent on the dimension, shape, and electrical conductivities of the parts of the aircraft body forming the conductive loops, but independent of the orientation of the aircraft.

Substituting equations (4 (9) into equations (1), (2) and (3), we obtain

H_L +A_L +H_L -LL+H_τ - π, + H_v M_L (10)

nil ΛΪΪ ΛΪΪ H_T +^ +H_L -LT+H_T 'IT + Hy -VT + ^-lt + ^-U + ^ -vt = M_T (11) dt dt. dt _y +A_y +H_L -LV +H_T .:?r+H *^+ — ■' +— - + — " v = Λf\, (12) ^r dt dt dt r ^\ We need to solve for (A_L, A_τ, A_v), (LL, TL, VL, LT, TT, NT, LV, TV, W), and (U, tl₅ vl, It, tt, vt, lv, tv, w) 21 unknown coefficients.

From the calibration flight data, we can set up the above three equations at each data point. So we end up with a linear system of equations to solve for the coefficients. If the attitude angles of the aircraft are constant on a straight-line segment of the flight path, there are only three independent linear equations on this segment under the assumption the earth's magnetic field is constant on a calibration flight covering a small area. Since there are four straight-line segments with different head angles in the calibration flight path, there are a total of 12 independent linear equations. So it is an under-determined problem to solve for the 21 unknown coefficients. However, the magnitude of eddy-current magnetic field is much smaller than the permanent magnetic and induced magnetic fields and we can first ignore the eddy-current terms and solve for the 12 (A A_τ, Av) and (LL, TL, VL, LT, TT, VT, LV, TV, W) factors for the permanent magnet and induced magnetic dipoles. The eddy-current coefficients (11, ti, vl, It, tt, vt, lv, tv, w) are calculated by regression dynamically line-by-line on the survey data.

For normal survey data, the correction algorithm first removes the permanent and induced magnetic effects of the aircraft using the coefficients calculated from the calibration flight data. This is done in the LTV coordinate system. The corrections formulas are as follows:

H_L ^M_L -A_L -M_L - L-M_τ -TL-M_Ϋ - VL (13)

H_T =M_T -A_T ~M_L -LT-M_T -TT-M_V > VT (14)

Hy =M_y -A_v -M_L -LV-M_τ -TV-M -W (15)

The corrected data are then rotated into the NED coordinate system. After this, a coefficient tensor (11, t , vl, It, tt, vt, lv, tv, w) for the eddy-current magnetic effect of the aircraft is computed from the data by regression on a line-by-line basis. The eddy- current magnetic effect is then computed at each point by multiplying the inducing time-derivative field with -the eddy-current coefficient tensor an subtracted from the data. There are then a number of additional steps for survey data correction:

Removal of residual aircraft attitude effect

After the correction of the permanent, induced and eddy-current magnetic effects of the aircraft, a regression of the corrected data on the aircraft attitude is done on each survey line and the residual effect of the aircraft attitude is removed from the data.

Moving-window reduction of noise

After tiie removal of residual aircraft attitude effect, it was noted that the three component magnetic data were still correlated with the aircraft attitude angles and their time derivatives in some portions of a survey line. Therefore a moving-window regression of the data on the aircraft attitude angles and their time derivatives is done along each survey line to reduce noise. The window size is a parameter in the range of 10 to 100 seconds.

Levelling vector magnetic data line-by-line using TMI data

If the magnetic component sensors are less accurate than the TMI sensor, one may wish to correct the mean levels of the North, East and Down components on each line to those calculated from TMI data to remove stripes in the component data. The processing algorithm has an option to do this if desired for noise reduction.

Vector magnitude correction

If the magnetic component sensors are less accurate than the TMI sensor, one may wish to correct the North, Eas and Down components so that the magnitude of the vector is the same as the TMI value at every station. The processing algorithm has an option to do this if desired for noise reduction.

Vector Residual Magnetic Intensity (VRMI)

The algorithm also calculates the vector residual magnetic data by the subtraction of a constant vector magnetic field. This constant field is often the average earth's main field in the survey area or the IGRF field if required. The NED components of the vector residual magnetic data and its magnitude (the vector residual magnetic intensity or VRMΪ) are all included in the output.

Of course, other parameters such as inclination and declination of the vector residual data can also be calculated and used in data interpretation. Fig. 2 shows a comparison of the data using the new and old techniques to remove the aircraft effec A visual inspection suggests a noise reduction improvement of a factor between 3 to 10. It should be noted that there is also a difference in the low frequency component in the corrected data in some cases (e.g. top of Fig. 2). This may be because tiie assumption of a sine head angle correction may not be strictly valid although it is a good approximation.

Fig. 3 shows a comparison of the Vector Magnetic Residual Intensity (VMRI) of data using the new and old techniques to remove the aircraft effect. The VMRI is the magnitude of the residual magnetic vector after subtracting the vector IGRF earth field from the data. The improvement using the new technique for vector magnetic data processing is illustrated clearly in Fig. 3,

It will be appreciated by persons skilled in the art that numerous variations and/or modifications may be made to the invention as shown in the specific embodiments without departing from the spirit or scope of the invention as broadly described. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive.

References

Assous, E. C, and Petilion, J., 1997, Method and device for simultaneous identification and correction of errors due to magnetic perturbations and to misalignments in the measurements of a magnetometer, US Patent 5654635.

Leliak, P., 1961, Identification and evaluation of magnetic field sources of magnetic airborne detector equipped aircraft, IRE Transactions on Aerospace and Navigation Electronics. Passier, F. A., 1970, Aircraft magnetometer system with means to compensate said system for disturbing magnetic field generated by the aircraft, US Patent 3,530,375.

Rice, J. A. Jr., 1993, Automatic compensator for an airborne magnetic anomaly detector, US Patent 5,182,514. Tolles, W. E., 1954, Compensation of aircraft magnetic fields, US Patent 2,692,970.

Tolles, . E., 1955, Magnetic field compensation system, US Patent 2,706,801. van Leeuwen, E.H., 2000, BHP develops airborne gravity gradiometer for mineral exploration: The Leading Edge, VoL 19, No 12, 1296-1297.

Claims

CLAIMS:

1. A method of airborne geophysical surveys and the processing of vector magnetic data from those surveys to remove the aircraft effect, comprising the steps of: conducting a calibration flight in the survey area, comprising straight line segments at constant attitude in different directions, and measuring the aircraft attitude and the total magnetic field relative to the heading angle, elevation angle and bank angle of the aircraft in each segment; calculating coefficients representing the permanent magnetic effect in the heading angle direction, elevation angle direction and bank angle direction from the calibration flight data assuming the earth's main field is constant; calculating coefficients representing the induced magnetic effect in the heading angle direction, elevation angle direction and bank angle direction from the calibration flight data assuming the earth's main field is constant; conducting an airborne geophysical survey, and measuring the total magnetic field relative to the heading angle, elevation angle and bank angle of the aircraft in each segment; then, removing the permanent magnetic effect from the survey flight data by subtracting the magnetic field components calculated using the coefficients representing the permanent magnetic effect from each of the components of the total magnetic field vector relative to the heading angle, elevation angle and bank angle of the aircraft; removing the induced magnetic effect from the survey flight data by subtracting the magnetic field components calculated using the coefficients representing the induced magnetic effect from each of the components of the total magnetic field vector relative to the heading angle, elevation angle and bank angle of the aircraft; rotating the corrected survey flight data into the NED coordinate system; computing a coefficient tensor for the eddy-current magnetic effect by data regression applied to each line of the survey flight, computing the eddy-current magnetic effect at each point by multiplying the tensor by the inducing time-derivative field, and then subtracting the components of the eddy-current effect from the survey data in the NED coordinate system.

2. The method according to claim 1, comprising the further step of applying a further regression step to the corrected data on the aircraft attitude for each survey line.

3. A method according to claim 1 or 2, comprising the further step of applying a moving-window regression of the data on the aircraft attitude angles and their time derivatives along each survey line.

4. A method according to claim 3, wherein the window is in the range 10 to 1 ^QO seconds.

5. A method according to any preceding claim, wherein the International Geomagnetic Reference Field (IGRF) is used for reference in the calculation of correction coefficients.

6. A method according to any preceding claim, comprising the further step of using total magnetic intensity (TMI) data to correct the mean levels of the North, East and Down components on each line.

7. A method according to any preceding claim, comprising the further step of using TMI data to correct the North, East and Down components so that the magnitude of the vector is the same as the TMI value at every station.

8. A method according to any preceding claim, comprising the further step of outputting corrected NED components of the vector residual magnetic data, or its. magnitude (the vector residual magnetic intensity or VRMT), or both, to the survey database.