https://ntrs.nasa.gov/search.jsp?

R=19850023272 2019-10-17T11:13:01+00:00Z

General Disclaimer

One or more of the Following Statements may affect this Document

This document has been reproduced from the best copy furnished by the
organizational source. It is being released in the interest of making available as
much information as possible.

This document may contain data, which exceeds the sheet parameters. It was
furnished in this condition by the organizational source and is the best copy
available.

This document may contain tone-on-tone or color graphs, charts and/or pictures,
which have been reproduced in black and white.

This document is paginated as submitted by the original source.

Portions of this document are not fully legible due to the historical nature of some
of the material. However, it is the best reproduction available from the original
submission.

Produced by the NASA Center for Aerospace Information (CASI)

 r

E85' 10 .1031
/49'-w- Oe- 175-6 Y
FINAL TECHNICAL REPORT

NASA Grant No. NAG5-304

W. J. Hinze and L. W. Braile

Department of Geosciences
Purdue University
West Lafr!ette, IN 47907

aid 1 t -

R. R. B. von Fr?se
Dept. of Geology 6 M:aeralog
The Ohio State University
Columbus, Oh 43210

i
(E85-10103 NASA-CG-175614) 1MOVING THE NE_-315E5
GEOLOGICAL 197EBPSEIAIION CF BAGIETIC AMD THFU
GUATITY SATELLITE ANCIALIES Final Report N85-3 1600
(Purdue Univ.) 67 F HC A04/Mk A01 CSCL 08G Unclas
G3/4 3 00 103

SEC ^ ^^►s

Jti
1
 I
L , ',a%

This report describes research performed under NASA contract NAG5-304

entitled "IMPROVING THE GEOLOGICAL INTERPRETATION OF MAGNETIC AND GRAVIW

SATELLITE ANOMALIES". The original proposal include) a two-year research

plan that was to be initiated by 1 Jan. 83. The objective of the proposal

-aas to investigat, current limitations in the quantitative interpretation of

satellite-elevation geopotential field data and to develop technique; to

overcome them or to determine the limitations that they impose on geological

interpretation. Originally, the plan was to study orbital effects upon the

quality of the data, including the utilization of downward continuation

techniques, improved forward and inverse modeling procedures, and the

utility and procedures for geological interpretation of gradient and vector

geopotential satellite anomaly components. The investigation was to be

performed analytically and empirically using geologically and geophysically

constrained model studies, culminating in the application of the improved

interpretational techniques to select anomaly features of the MAGSAT data.

Funding constraints imposed by NASA delayed initiation of this program and

severely reduced the overall level of funding. However, despite these

i
modifications a significan portion of the original scope of research was

maintained.

A major result has been the preparation of an improved scalar magnetic

anomaly map for South America and adjacent marine areas directly from the

original MAGSAT orbital data. This effort involved specially developed

procedures for isolating crustal anomaly profiles which were subsequently

reduced to a constant elevation and common radial pole by equivalent point

source inversion. These procedures and resultant map were presented at the

1984 Spring meeting of the American Geophysical Union under the following

citation:

A
 2

1) Ridgway, J. R., W J. Hinze and L. W. Braile, 1984, MAGSAT scalar

anomaly map of South Ame • :ca, EOS (Am. Geophys. Union Trans.),
V. 65, p. 202.

These have also been documented under the following thesis citation:

2) Ridgway, J. R., 1984, Preparation and Interpretation of a Reviseu

MAGSAT Magnetic Anomaly Map Over South America, Unpubl. MS-thesis,
Purdue Univ., 121 pp.

The errors of numerically averaging satellite magnetic anomaly data for

geologic analysis have been investigated using orbital anomaly simulations

of crustal prisms by Gauss-Legandre quadrature integration. These simulations

have shown that numerical averaging errors make small and relatively minor

contributions to the total error-budget of higher orbital estimates ( 400 km),

whereas for lower orbital estimates the error of averaging may substantially

increase. As an alternative to numerical averaging, least-squares collocation

was also investigated and found to produce substantially more accurate anomaly

estimates as the elevation of prediction is decreased towards the crustal

sources. These results were presented at the 1984 and 1985 Spring meetings of

the American Geophysical Union under the following citations:

3) Goyal, H. K., R. R. B. von Frese, J. R. Ridgway, and W. J. Hinze,

1984, Geological analysis of averaged magnetic satellite
anomalies, EOS (Am. Geophys. Union Trans.), V.65, p. 202. t^

4) Goyal, H. K., R. R. B. von Frese and W. J. Hinze, 1985, Statistical

magnetic anomalies from satellite measurements for geologic
analysis, EOS (Am. Geophys. Union. Trans.), V. 66, p. 255.

These results have also been submitted for publication under the following

citation:

5) Goyal, H. K., R. R. B. von Frese, and W. J.

of satellite magnetic anomalies, Geophys
(submitted).
0 1 3

Quantitative geologic modeling of MAGSAT data has mostly focused to date

on the eastern portions of the prominent transcontinental magnetic anomaly of

the 11. S. This feature involves the crustal structure of the south-central

U.S., the Misissippi Embayment and central Kentucky which is related to the

mineralization and seismicity of a highly urbanized portion of the country. A

crustal magnetic model of the south-central anomaly based on MAGSAT data was

presented at the 1984 Spring meeting of the American Geophysical Union under

the following citation:

6) Starich, P. J., W. J. Hinze, and L. W. Brailt, 1984, The south-

central United States magnetic anomaly, EOS (Am. Geophys. Union
Trans.), V. 65, p. 202.

Fuller documentation of these results are -jntained in the following

t..-sls-

7) Starich, P. J., 1984, The South-Central United States Magnetic

Anomaly, Unpubl. MS-thesis, Purdue Univ., 76 pp.

To date our spherical-earth magnetic anomaly modeling procedure has

proved complex and cumbersome in many geologic applications. Accordingly,

simpler and faster flat-earth modeling was investigated and generally found

to be a practical alternative to the spherical-earth modeling algorithm when

dealing with gravity and differentially reduced to pole magnetic anomaly

data at satellite elevations. These results have been presented at the 1984

Spring meeting of the American Geophysical Union under the following citation:

8) Parrot, M. H., W. J. Hinze, L. W. Braile, and R. R. B. von Frese,

1984 9 A comparative study of spherical and flat-earth geopotential
modeling at satellite elevations, EOS (Am. Geophys. Union Trans.),
V. 65, p. 181.

They have also been documented for publication under the citation:

9) Parrot, M. H., W. J. Hinze, L. W. Braile, and R. R. B. von Frese,

1984, A comparative study of spherical and flat-earth geopoten-
tial modeling at satellite elevations, Geophys. Res. Lett.
(submitted).
 4

Three papers involving the tectonic analysis of MAGSAT anomalies for the

region (30°W, 60°E) (40 0 , 70°N) have been prepared by our group. These

include:

10) Oliver, R., W. J. Hinze and R. R. B. von Frese, 1983, Reduced to

pole long-wavelength magnetic anomalies of Africa and Europe, EOS,
(Am. Geophys. Union Trsns.), V. 65, p. 214.

11) Hinze, W. J., R. Olivier and R. R. B. von Frese, 1983, Euro-African

MAGSAT --omaly-tectonic observations, IUGG XVIII General Assembly,
Programme b Abstracts, V. II, p. 630; and

12) von Frese, R. R. B., R. Oliver and W. J. Hinze, 1983, Long-

wavelength magnetic: and gravity anomaly correlations of Africa and
Europe, IAGA Bulletin, v. 47, p. 140.

The first paper illustrates the utility of differential magnetic pole

reduction for geologic analysis of regional magnetic satellite anomalies,

where-s the second paper considers in some detail the tectonic implications

of reduced to pole Euro-African MAGSAT anomalies. The third paper

illustrates the synergism of MAGSAT and regional free-air gravity anomalies

for enhanced geologic interpretation.

Comparison of the reduced to pole MAGSAT anomalies for Euro-Africa with

the South Amer ; zan data has shown the strong correlation of anomalies along

the Atlantic rifted margins of the continents. This is a very interesting i

result, as to date it has been the mgnetic anomalies of the oceans which in

addition to paleomagnetic studies have provided much of the critical

evidence and details on continental drift and the evolution of the oceans.

To investigate this further, we reduced the 2°-averaged MAGSAT anomalies to

a common elevation and radial pole for the eastern Pacific Ocean, North and

South America, the Atlantic Ocean, Europe, Africa, India, Ausralia and

Antarctica. The results indicat!d that continental anomalies have prominent

affiliations with ancient features and that they demonstrate remarkably

dPrQiled correlation of lithospheric magnetic sources across rifted margins

 `7
5

5m
when plotted on a reconstruction of Pangea. Accordingly, these anomalies,

first observed as a result of NASA's magnetic satellite programs, describe

geologic conditions of considerable age and provide new and fundamental

constraints oa the geologic evolution of the continents and their

reconstructions.

These results were presented as an invited paper at the recent GRM

conference and in a paper presented at the 1985 Spring meeting of the

American Geophysical Union under the following citations:

13) von Frese, R. R. B., and W. J. Hinze, 1984, Continental and oceanic
magnetic anomalies: Enhancement through GRM, Geopotential
Research Mission Science Conference, Univ. of MD., Abstracts (Oct.
29-31, 1984), 4 pp.

14) von Frese, R. R. B., W. J. Hinze, R. Olivier, and C. R. Bentley,

1985, Continental magnetic anomaly constraiA s on continental
reconstruction, EOS (Am. Geophys. Union Trans.), V. 66, p. 255.

These results have also been documented for publication as follows:

15) von Frese, R. R. B., W. J. Hinze, R. Olivier, and C. R. Bentley,

1985, Regional magnetic anomaly constraints on continental
rifting, Geology (in-press).

In general, considerable and significant progress has been made under

the auspices of this research program. This incudes thirteen papers that

have been presented or submitted for publication and the completion of two

master's theses that rave been partially supported by this contract.

Attached is an APPENDIX which includes the papers submitted for publication

and the abstracts for presented papers and theses supported by the research

program. The abstracts and papers are listed in the APPENDIX in the order

that they are cited in this report.

I
i

Citation ail (Presentation Abstract) -a) I

MAGSAT SCALAR ANOMALY MAP OF SOUTH AMERICA

N8 s - 315 8g
J.R. Ridgway, W.J. Hinze and L.W. Braile
Department of Geosciences
Purdue University
West Lafayette, IN 47907

A scalar magnetic anomaly map has been prepared for South America and

adjacent marine areas directly from original MAGSAT orbits. The preparation

of the map poses special problems, notably in the separation of external field

and crustal anomalies, and in the reduction of data to a common altitude.

External fields are manifested in a long-wavelength ring current effect, a

medium-wavelength equatorial electrojet, and short-wavelength noise. The

noise is reduced by selecting profiles from "quiet" periods (k P < 3), and

since the electrojet is confined primarily to dusk profiles, its effect is

minimized by drawing the data set from dawn profiles only. The ring current

is corrected through the use of the standard ring current equation, augmented

by further filtering with a Butterworth bandpass filter. Under the assumption

that the time-variant ring current is best removed when a replication of

redundant profiles is achieved, a test set of 25 groups of 3 nearly coincident

orbits per group is set up for filtering with a range of long-wavelength

cutoffs, spanning 22 degrees to infinite wavelength, to determine which cutoff

best replicates the residual profiles. Replication is determined by linear

regression, which results in a correlation coefficient, a slope, and an inter-

ceps. By using these parameters in a triple test, the long-wavelength cutoff

which best removes the ring current is found to be 50 degrees. Profiles thus

filtered differ primarily in amplitude due solely to satellite altitude

differences. These differences are then normalized b y an inversion of the

profile data onto a grid of equivalent point dipoles, and recalculated at an

altitude of 35U km. The resulting map, when compared to the 2' averaged map,

shows more coherent anomalies, with notable differences in the region affected

by the electrojet, and promises much in regard to improved geologic interpre-

tat ion.
 r
citation #2 (Thesis Abstract)
N85-31587
MAGSAT. SATELLITE MAGNETIC ANOMALY MAP OVER SOUTH AMERICA

J.R. Ridgway
De p artment of Geosciences
Purdue University
West Lafayette, IN 47907

A scalar magnetic anomaly map has been prepared for South America and

adjacent marine areas directly from original MAGSAT orbits. The preparation

of the map poses special problems, notably in the separation of external field

and crustal anomalies, and in the reduction of data to a common altitude.

External fields are manifested in a long-wavelength ring current effect, a

medium-wavelength equatorial electrojet, and short-wavelength noise. The

noise is reduced by selecting profiles from "quiet" periods (Kp < 3), and the

effect of th= electrojet is minimized by drawing the data set from dawn I .1^

profiles only. j,

The ring current is corrected through the use of a standard equation,

augmented by furtt-.e. digital band-pass fat-ring. Under the assumption that

the time-variant ring current is best removed when a replication of redundant

profiles is achieved, a test set of 75 pairs of coincident orbits are filtered

with a range of long-wavelength cutoffs to determine which cutoff best

replicates the residual profiles. Replication is determined by linear

regression, which results in a correlation coefficient, a slope, and an

intercept. UEing these parameters in a triple test, the long-wavelength

cutoff which best removes the ring current is found to be 5U degrees.

Profiles thus filtered differ primarily in amplitude due solely to

satellite altitude differences. These differences are normalized b y an

A^ ti

7
Y
'^_
inversion of the profile data into a grid of equivalent point dipoles, and

recalculated at an alL. t ude of 350 km. The low latitudes in the study area

cau_e instability in the inversion, necessitating sepaia-e irversions of

several sub-areas which are subsequently merged. The resulting map more

accurately depicts crustal anomalies around the geomagnetic equator than

previous maps of South America.

Crustal anomalies reduced-to-the-pole exhi luit marked correlations to

kno wn tectonic features, and snow promise for unproved geologir intepretation.

^*nj
V

^.
 N85 - 3I 5 g8
Citation X13 (Presentation kbstract)

GEOLOGIC ANALYSIS OF AVERAGED MAGNETIC SATELLITE ANOMALIES

H.K. Goyal (1 , R.R.B. von Frese (1 b J.R. Ridgway (1) and W.J. Hinze(1)

To investigate rela;ivt advantages and limitations for quantitative

geologic analysis of magnetic satellite scalar anomalies derived from arith-

metic averaging of orbital profiles within equal-angle or equal-area

parallelograms. the anomaly averaging process was simulated by orbital

profiles computed from spherical-earth crustal magnetic an..)maly modeli.-,g

experiments using Gauss- T .egendre quedrature integration. The results indicate

that averaging can provide reasonable values ar satellite elevations, where

contributing error factors within, a given parallelogram include the elevation

distribution of the data, and orbital noise and geomagnetic field attributes.

Various inversion schemes including the use of equivalent point dip:)les are

also investigated as an alternative to arithmetic averaging. Although inver-

sion can provide improved spherical grid anomaly estimates, these procedures

are problematic in practice where computer scaling difficulries frequently

arise due to a combination of factors including large source-to bservation

distances (> 400 km), high geographic latitudes, and low geomagnetic field

inclinations. Finall y , a comparison of averaged scalar anomalies over South

America taken from the global <2 * > MAGSAT anomaly map with anomalies derived

from a detailed analysis of orbital profiles by equivalent point source

inversion indicates that external field and variable elevation effects

contribute significantly to distortion in the <2 * > anomalies.

(1) Dept. of Geology and Mineralogy, Ohio State University, Columbus, 04

43210

(2) Dept. of Geosciences, Purdue University, West Lafayette, IN 47907

is
 Dy
N8541589
Citation #4 (Presentation Abstract)

STATISTICAL MAGNETIC ANOMALIES FROM SATELLITE MEASUREMENTS FOR GEOLOGIC

ANALYSIS

H.R. Goyal (1 , R.R.B. von Frese (l , W.J. Hinze(2)

The errors of numerically averaging satellite magnetic anomaly data for

geologic analysis are investigated using orbital anomaly simulatio:ts of

crustal magnetic sources by Gauss-Legendre quadrature integration. These

simulations suggest that numerical averaging errors constitute small and

relatively minor contributions to the total error-budget of higher orbital

estimates 0 400 km), whereas for lower orbital estimates the error of averag-

ing may increase substantially. Least-squares collocation is also investigat-

ed as an alternative to numerical averaging and found to produce substantially

more accurate anomaly estimates as the elevation of prediction is decreased

towards the crustal sources.

(1) Dept. of Geology and Mineralogy, Ohio State University, Columbus, OH

43210

(2) Dept. of Geosciences, Purdue University, West Lafayette, IN 47907

I 1

_ ^rr
 N85-31590
Citation #5 (Refereed Paper)

BINNING OF SATELLITE MAGNETIC ANOMALIES

H. K. Goyal l , R. R. B. von Frese l , and W. J. Hinze2

l Dept. of Geology & Mineralogy 2 Dept. of Geosciences

The Ohio State University Purdue University
Columbus, OH 43210 W. Lafayette, IN 41907

Abstract

The errors of numerically averaging satellite magnetic anomaly data

for geologic analysis are investigated using orbital anomaly simulations of

crustal magnetic sources by Gauss-Legendre quadrature integration. These

simulations suggest that numerical averaging errors constitute small and

relatively minor contributions to the total error-budget of higher orbital

estimates (^ 400 km), whereas for lower orbital estimates the error of

averaging may increase substantially. Least-squares collocation is also

investigated as an aiternative to numerical averaging and found to produce

substantially more accurate anomaly estimates as the elevation of prediction

is .;ecreased towards the crustal sources.

Introduction

artificial, earth-orbiting satellites are making increasingly available

consistent, regional-scale magnetic anomaly data for geologic analysis. These

data are frequently presented as numerically averaged anomalies of orbital

profile anomaly values within equal-angle or equal-area parallelepipeds or

bins. (e.g., Regan et al., 1915; Langel et al., 1982; Ritzwoller and Bentley,

983). The averaged anomaly is normally assigned to the center of the area at

he arithmetically averaged elevation of the data within the bin. In general,

t is felt without quantitative basis that the averaging process limits the

tility of anomaly maps prepared from these data in g eologic analysis, but

he advantages to geologic interpretation of mere sophisticated and costlier

r

S. 2

data processing are also not clear. The oojective of this study is to
investigate the errors of numerical averaging for satellite magnetic anomaly

estimation using systematic spherical - earth magnetic anomaly simulations and

to evaluate, as an alternative to arithmetic averaging, magnetic anomaly

estimation by least-squares collocation.

Numerical Simulations and Procedures

The basic data of this study are orbital scalar magnetic anomaly values

of two dimensional, 40 km thick crustal spherical prisms with widths of 50,

100, 200, 300, and 500 tom (Fig. 0. The magnetic anomalies of these radially

( normally) polarized prisms were calculated assuming 3 A/m magnetization by

the Gauss - Legendre quadrature integration procedure ( von Frese et al. (1981)).

The anomaly values were computed at 0.25° intervals on 40"-length orbits

that are at 25 km levels over elevations ranging from 100 to 70U km. The

radial polarization and two dimensional assumptions were made to generalize

the results while simplifying the computations. f

Ii
Numerical averaging simulations were conducted using the procedure

Outlined by Langel et al. (1982) to obtain 2'-averaged scalar magnetic tt^

anomalies from MAGSAT. This procedure involves computing the mean and
f

standard deviation of the anomaly values within a 2°-parallelepiped, elimina-

ting anomaly values which are not within two standard deviations of the mean,

and then recomputing the mean of the remaining values as the 2 '-averaged

anomaly estimate. To evaluate numerical averaging errors, the averaged

anomaly values were computed from the simulated 2 * -bins with elevations

ranging from 100 to 300 km, 300 to 500 km, and 500 to 700 km, and compared

to the modeled anomaly values at the center of each bin at elevations,

respectively, of 200 km, 400 km, and 6UU km.

Least-squares collocation, a statistical estimation technique which is

widely applied to problems in physical geodesy (e.g., Moritz, 1972), was also
 3

used in this study. For satellite magnetic anomaly prediction, the observea

anomaly values, LT, Within a bin, may be modeled as

IT = L'1C+N.

When the crustal anomaly, LTC, and random noise, N, are not correlated, a

least-squares estimate of the crustal anomaly can be obtained from

LTC p - C(LTC,LT)TC(LT,LT)-1LT,

where C(LTC,LT) is the cross-covariance matrix between LTC and LT, and

C(LT,LT) is the covariance matrix of observations.

For optimal prediction, the statistical behavior of the anomalies can be

assessed through an appropriate covariance function. In this study,

covariances for all separation distances between observations within a 2*-

bin were computed by forming the corresponding anomaly products which were,

in turn, sorted into successive uniform intervals of the separation distance.

For each interval of the separation distance, the average of the anomaly
IM
products, normalized to the covariance value at zero separation distance, was

plotted against the mean separation distance to obtain the normalized

covariance function shown in Fig. 2. For a spherical earth the separation

distance, K12, between points P(r1,6 1 ,4 1 ) and Q(r 2 ,6 2 ,4 2 ) is given by

K12 - (r12+r22-2r1r,)(cos8lcos62+sin6lsin62cos(41-42))1/2

where (rl,el,0i) and (r2 ,62,42) represent the radial, co-latitude and

longitude coordinates for P and Q, respectively.

The normalized covariance functions for all bins considered in this study

were found to follow in a least-squares sense the functional form given by

Cn (R) - 1.00-(0.78*10-2)R-(0.2*10- 4 )1:2,

where C n(R) is the normalized covariance value at separation distance, R.

Covariance values for the various separation distances are given by

C(R) - C (R)*C(0),
n

1
 4

wnere C(0) is the value of the covariance function in nT 2 at zero separation,

and R is the separation distance in km. This relationship was observed to

hold for R < 100 km. For the covariance values in each bin, the correla-

tion length was close to 56 km, where the correlation length, &, is the value

of R for which C(&) - C(0)/2. Accordingly, observations within a radius of

about 100 km of the prediction point were used to form the elements of the

covariance matrix C(LT,LT).

Results and Discussion

Typical examples of the radially polarized anomaly signals considered in

this study are shown in Fig. 3. The upper panel (Fig. 3.A) illustrates how

the anomalies broaden and increase in amplitude at 2UU km elevation witn

increasing source widths, whereas the lower panel (Fig. 33) shows how the

magnetic anomaly of a crustal prism of width 20U km increases in amplitude and

wavenumber with decreasing orbital elevation. Similar observations hold at

the other elevations for all the source widths considered in this study.

Fig. 4 provides characteristic results of how the averaging procedure

described by !.angel et al. (1982) performed in these simulations. Tne dolled

line in each panel shows the small relative deviation (nT) of the estimates

which are defined as the modeled (true) anomaly (solid line) minus the 2°-

averaged anomaly value obtained from analysis of all the computed anomalies

Within each bin, which in this case nas no values perpendicular to the

profile. However, in the normal preparation of satellite magnetic anomaly

maps, orbits with large external magnetic field activity indices are routinely

rejected to enhance crustal anomaly components. This frequently results in

severely decimated coverage with respect. to the anomaly values available for

averaging within a bin. To simulate this condition, the computed values

within each bin were randomly decimated to 20% the original coverage.
 5

Characteristic 2°-averaged anomaly aeviat ions for the sparse data coverage are

plotted as the dashed curves on the panels of Fig. 4.

In general, Fig. 4 shows that maximum deviations for both dense and

sparse data coverage principally occupy the flank and peak regions of the

anomalies, where the 2' • -averages derived from sparse data coverage show the

largest and most erratic deviations. Accordingly, to further evaluate errors

of averaging, subsequent study focused on 2'-averaged predictions from the

decimated data for the various crustal prisms at the locations of the peak

amplitude, 50% peak amplitude, and WX peak amplitude of the true !modeled)

anomaly profiles over the principal elevations of 20U km, 4UU km, and COO km.

An overview of these results is presented in TABLE 1, where at each of the

principal elevations the true amplitude at the point of prediction is listed

along with A) the max'.mum averaging errors as defined by (deviation/true

value)*100% and B) the maximum averaging deviations. Corresponding collo-

catior errors and deviations in TABLE I are not trivial functions of

elevation because the prediction is also affected by such factors as the

variable number and distribution of data within the bins (Cruz, 19oi).

These results suggest ths: the simple averaging procedure provides

remarkably good anomaly estimates at elevations of 400 km and greater for the

simulations considered here. In practice, the location of a satellite

magnetometer measurement is typically known only to within plus or minus

several tens of meters of the true position. This mislocation can produce

errors in the calculation of the geomagnetic reference field as large as a

few nanoteslas at 40U km and which increase with increasing orbital

elevations. 'Wiese errors are introduced wnen the reference field is

subtracted from the observations to obtain the crustal anomaly residuals.

External magnetic field effects, which are currently not well understood ana

difficult to account for, constitute an even greater error source in these

 ^` vim. .r.^^ r. ^^^ .^-...-_ - - ---- _...a •.^1

.' n
residuals. Hence, inaccuracies due to numerical averaging represent only a

minor contribution to the overall error-budget of satellite magnetic anomalies

at these elevations. For lower elevations on the other hand, such as at 200

km where future satellite missions may be orbited, mislocation of the

magnetometer sensor introduces errors of only 1 or 2 nT or less, yet numer-

ical averaging can produce significantly large errors according to these

simulations.

For comparison with the numerical averages, predictions based on least-

squares collocation were also obtained for these sparsely populated

bins. Differences between the anomaly estimates produced by the two methods

smaller than or equal to 1.0 nT were attributed to numerical roundoff errors.

For any differences above this cutoff, collocation yielded a significantly

more accurate estimate than numerical averaging. Out of the 45 sparsely

l
populated bins tested at the various elevations and source widths, col ocation

yielded significantly improved anomaly estimates for roughly 332 of them. The

improvement of collocation predictions over averaging as defined by (1-(col-

location deviation/averaging deviation))*1002 was found to range from 511 to

992 over all the source widths (Fig. 5.A). However, improvement was more

concentrated and significant at the lowest principal elevation (200 km) i

considered in this study (Fig. 5.B and TABLE 1).

Coaclusions

Crustal magnetic anomaly signals over satellite orbits were simulated to

investigate numerical averaging as an anomaly estimator. Averaging is a

convenient procedure for reducing satellite magnetometer data to manageable

proportions for geologic analysis, although the precision of averaging as an

anomaly estimator involves significant problems concerning spatial and

amplitude smoothing of the satellite magnetic observations. The results of

the simulations considered in this study suggest that the error of numerical

1
 7

averaging constitutes a small and relatively minor component of the total

error -budget of higher orbital anomaly estimates (^ 4UU km), whereas for

lower orbital estimates numerical averaging error increases substantially.

As an alternative to numerical averaging, least-squares collocation was

investigated and observed to produce substantially more accurate anomaly

estimates, particularly as the orbital elevation of prediction was decreased

towards the crustal sources. In contrast to averaging, collocation is a

significantly more resource-intensive procedure to apply oecause of the

practical, but surmountable problems related to establishing and inverting the

covariance matrix for accurate anomaly prediction. However, as demonstrated

by these simulations, collocation may be much more effectively used to exploit

the anomaly details contained in the lower orbital satellite magnetic data for

geologic analysis.

Acknowledgements

The authors thank J. Y. Cruz of the Dept. of Geodetic Science at Ohio

State for stimulating discussion and initial software for im p lementing least-

squares collocation. Financial support for this investigation was provided by

the Goddard Space Flight Center under NASA contract hAGi-304.

.:M
References kited

Cruz, J. Y., 1983, Experience With Altimeter Data Gridding, t.ept. 047, Dept.
of Geodetic Sci., Tne Uh3o State University, Columbus, Uri.

Langel, R. A., J. D. Phillips and R. J. Horner, 1982, Initial scalar

magnetic anomaly map from MAGSAT, Geophys. Res. Lett., V. `l, p. 209-272.

Moritz, H., 1972, Advanced Least-Squares methuda, Rept. #175, Dept. of

Geodetic Sci., The Ohio State University, Columbus, Oh.

Regan, R. D., J. C. Cain and W. M. Davis, 1975, A global magnetic

anomaly map, J. Geophys. Res., v. du, p. 794-802.

Ritzwoller, M. H. and C. R. Bentley, 1983, Magnetic anomalies over

Antarctica measured from MAGSAT, in (Olivier, R. L., P. R. James and
J. B. Jago, eds.) Antarctic Earth Science-Fourth International
Symposium, Cambridge Univ. Press, NY, p. 504- 50.
 8

References (Cont.)

von Frese, R. R. B., W. J. Hinze, L. W. Braile and A. J. Luca, 1981,

Spherical earth gravity and magnetic anomaly modeling by Causs-
Legendre quadrature integration, J. Geophys., v. 49, p. 234-242.

TABLE I.A - Maxim-.,a Prediction Errors - Averaging vs. Collocation

(Sparse Data Coverage Simulations)

Orbital Elevation True Anomaly Amplitude Averaging Error Collocation error

[km] [nT) 1%) 1%)

200 11.6 86.2 0.7

400 3.0 61.3 26.8
600 2.2 24.7 7.0

TABLE I.B - Maximum Prediction. Deviations - Averaging vs. Collocation

(Sparse Data Coverage Simulations)

;,
Orbital Elevation True Anomaly Amplitude Averaging Collocation
Deviation Deviation
[km) [nTJ [nT) [nT)

200 99.3 25.0 0.4

400 49.0 2.7 1.0
600 2.2 0.5 0.2
J

Figure Captions

Fig. 1 Geometric and physical properties used for the satellite

orbital magnetic anomaly simulations. This study focused on
anomaly estimates for A a 2%

Fig. 2 The least-squares polynomial curve in separation distance, R,

fitting the normal:zed covariance functions of this study is given
by Cn(R) - 1.00-(0.78*10-2)R-(0.20*10-4)R2.

_Fig. 3 Characteristics of radially polarized magnetic anomalies for

different crustal prism widths at 200 km elevation (A), and at
different satellite elevAtions for a crustal prism 200 km
wide (B).

rte`

T -- -. M f4


Figure Captions (Cont.)

Fig. 4 Comparison of true anomalies due to a 200 km vide cructal prism

with corresponding deviations of 2'-averaged estimtes derived from
dense (dotted) and sparse (dashed) data coverge at (A) 2CO km
elevation, (B) 400 km elevation, and (C) 600 km elevation.

! Fig. S Percent improvement of collocation estimates over those obtained by

averaging plotted as a function of (A) the source widths and (B)
the principle satellite elevations considered in this study.
Numbers beside superimposed data points give the number of data
points in the cluster.

FIGURE 1

PARAMETERS FOR MAGNETIC SATELLITE ANOMALY AVERAGING SIMULATIONS

to
e /ice / ^ ^ _ ^'`.\ \ \ ^
►oO^l^

/// / / \ \\ 190

?0
OO^^ 04,4
ti ^ i

I O k m 1
11
Magnotlzatlon: 3 A/m
50, 100. 200, 300, 500 km

t
 U-^
FIGURE 2
0
0
TM
Y 0
Co
O
O
U-) .,
Y Y
O
O O O 0 v
O M O
M r v z
I I
0
O O
O O
CD l!')
O 0
co W N I
^ F— F-
^, Q
to ^ m W
2 O m cn
O W O
E ia'
cc O
W W W
3 ►— CL 0
CD z CL N
El E)
0
0 uA 0
^r 0
( Z 1N) IINHI^HA03
 i

FIGURE 3

125- 125
SOURCE WIDTH 500

75 " i , SOURCE WIDTH 200 75

J , SOURCE WIDTh 50
25- , 25

-25
77 -25
-20 -15 -10 -5 0 5 10 15 20
A WEST - ERST PROFILE (DEGREE)
200 200
z 1 ^_^

120 ANOMALY AT 100 KM 120

ANOMALY AT 300 KM
-- _
z 40 RIvOMAL Y AT Kr
40
500
I i
Z
r l

-40 -40
20 -15 -10 -5 0 5 10 15 2C
B WEST -- ERST PHOF I LE (DEGREE)
 FIGURE 4

i I
TRUE RNOMRLY l
I I ,

ANOMALY D E V I AT 10 N (100%
, DATR COVERAGE)

I
- ANOMALY DEVIATION 120% DATA COVERAGE)

20 45 -10 -5 0 5 10 15 2a
WEST - ERST PROFILE (DEGREE)
TRUE ANOMALY- I, ANOMALY DEVIATION (100%
DATA COVERAGE) -4-
5 , ,, 0.57
ED

5 I -1.5 cr-

w
ANOMAL v, DEVIATION (20% DATA COVERRGE)
5 -3.5 °
-^n -1S - Ir. -S n S In 1- 20
B WEST - EAST PROFILE (DEGREE)
3.0
RNOMRLY DEVIATION (100%
TRUE ANOMALT^ z
ORTA COVERAGE)
1 1 i N 1 1
,; 1.0 z
0

cl
X: 0 I 1.0
CJ
z ANOMALY DEVIATION (20% DATA COVERAGE) >
CL wl
-6 -3.0 °
20 -15 -10 -5 0 5 10 15 2 0
C WEST - ERST PROFILE (DEGREE)
 FIG RE S

ID O

ELEVATION 200 KM ELEVATION 400 KN
100 100

z
w
w 75 75
0
cc
w
'^ 5 0 50
0 100 200 300 400 500
A SOURCE WIDTH (KM)

p SOURCE WIDTH 50 KM
(D SOURCE WIDTH 100 KM p SOURCE WIDTH 300 KM

2^ SOURCE WIDTH 200 KM p( SOURCE WIDTH 500 KM

100 100
7,
I

w
^
63 I
I
i I I I i
2
LLJ 75 75
CD i 1
I
I ^ I I I i
r

I
I I
50 50
100 200 300 400 500 600

B ELEVATION (KM)

'N i
 C

Citation Y6 (Presentation Abstract) 1^^

4
N 85 -315 91
THE SOUTH-CENTRAL UNITED STATES MAGNETIC ANOMALY s

P.J. Starich, W.J. Hinze and L.W. Braile

Department of Geosciences
Purdue University
West Lafayette, IN 47907

A positive magnetic anomaly, which dominates the MAGSAT scalar field

over the south-central United States, results from the superposition of

magnetic effects from several geologic sources and tectonic structures in the

crust. The highly magnetic basement . s of this region show good correla-

tion with increased crustal thickness, above average crustal velocity and

predominantly negative free-air gravity anomalies, all of which are useful

constraints for modeling the magnetic sources.

The positive anomaly is compcsed of two primary elements. The western-

most segment is related to middle Proterozoic granite intrusions, rhyolite

flows and interspersed metamorphic basement rocks in the Texas panhandle and i
f
eastern New Mexico. The anomaly and the magnetic crust are bounded to the

west by the north-south striking Rio Grande Rift, a zone of lithospheric

thinning and elevated heat flow, in central New Mexico. The anomaly extends

eastward over the Grenville age basement rocks of central Texas, and is

terminated to the south and east by the buried extension of the Ouachita

System.

The northern segment of the anomaly extends eastward across Oklahoma and

Arkansas to the Mississippi Embayment. It corresponds to a general positive

magnetic region associated with the Wichita Mountains igneous complex in

south-central Oklahoma and 1.2 to 1.5 Ga. felsic terrane to the north. The

magnetic rerrane terminates along a roughly east-west line in southern Kansas.

A subdued northeasterly extension of the anomaly, from southwest Missouri

into the Great Lakes region, appears to be related to the felsic terrane which

extends northeast across the Midcontinent.

 r^

Citation #7 (Thesis Abstract)
N 85 - 315 92
THE SOUTH-CENTRAL UNITED STATES MAGNETIC ANOMALY

P.J. Starich
Dept. of Geosciences
Purdue University
West Lafayette, IN 47907

The South-Central United States Magnetic Anomaly is the most prominent

positive feature in the MAGSAT scalar magnetic field over North America. The

anomaly correlates with increased crustal thickness, above average crustal

velocity, negative free-air gravity anomalies and an extensive zone of Middle

Proterozoic anorogenic felsic basement rocks.

The anomaly and the magnetic eruct are bounded on the west by the north-

striking Rio Grande Rift, a zone of lithospheric thinning and high heat flow

in central New Mexico. The anomaly extends eastward over the Grenville age

basement rocks of central Texas and is terminated to the south and east at the

burial extension of the Ouachita Orogenic System which is the southern edge of

the North American Craton. The anomaly also extends eastward across Oklahoma

and Arkansas to the Mississippi Embayment. A subdued northeasterly extension

of the anomaly continues into the Great Lakes region. The feature terminates

along the east-west boundary of the felsic terrain in southern Kansas.

Spherical dipole source inversion of the MAGSAT scalar data and subse-

quent calculation of reduced-to-pole and derivative mape provide additional

constraints for a crustal magnetic model which corresponds geographically to

the extensive Middle Proterozoic felsic rocks trending northeasterly across

the United States. These felsic rocks contain insufficient magnetization or

volume to produce the anomaly, but are rather indicative of a crustal zone

which was disturbed during a Middle Proterozoic thermal event which enriched

magnetic material deep in the crust.

J
 ^g
Citation #8 (Presentation Abstract)

N85-31593
A COMPARATIVE STUDY OF SPHERICAL AND FLAT-EARTH
GEOPUTENTIAL MODELING AT SATELLITE ELEVATIONS

M.H. Parrott (1) W.J. Hinze (l) L.W. Braile (l) and R.R.B. von Frese(2)

Flat-earth modeling is a desirable alternative to the complex spherical-

earth modeling process, but do errors invalidate the use of flat-earth assump-

tions at satellite elevations? These methods were compared using 2^-D flat-

earth and spherical modeling to compute gravity and scalar magnetic anomalies

along profiles perpendicular to the strike of variably dimensioned rectangular

prisms at altitudes of 150, 300, and 450 km. Comparison was achieved with

percent error computations (sphericai-flat/spherical) at critical anomaly

points. At the peak gravity anomaly value, errors are less than ±5% for all

prisms. At T and T U of the p eak, errors are generally less than 10% and 40%

respectively, increasing to these values with longer and wider prisms at

higher altitudes. For magnetics, the errors at critical anomaly points are

less than -10% for all prisms, attaining these magnitudes with longer and

wider prisms at higher altitudes. In general, in both gravity and magnetic

modeling, errors increase greatly for prisms wider than 500 km, although

gravity modeling is more sensitive than magnetic modeling to spherical-earth

effects. Preliminary modeling of both satellite gravity and magnetic anoma-

lies using flat-earth assumptions is justified considering the errors caused

by uncertainties in isolating anomalies.

(1) Dept. of Geosciences, Purdue University, West Lafayette, IN 47907

(2) Dept. of Geology and Mineralogy, Ohio State University, Columbus, OH

43210

1
 N95
%F 9 4
Citation #9 (Refereed Paper)

A COMPARATIVE STUDY OF SPHERICAL AND FLAT-EARTH

GEOPOTENTIAL MODELING AT SATELLITE ELEVATIONS

M.H. Parrott, W.J. Hinze, L.W. Braile

Department of Geosciences
Purdue University
West Lafayette, In 47907

and

R.R.B. von Frese

Department of Geology and Mineralogy
Ohio Stare University
Columbus, OH 43210

*;

July, 1984

1
n^
 i

•.6._ '

ORIGCNAL PAGr_ iQ
ABSTRACT 3F POOR QUALITY

Fla earth and spher_ca:-earth ;,eo;)otent_aI r.03e_inZ of cr_._al

ar,cnaly sources at satell:te elevations are conpare.-' by comb, t:r.G

gravity and scalar maSnetic anomalies perpendicular to the strike of

variably Cinensioned rectangular prisms at altitudes of 150, 300, an::

450 km. Results indicate that the error caused by the flat-earth
i

approximation is less than 10% in most Geo::,etric conditions.

Generally, errors increase with larger and wider anomaly sources at

hi_r her altitudes. For most crustal source nodelinE ap,-licatior.s at I

conventional satellite altitudes, flat-earth nodelin,-, can be justified

aad is nu_lerically efficient. I

INTRCDJCTION

It is well known that eeopotential mode 1i r,- at sate:':te

elevations using flat-earth assumptions disre3ards the sphericity of

the earth. To eliminate resultint errors, spherical-earth modeling

has been developed and implemented is recent years (von Frese et ai.,

19E1). however, currently available spherical modelin al-orith-s are

far more complex than fiat-earth a1FcrI'thns, res,:it i:.0 i n a 1

ccn!i j era.ae increase in ccz".static.".al effcr.. Le:a:: °e of the r:-ea..

co^.p.statiO::al erficiency :..0 ] ca.:ses flat — t-art.) MOueli no to De d

desiratle alternative, it is therefore i_ 7ortant to under Stan ,., the

errors involved with the use of flat-Barth assu'_"tions at satellite

_ elevations. A comparative study of the two modelirZ; procedures was

,,.er.°or:-,ed to deter - ine the errors res.:itin; fro= the use of t!.e
fiat-eartt assumptions for both rravit; an: scalar !LaZnetic anomalies

due to various source geometries.

`.Z


AKA:.YSIS PP,OCEDUR7

A cccy,:tational approach was selectee: for the e."ror analysis.

F iat-earl..", ceceline uses 2112 dirJer..:c:,^_ ;ra'.ity and t^a; netic

modelini, proarans, w;:ich assw'e hoi:o — - , k-ous roter.tia: field

characteristics throughout ;he anomalous body. Spherical modeling was

performed by an algorithm which computes gravity or :ja-r.etic anomalies

by concentrating the potential field characteristics in a specified

number of point sources distributed th,rourhout the anomalous body.

The comparative study involved computation of iravity and scalar

magnetic anomalies, using both modeliZg techniques, along common

profiles perpendicular to the stri p e of variably dimensioned

rectan-ular prisms. These computations were made at observation

I

elevations of 150, 300, and 450 km for bodies with widths varyink; from
i

50 to 2000 ^-n, and strike lengths of both 2 and 10 tires the

elevation. The radial (vertical) dimension of the boc'ies v , as 'eld

constant, extending from the sirface to a depth of 1:0 km. Figure 1

shows the specifications of the prisms and observet_cn profiles. T':e s

density contrast (0.25 g/cc) and susceptibility contrast (0.005

3
en-./cn ) were arbitrarily chosen, anc are simply scaling factors for

t'.-.e ;6r; --z e5 of tC:15 study. :121-netic ano.alies were calculated

v ertica ll ,olar: .z."t
^ :on a :r cs5 `h
^,e er-.ir a o u rca e
s out ocy to

facilitate the comparisons, but test` r. Z ci s.ie, e_ ;rofiles ::es

o c o
perfcrced using inclinations of 75 , 60 , and 45 . The profiles were

co::;ared by calculating the percent error (sp:hericLi-flat/:^,hericai)

at critical anomaly points alone the profiles. This a:etho.: o.`

analysis eliminates rob,jective scaling factors. Cravity anomalies

we-e compared at the maxiouc anomaly value, an.'. at 112 and 1/10 of the

J RICAAAL PAGE - (S
OF POOR QUALITY

r
 _ r_-.rte.-.^^^ -« . •^
^- .w r-^ „'_..^.- -`-^•

CRIGiNAL PAGE' iS
OF
^1 1Q' TY
:^::=u.:. Fa^net:c anomalies were compared at the m
, a - 1/2 of the

caxir:::z, and at tne =i.nimu- value. Also, the offset of tne =inimu¢

,Z%etic ano:.aly values was calculated.

RESULTS

Errors for the various source geometries and observation elevation

for gravity and maSnetic modeling are shown on Figures 2-8, with each

fiEure slowing the percentage error for variable prism width and

observation elevation, at strike lengths of 2 and 10 times the

elevation. Figures 2-4 show the error at critical gravity anomaly

points. Figure 2 indicates the errors at the maximum value as being

less than 5; in all cases. At 1!2 and 1/10 of the maxiex--, Figures 3

and 4 show a trend of increasing error with increasing distance from

the anomaly r^.idpo _r.t. ;,arimum errors are 12f at 112 of the ;eak

value, and 405 at 1/10 of the peak, increasinS to those values with

increasin;; prism width, strike length, and observation elevation.

Figures 5-8 show the errors resultinS fro= =P-Fnetic modeling usin3

flat-earth assumptions. Figures 5 and 6 snow the errors at the

maxi=u= and 112 of the maximum as being iesc than 12$ and 11`%

res^ec__vely, ir.creas_rg tc those values with ircre_s_n^ ;rise ti__. n,

e 1en -.- th, ar.- o ' -serv::t_en elevation. e; a 1.'!.i_U_ ba_ ie,

F.:-zj re 7 sh o ws ':he errors to t)e less than 2: for all cases. Use of

the two modelin,, techni;ues also creates an offset in the Lini =''jm

nno=aly value, which is indicated by Figure E, with the offset

increasing with increasing observation elevation ar.d strive length and

decreasing prism width. Several of the con;arisons were repeated

o 0 o
usirZ inclinations of 75 , 60 , and 45 with si_ilar resu:.ts. A.

these inclinations, each test showed the error at the maxirun value to

vary not p ore than 1.13 f r om the vertical polarization case. 1

0q%-"N W---


I;

7 i :-ures e and 10 show the extre_=e examples for the comparison of I _,

bot'.. Sravity and :aSnetic modeling usinS fiat-earth (dashed line) an?

spherical-earth (solid line) modelinS. These figures de_onstrate the

range in error anticipated ,using flat-earth modeling for data acquired

at satellite elevations.

COI&IIUSIONS

Generally, preliminary modellnE of both naSnetic and gravity

satellite anomalies using flat-earth assumptions is justified.

However, a modeler should use flat-earth assumptions only with full

awareness of the errors involved, and with added discretion when

modelinS bodies wider than 500 km. For these bodies, gravity modelinS

comparisons show errors increasinS greatly as compared to those

associated with smaller bodies, at all observation elevations.

Magnetic modeling shows similar large increases in error, but only at

observation elevations of greater than 300 km. At an elevatir.n cf 150

I.
ko, errors are relatively constant for all bo.-y dimensions, as

magnetic modeling is less sensitive to spherical-earth effects than

gravity modeling. Increasing errors alonS the flanks shoul r. not b-^ of

great concern to the modeler, because those 1_^Ze percent;_!:E errors

reflect lii_te.. :fferEnce! in ma"nitude die to the s_ali 2.;- litu: es

along the flanks, and also because of errors associated with 11

uncertainties in precisely isolatinS anomalies.

REFEREhCE

1. von Frese, R.ri.B., W.J. Hinze, L.W. Braile an 4 A.J. Luca,

I
Spherica'-eij•th gravity and magnetic mojelirZ by Gauss-Le;,endre

quadrature integration, J. Geofl'r.YSics, _^_

O , 234-242, 1961.
i

FIGJRZ CAPTION'S iP MOR QUALITY

Figure 1. "pecifications of anomalois prises and observation profile.

F:-ure 2. ercentage error (spherical-flat/spherical) at raxi=us

;;ravity an y - '• values with strike length c a) 2 f elevation and b)

104elevation.

Figure ?. Perce-tape error (spherical-fiat/spherical) at ore-half

maximum gravity anomaly value with strike length. = a) 2 f elevation and

b) 10felevatiop.

Figure 4. Percentage errcr ,'spherical-flat/spherical) at one-tenth

maximum gravity anomaly value with strike leagth = a) 2 0 elevation and

b) 100elevation.

Figure 5. Percentage error (spherical--flat/spherical) at r1e.:imim

magnetic anomaly value with strike iei,r.. = a) 2 1 elevstion and b)

10felevation.

Figure 5. Percentage error (spherical-flat/spherical) at one-half

maxiourr aagneLic anomaly value with strike length = a) 2 f elevation and

^r
b) 10#tlevation.

Figure 7. Percentage error (spherical-flat/s}herical) at mini_um

7-c-netic anoLla'- value with s:rike lenzt'% = a) 2 f Ele v a`ic n. an_ :)

! J;elevation.

Fi;ll re &. Offset of r2inimum ma:'-etic an. =-cl)' values with strike

ler,gth = a) 2*elevation an:: b) 10felevation.

Fi 'sure 9. Conpari:^on of Sravity ano_Elies c,,::pta ed by fist-earth

(dashed line) and spherical-earth (solid line) methods, with

dimensions o; : a) z=150 kill, F;_50 kri, str_ke=2z and b) z=45;D 1CM, w=2000

ks, stri.ke=10z.

— - _ ,
-^ ►^.^F ^ =^- _ . ...i.
^ 'fie
Figure 10. Comparison of magnetic anozalies cozputed by flat-earth

(dashed line) and spnerical-earth solid-iine) cet ods, with

di=, ensions of: a) z=150 ka, w=50 km, strl;ce_2z and b) z=450 km, =2000

ks, strike=10z.

Figure I , . Linear regression_ analysis of extreme exE=ples c.`' tt.e

gravity anomaly comparisons.

Figure 12. Linear regression analysis of extreme exar-ples of the

magnetic anomaly comparisons.

t
1

— — TU
 i

)n

r ivv2c
 pr j
Aw
rN
3 ,
it
S
f'•I
^/
L
1
•U
F--
1
3
N
^n
.O
O o o CD CD
.-.r
m LO Cr N I
i A
-^ o
i
m Y
0
2
F-
O
3
N ^
CD ►fir
CL
ti
t

I ^^
L ^
4- ^11*-^ O
I
3
N
X Y Y
CD 1^-r
LO O 0
11 II II
Z N N N fI
1
cD E O
^I
0
I
• O O O 0
O I
m
N O
--+ CO (D N
d^dd^ 1N^33d^d
S
a
I I I ^ ^ I I
0
s
Y i
c ^ rte-
2 i
O
3
v ^
O
i
..
C>
I —4
O
O O
• O -^ N m Y
Q Cr N
dOd^^ 1N^^d^d
 s^
Y
CD
0
s
N
LO
Y Y Y CD 1--4
0-
of
Ln
11 II II
N
ED E)
s
I
CD
C C:) I C
^.^ C^ C^ 0 CD
m (D U- m cli
V-
CD
c
r
ED
3
N ^
Z
Y E
Y Z
Y
O
0 O l)
.r c ?
11 11 II
IC P I CD
I ^ I
^ I
u
,_..^
1
0
I
O U7 O
80^8^ 1N338^d
 Now
I !
n ^
SC
O
^' I
r20
3
I ;I
O CO
a..
O
^„J LJ v v
N
4... M
• O
cz CO '"' "'
O N
I I I I 1 I
dOd^J^ .1N^^d^d
s
c
I 1 I I ^ ! ^ I 1 --^ o
1
I_ ^
O Y
^ I
2
3
N
CD Cc
I r
C_ O O
c p N
1N33^3J

n
F'Z^
E . 01
M z
Y
,O I
S
O
3
N
Y
O 0
Q'
^ ^ S
11 II II
N N N
0 0
CD
N
G C'') CG 07
m

I
I`
-^ o
E3
Y
^° I
. S
0
3
N ^
Y Y
O
.-4 of
Ypp
I^f) O CS, IL
II 11 II
N N N
c s e
W .W
J
I I 0 ^ ^ ^ . O
• . .
O N ^' CO co
CD I
'^ N
a ! I I 1 I

1
IS
ORIGINAL PAQU
OF POOR QUALITY
Zl
O
S
O
3
r X:
Y ]p^C Y CID
^ ~'
LL O LLo If
M
Q_
II 11 II
N N N
W
Z
J r.
I ' I I I -^ O
O
U)
CU
I I I I
I ^ I ^ I ^ I ^ I ^ i
-^ o
0^ I
I
O
3
ti S
O
Lo
I I^ I I I ^I
Ln O LI3 •
Q^ ^ I I I I
i'
^0^^^ 1N3^^^d
 Q
►I ^l ►1
^! I
N fJ N
I-w
j
^ I

Y
2 i
0 I-
3
N X:
O M I
of —i — o
It 6 o
0 0
N 0
4 0
m 7
I
i
-- 0
C Y Y ^ I
p
L!3 Q ^
11 ^I H r
m
.O
O
3
N ^
CD ►^-1
9-4
of
r
d
O
0V
v v `^
r-4
n 0
Q u)
0 0
M
0
N
0
x4s.-• *

^Y
V
I —O
O
OWGINAL PAGE' FS `I
OF POOR QUALITY 1
ti
i
I
l
U
O z
O Q
O co
LrD
O
1
1
t
I
i T--T-T
L0 CD
CV O Lf^ O Lf O
.--4 [- Lf) CV
^dJW-,^ IN^;^NH
r
I
LU
U
Z
O
V^
Y^r
O
Im
•
O, U^ O
p • O I ^,
1 !
^dJW -l^^Nl.10Nd I^
l^

I ` I O
YI
LJ
U
Z
Q
LO
• O O
O Lf) O • O
mCV .--^ r" Lr I I
1N-A^HWONH
I
wI
U
Z
CL
F-
^r^
O
• O
O U-) O O
•--^
a N W-4 Ll7 I
i^
1N-AIHWONH
► 7 % s•^
r. ^G^' ^'^ ,ems
 wr

20 .0-1 , , 1 1 , , , 11 , 1 11 11 11111 -
SOURCE DIMENSIONS
X: Y: i
W-50 KM W-2000 KM
rYr
/
L- 1 50 KM L-4500 KM Y /
r
Z-450 KM Z-450 KM r iCi
CL Y i
CD Y x

I Y ^(
J
Y
U x

ckf
W
10.0- Y
Y

x
-

U)

>r

REGRESSION ANALYSIS
X: Y:
r- 1 .00 r- 1 .GO

M-0.99 m- 1 .00

b-0.14 b-0.67
0
10.0 2010
FLAB-MGRL

c
11

i "+r w
 20.1. I_
SOURCE DIMENS'ONS

X: Y:
W - 50 KM W - 2000 KM Y
I Y
L- 1 50 KM L-4500 KM
Z- 1 50 KM Z-450 KM / /Y
Z i0.Q: Y
I /

i
X
, /Y
11..E X

0-
^ X

0/

-10 T^
-10.0 0 10.0 20.0
FLAT-NT

r^ 3.
c
fi
rn

^.
 -D/ O

Citation #10 (Presentation Abstrac ) N85-3

15 95

REDUCED TO POLE LONG-WAVELENGTH MAGNETIC ANOMALIES OF AFRICA AND EUROPE

R. Olivier (l) W.J. Hinze (1 ^ R.R.B. von Frese(3)

To facilitate analysis of the tectonic framework for Africa, Europe and

adjacent marine areas, magnetic satellite (MAGSAT) scalar anomaly data are i

differentially reduced to pole and compared to regional geologic information

and geophysical data including surface free-air gravity anomaly data upward

continued to satellite elevation. 050 km) on a spherical earth. Comparative

analysis shows magnetic anomalies correspond with both ancient as well as more

recent Cenozoic structural features. Anomalies associated with ancient struc-

tures are primarily caused by intra-crustal lithologic variations such as the

crustal disturbance associated with the Bangui anomaly in west-central Africa.

In contrast, anomalies correiative with Cenozoic tectonic elements appear to

be related to Curie isotherm perturbations. A possible example of the latter

is the well-defined trend of magnetic minima that characterize the Alpine

orogenic belt from the Atlas mountains to Eurasia. In contrast, a well-

defined magnetic satellite minimum extends across the stable craton from

Finland to the Ural mountains. Prominent magnetic maxima characterize the

Arabian plate, Iceland, the Kursk region of the central Russian uplift, and

generally the Precambrian shields of Africa.

(1) Institute of Ge•)physics, University of Lausanne, 5 rue de 1'Universite,

1005 Lausanne, Switerland

(2) Dept. of Geosciences, Purdue University, West Lafayette ; IN 47907

(3) Dept. of Geology and Miner logy, Ohio State University, Columbus, OH
43210

t
 Citation #11 (Presentation Abstract) 8 5- 3 1 5 9 6
N
EURO-AFRICAN MAGSAT AN011A1.Y-TECTONIC OBSERVATIONS

W.J. Hinze (l I R. Olivier (Z ^ and R.R.B. von Frese(3)

Preliminary satellite (MAGSAT) scalar magnetic anomaly data are compiled

and differentially reduced to radial polarization by equivalent point source

inversion for comparison with tectonic data of Africa, Europe and adjacent

marine areas. A number of associations are evieent to constrain analyses of

the tectonic features and history of the region. Rift zones and aulacogens,

for example, tend to be magnetically negative. The most intense positive

anomaly of the region is the Bangui anomaly which has been interpreted as due

to a deep crustal positive magnetization source. There are no near-surface

sources which will explain this anomaly. By contrast, the next most intense

positive anomaly is over the Kursk region in the Russian Ukraine. This

anomaly extends 450 km in a northeasterly direction and is roughly 150 km

wide, and is caused according to aeramagnetic anomaly interpretations by near-

surface, intensely magnetic ferruginous quartzites. Apparently there is

sufficient long-wavelength energy in these near-surface anomalies far them to

be observed at satellite elevations. The Precambrian shields of Africa and

Europe exhibit varied magnetic signatures. All shields are not magnetic highs

and, in fact, the Baltic shield is a marked minimum. The reduced to the pole

magnetic map shows a marked tendency for northeasterly striking anomalies in

the eastern Atlantic and adjacent Africa, which is coincident to the track of

several hat spots for tha past 100 million years. However, there is little

consistency in the sign of the magnetic anomalies and the track of the hot
 .--, ti
spots. Comparison of the radially polarized anomalies of Africa and Europe

with other reduced to the pole magnetic satellite anoiaaly maps of the Western

Hemisphere support the reconstruction of the continents prior to the origin of

the present—day Atlantic Ocean in the Mesozoic Era.

(1) Dept. of Geosciences, Purdue University, West Lafayette, :N 47907

(2) Institute of Geophysics, University of Lausanne, 5 rue de 1'Llniversite,

1005 Lausanne, Switerland

(3) Dept. of Geology and Mineralogy, Ohio State University, Columbus, OH

43210

.A
 Citation 012 ;Presentation Abstract) .J
N8 5 - 315 S7
LONG-WAVELENGTH MAGNETIC AND GRAVITY ANOMALY CORRELATIONS
ON AFRICA AND EUROPE

(2) and W.J. Hinze(3)

R.R.B. von Frese (0 R. Olivier

Regional geopotential anomalies and their correlations provide important

constraints for investigating the mesatectonic framework of Africa and Europe.

Accordingly, preliminary satellite (MAGSAT) scalar magnetic anomaly data are

compiled for comparison with long-wavelength-pass filtered free-air gravity

anomalies and regional heat-flow and tectonic data. To facilitate the

correlation analysis at satellite elevations over a spherical-earth, equiva-

lent point source inversion is used to differentially reduce the magnetic

satellite anomalies to the radial pole at 350 km elevation, and to upward

continue the first radial derivative of the free-air gravity anomalies.

Correlation patterns between these regional geopotential anomaly fields are

quantitatively established by moving-window linear regression based on

Poisson's theorem. Prominent correlations include direct correspondences for

the Baltic Shield, where both anomalies are negative, and the central Mediter-

ranean and Zaire Basin where both anomalies are positive. Inverse relation-

ships are generally common over the Precambrian Shield in northwest Africa,

the Basins and Shield in southern Africa, and the Alpine Orogenic Belt.

Inverse corelations also persist over the North Sea Rifts, the Benue Rift, r A

more generally over the East African Rifts. The results of this quantitative

correlation analysis support the general inverse relationships of gravity and

magnetic anomalies observed for North American continental terrane which may

be broadly related to magnetic crustal thickness variations.

(1) Dept. of Geology and Mineralogy, Ohio State University, Columbus, OH

43210

(2) Institute of Geophysics, University of Lausanne, 5 rue de l'Universite,

1005 Lausanne, Switerland

(3) Dept. of Geosciences, Purdue University, West Lafayette, IN 47907

s 'a
N85-w31598
Citation #13 (Presentation Abstract)

COFTINENTAL AND OCEANIC MAGNETIC ANOMALIES:

ENHANCEMENT THROUGH GRM

Ralph R. B. von Frese William J. Hinze

Pept. of Geology & Mineralogy Dept. of Geosciences

The Ohio State University Purdue University

Columbus, Ohio 43210 W. Laiayette, Indiana 47901

Earth-orbiting satellites over the past two decades have provided

consistent global magnetic anomaly data sets which have given us unique
insight on regional petrologic variations of the crust and upper mantle,
and crustal thickness and thermal perturbations. Verification of the
satellite magnetic data has been demonstrated by quantitative comparisons
with aeromagnetic anomalies of the conterminous U.S. and western Canada.
In contrast to the POGO and MAGSAT satellites, the GRM satellite system
will obit at a minimum elevation to provide significantly better resolved
lithospheric magnetic anomalies for more detailed and improved geologic
analysis. In addition,,.GRM will measure corresponding gravity anomalies to
enhance our understanding of `the gravity field for vast regions of the
earth which are largely inaccessible to more conventional surface mapping.
Crustal studies will greatly benefit from the dual data sets as modeling
has shown that lithospheric sources of long-wavelength magnetic anomalies
frequently involve density variations which may produce detectable gravity
anomalies at satellite elevations. Furthermore, GRM will provide an
important replication of lithospheric magnetic anomalies as an aid to
identifying and extracting these anomalies from satellite magnetic
measurements.

After the question of the formation of the earth itself, the most
fundamental problem of the geosciences concerns the origin and
characterization of the continents and oceans. An essential difference in
the earth is between the continents and the oceans which is reflected in
gravity via the Bouguer anomaly. However, as in the case of free-air and
isostatic gravit y anomalies, satellite magnetic measurements indicate no
overwhelming difference between these regions. This is illustrated in
Figure 1 which shows scalar 2°-averaged MAGSAT anomalies differentially
reduced to the radial pole of intensity 60,000 nT at 400 km elevation for
the eastern Pacific Ocean, North and South America, the Atlantic Ocean, and
Euro-Africa. These radially polarized anomalies have been adjusted for
differential inclination, declination and intensitv effects of the
geomagnetic field, so that in principle these anomalies directly reflect
the geometric and magnetic polarization attributes of crustal magnetic
sources. Characteristics of the data given in Figure 1 include the
amplitude ra y (AR), amplitude mean (AM), contour interval (CI), and the
no:malizatioi iplitude (AMP) for the radially polarizing field.

A subtle indication of the difference between oceans and continents is

obtained by computing mean magnetic anomalies for the region shown in
Figurc 1. The analysis indicates that the mean magnetic anomaly of the
continents (0.3 nT) is greater than the average for the oceans (-1.$ nT).
This observation is compatible with the shallow crust of the ocean basins
f
and evidence that suggests that the Moho is a magnetic boundary between

i^
-'n%

-2-

crustal magnetic and upper mantle non-magnetic rocks. However, this result
may be muted by remanence effects which for regional crustal magnetic
sources are generally not well known and at MAGSAT elevations are not well
resolved.

Low-level observations of the oceans indicate that most of the

magnetic anomalies reflect the age of the crust and are caused by the
acquisition of remanent magnetization in the reversing magnetic field as
the rocks pass through their Curie Point. However, as shown in Figure 1,
only the broader scale Mesozoic and Cretaceous quiet zones seem to be
clearly affiliated, respectively, with pronounced negative and positive
radially polarized anomalies at MAGSAT elevations. In contrast, the
satellite elevation magnetic anomalies of the continents seem to have a
predominance of induction effects. There are remanent effects, but these
are characteristically high wavenumber anomalies and are attenuated at
higher elevations in the continental areas. The increased anomaly
resolution derived from GRM's 160 km elevation orbits will significantly
contribute to an understanding of the role which remanence has in causing
regional anomalies of the oceans and continents.

Continental satellite magnetic data show an apparent sharp truncation

and even parallelism of the anomalies along the leading edges of the North
and South American plates, whereas across trailing plate continental
margins prominent anomalies tend to continue into the ocean. Detailed
analysis of the MAGSAT orbital data for South America indicates that the
trailing edge anomalies have no apparent relationship to external fields,
so these anomalies appear to be internally derived. Many of these
anomalies show a striking parallelism with the tracks of hotspots
particularly in the south Atlantic, although there is little consistency in
the sign of the radially polarized anomalies and the hotspot tracks.
Gravity and bathymetric correlations also suggest possible affiliation of
some of these anomalies with subsided continental fragments. Clearly,
information on these features is very limited and their origin is an
important area of inquiry. High resolution magnetic data and correlative
gravity anomalies provided by GRM will significantly facilitate
understdnding their origin and possible role in the evolution and dynamics
of the continents and oceans.

The radially polarized anomalies of Figure 1 permit testing the

reconstruction of the continents prior to the origin of the present day
Atlantic Ocean in the Mesozoic Era. Indeed, as demonstrated in Figure 2,
the radially polarized MAGSAT anomalies of North and South America, Euro-
Africa, India, Australia and Antarctica exhibit remarkably detailed
correlation of regional magnetic crustal sources across rifted margins when
plotted on a reconstruction of Pangea. Obviously, these results suggest
great ages for the geologic conditions which these anomalies desribe and
provide new and fundamental constraints on the geologic evolution of the
continents. The high resolution regional magnetic and correlative b:avity
anomaly data potentially available from the GPM offer the clear promise to
improve quantitative geologic modeling of these features and to detail
their development through geologic time.

Ni

. . . . . . . . . . .
FIGURE 1 co O O O (n
O O O O O O O O
r tt^ l! f 7 N N M 7
I I t I
.09 0 .09
ORIGINAL PAGE" IP o c .e 8-
OF POOR QUALITY .OS .OS
4
o ^J
.0h
q .0h
o e
.OE .oE
4b
P
W
.Oz .0z
H ^ J a O
W
.01
Z ,OI
< C O
11 •0 ^ ,0
V O
Z . 0 1- '1-
ff w C ` v
V 0
<c
f.02- NO- '2-
r- II c e
o
' oe-
^ < u
' oh-
1b' ci
e o
o ./
,OS- e os-
e ,^
y g
i e o
O '09- - 09-
ui ♦ ° 0
N II !I/ O
ac N o .OL- f OL_
O
N
XDo- 09-
cc
^ 906- o 06-
0 V
E
< P ool- 001-
6-
.OTT- oil-
e
I
' o
0zt-
a 0
e
.OEI- OEi-
1 a
.Ohl e Ohl-
O
r-
O O O O O
N
O
—
O O O O O
to In en N R7 ?
1 i 1 1

OWGI AL F AUSE' lS
«+ OF PCOR Q
w c
8 c
w
V1
pow-
u 4. ,\\
c N \\\\ ♦ \♦
Y^ \/
d
0
IN,
t+NW, O u o v+
•r •^ r• `
%
0 N N ^
NuNc `.
A n o \\.
CID \\
►~ 4A u
c a+ .... .
u
CL .h \\\.\ \.\
W
t+
J
\\
\
n
0
N ^ .\
D c
co
{
OC 0
V `
O I .\
J "` \\L
♦V
i ^1I
0
t
r ` \\\
. \ \\\
d N\\ \,
\\..
I\.\ \
H
c
\
O
i A^
1
FIGURE 2
I
n
N
 ....
PIC-frai-. 0 7"0

Cit a tion #14 (Presentation Abs tra ct) 1

N85-31 5 99
CONTINENTAL MAGNETIC ANOMALY CONSTRAINTS ON CONTINENTAL RECONSTRUCTION

R.R.B. von Frese (1), W.J. Hinze (Z), R. Olivier ( ) C.R. Bentley(")

Cruatal magnetic anomalies mapped by the MAGSAT satellite fo g North and

South America, Europe, Africa, India, Australia and Antarctica and adjacent 1.1.

marine areas have been adjusted to a common elevation of 400 km and differer.-

tially reduced to the radial pole of intensity 60,000 nT. These radially

polarized anomalies are normalized far differential in.linatiun, declination

and intensity effects of the geomagnetic field, so that in principle they

directly reflected the geometric and magnetic polarization attributes of

sources which include regional petrologic variations of the crust and upper

mantle, and crustel thickness and thermal petturbations. Continental anoma- i

lies demonstrate remarkably detailed correlation of regional magnetic sources

across rifted margins when plotted on a reconstr uction of Pangea. According-

ly, they suggest further fundamental constraints on the geologic evolution of

the continents and their reconstructions.

T
(1) Dept. of Geology and Mineralogy, Ohio State University, Columbus, OH
4321,)
(2) Dept. of Geostierces, Purdue University, West Lafayette, IN 47907

(3) Institute of Geophysics, University of Lausanne, 5 rue de 1'Universite,

1005 Lausanne, Switerland

(4) Dept. of Geology 3 Geophysics, University of Wisconsin-Madison, Madison,

WI 53706

1
qL

.A"
 Citation #15 (Refereed Paper) N OB 0316 00
REGIONAL MAGNETIC ANIMALY CONSTRAINTS ON CONTINENTAL RIFTING
R. R. B. von Frese l , W. J. Hinze 2 , R. Olivier 3 , and C. R. Bentley4

1 Dept. of Geology b Mineralogy 2Dept. of Geosciences

The Ohio State University Purdue University
Colu • .tbus, OH 43210 W. Lafayette, IN 47907

3 Institut de Geophysique 4 Dept. of Geology b Geophysics

Universite de Lausanne University of Wisconsin Madison
Rue de l'Universite 5 Madison, WI 53706
1005 LAUSANNE, Switzerland

Abstract

Lithospheric magnetic anomalies mapped by the MAGSAT satellite for

North and South America, Europe, Africa, India, Australia, Antarctica, and t

adjacent marine areas have been adjusted to a common elevation of 400 km and

differentially reduced to the radial pole of intensity 60,000 nT. In

principle, these radially polarized anomalies are normalized for differ-

ential inclination, declination and intensity effects of tae geomagneCic

field. Hence, they directly reflect the geometric and magnetic polarization I^I
I
attributes of lithospheric magnetic sources which include regional petro-

logic variations of the crust and upper mantle and crustal thickness and

thermal perturbations. Continental anomalies demonstrate remarkably

detailed correlation of regional lithospheric magnetic sources across rifted

margins when plotted on a reconstruction of Pangea. Accordingly, these

anomalies provide new and fundamental constraints on the geologic evolution

and dynamics of the continents and oceans.

Introduction

Satellite magnetic observations are prnviding new insight into the

structure and petrophysics of the earth's crust. Initial compilations of

1

satellite magnetic anomalies for Lithospheric applications involved scalar

2
a .

magnetometer observations for the conterminous U.S. collected by the Russian

COSMOS 49 satellite at 375 km mean altitude ( Zietz et al., 1970), as well as

quasiglobal 1°-averages of scalar anomaly data derived from NASA's three

Polar Orbiting Geophysical Observatory ( POGO) satellites over orbital eleva-

tions ranging between 400 and 700 km (Regan et al., 1975). For geologic

analysis, however, the best resolved anomaly data set to date has been

obtained by NASA's most recent magnetic satellite mission, MAGSAT, which was

orbited at altitudes ranging from 200 to 550 km (Langel, 1980 b 1982).

Satellite magnetic anomalies have a limited history of geological

interpretation. However, as reviewed by Mayhew et al. (1985), the origins of

these anomalies appear to include regional petrologic variations of the crust

and upper mantle, and crustal thickness and thermal perturbations. Frey et

al. (1983) have noted the geographic coincidence of many satellite magnetic

anomalies with ancient shields and cratons, suggesting that those anomalies

may be of great age and predate the breakup of Pangea. Indeed, they plotted

predominantly total field POGO satellite magnetic anomalies directly on a

Pangea reconstruction to show that prominent anomalies appear to abut each

other along the rifted continental margins.

Previous studies, however, have been limited by the use of variable L-1 1

elevation magnetic anomalies which are also distorted to various degrees of

severity by the attitude and int:nsity variations of the main geomagnetic

field over tha continents in their present configuration. The regional

correlations of lithospheric magnetic sources for South America, Africa and

i
India are particularly problematic when total field magnetic anomalies are

used because these magnetic anomalies exhibit continuously variable

1{
relationships to their sources from the geomagnetic poles to the equator

where inueraion of the anomaly signs occurs. I Y^

`!1.
• . 3 1 '1

To facilitate regional scale correlation analysis, however, total field

magnetic anomalies can be adjusted for variable elevation and differential

inclination, declination and intensity variations of the geomagnetic field

by equivalent point source inversion (von Frese et al., 1981). This

procedure involves least squares matrix inversion of magnetic anomaly data

using an appropriate geomagnetic reference field model to determine magnetic

susceptibilities for a spherically orthogonal array of point dipoles. An

estimate of the magnetic anomalies at a common elevation and reduced to a

radial (normal) pole of constant field strength is obtained by simply

recomputing the equivalent point dipole anomalies at constant e:evation

assuming radial inclination at source end observation points and polarizing

the dipoles by a uniform field strength. Geologic applications which

demonstrate this procedure are given by von Frese et al. (1982a) for North

America, Hinze et al. (1982) for South America and Olivier et al. (1982)

for Europe and Africa.

If remanence can be ignored, then radially polarized anomalies are in

principle centered over their sources and anomaly variations directly

reflect the geometric and magnetic polarization variations of the

lithospheric sources. This has profound implications for continental or

global scale tectonic analyses as geologic source regions may be compared

directly in terms of their radially polarized magnetic anomaly signatures.

The degree to which continental remanent magnetization may affect regional

magnetic anomalies remainE uncertain, as the magnetic properties of the crust

are not well known. However, thermal remanent effects which commonly

complicate the magnetic analysis of upper crustal rocks are

characteristically high wavenumber anomalies that are attenuated at satellite

elevations. Within the lower crust, magnetic susceptibility will be

 4

increased over certain temperature ranges and thermal remanent magnetization

will je decreased by the higher ambient temperatures and increased age of the

rocks (Wasilewski et al., 1979). Furthermore, viscous magnetization directed

along the earth's magnetic field will increase as temperature increases, thus

enhancing total magnetization. In general, limited evidence to date

tentatively suggests that remanent magnetization on a large scale in the

crust .s relatively unimportant (Mayhew, et al., 1985).

To investigate the reconstruction of the continents prior to the origin

of the present day Atlantic Ocean in the Mesozoic Era, 2°-averaged MAGSAT

anomalies produced by Langel et al. (1982) for North and South America,

Euro-Africa, India and Australia, and 3°-averaged anomalies for Antarctica

from Ritzwoller b Bentley (1983) were differentially reduced to the pole at

constant elevation by equivalent point source inversion. Stable point

dipole models were produced which fit the averaged anomalies with negligible

error using the IGS'75 geomagnetic reference field model updated to 1980.

The corresponding MAGSAT anomalies at 400 km elevation differentially

reduced to the pole with a normalization intensity of 60,000 nT are plotted

in Figure 1 on a late Triassic reconstruction of the continents after Smith

et al. (1981).

Pangea and Radially Polarized 2°-AveraRed MAGSAT Anomalies

When normalized by differential reduction to the pole, the MAGSAT

data indicate remarkably detailed continuity of regional magnetic crustal

anomaly sources across the rifted continental margins. The sources of the

magnetic anomalies on the Pangea reconstruction are many and complex, and

only a relative few have been quantitatively investigated using constraining

geologic and ancillary geophysical data. These include aspects of the

prominent positive U.S. transcontinental magnetic anomaly which consists of

1
 S

three peaks centered over the Colorado Plateau, the Anadarko Bssin and the

southern Cincinnati Arch in central Kentucky. The anomaly from the central

peak eastwards corresponds to a well-defined regional trend of free-air

gravity minima. This together with limited seismic evidence suggests that

the transcontinental magnetic feature may characterize a region of thicker

crust (von Frese et al., 1982a).

I
Isotopic evidence indicates that the central and eastern peaks
i
delineate a middle Proterozoic terrane of basement rocks (Van Schmus b
i
Bickford, 1981). To the west the central peak is separated from the !

Colorado Plateau by magnetic minims over the Rio Grande Rift which reflect

thinned magnetic crust due to increased heat flow. To the ea:-. the

transcontinental feature is also breached by minima which have been

1
related by spherical earth modeling to the paleo-rift structure of the

Mississippi Embayment (von Frese, 1982). Here however, crustal seismic

Pn-velocities are anomalously high (Braila et al., 1984), so the

decrease in crustal magnetization is probably related to a lithologic

variation.

The Takatu and Amazon River rift systems of northeastern South America

are also characterized by positive gravity and negative magnetic anomalies

analogous to the regional inverse correlations observed for the Mississippi

Embayment and Rio Grande Rift (Longacre et al., 1982). Synthesis of the

radially polarized MAGSAT data with deep crustal magnetization constraints

indicates tectonic models which relate the negative magnetic anomalies to

a failed-rift component defining nonmagnetic blocks within the lower crust.

The prominent positive Bangui Anomaly of westcentral Africa correlates

with regional heat flow and gravity minima (Olivier et al., 1982) and has

been modeled by Regan b Marsh (1982) as originating from a major intracrustal

1
 i
6
t ^!
lithologic feature. The Bangui Anomaly is bordered to the north by the Benue

Rift aad associated magnetic minima. The anomaly Las a northward projection

to the central Mediterranean Sea and an eastward extension onto Somalia which

is breached by magnetic minima over the East African rift system. To the

west across the Atlantic rift margin, the Bangui Anomaly projects

magnetically as positive anomalies associated with the Sao Luiz Crate,

I fl

northeastern. Brazil and the Central Brazilian Shield. The magne t ic bigntsivre

of the Sao Luiz Craton also co-relates well with the positive a: es of

the Precambrian Shield of southern northwest Africa. The Bangui Anomaly is

bordered to the south by a pronounced magnetic minimum which troughs at its

eastward extension over the Zaire Basin. This feature projects across the

Atlantic rift ma r gin as a comparable magnetic minimum over the Sao Francisco

Craton.

The prominent magnetic minimum over the Cape Orogen in South Africa

roughly corresponds with the East Antarctic minimum over Queen Maud Land and

the minima of southern India and Madagascar. The Antarctic positive anomaly

iarth.r east over Enderby Land is generally consistent with the Indian

positive over the northern shield and Himalayan rocks. The Antarctic high

over Wilkes Land is reflected by a strikingly comparable positive anomaly

a; overlying Archean -Proterozoi: cratonic blocks of southcentral and western

Australia. The prominent positive Australian anomaly is flanked on the west

by a magnetic minimum which follows the western boundary of the Tasman

Orogenic Zone and also includes the Eromanga Basin. This minimum may be

related to an elevated Curie Point isotherm as Sass b Lachenbruch (1978)

identify a high heat flow province here attributable to young magmatic heat

sources within the crust. The magnetic low over the Alelaide and Tasman

Orogens corresponds to an Antarctic minimum over the Ross Sea Embayment and


Transantarctic Mountains. This Antarctic minimum compliments an underlying

anomalously thin crust with high heat flow (Ritzwoller 6 Bentley, 1983) which

may be related to an ancient rift zone (Hayes b Davey, 1975) that probably

was active 65 Ma ago during separation of Antarctica from Australia.

For the most part, the radially polarized anomalies are sufficiently

consistent in amplitude and shape that they may be readily contoured

across or along the rifted continental margins. However, exceptions occur

because the trailing margins have been subjected to intense rifting and

related thermal activity which potentially can lead to anomalies of the

magnitude and scale that could be observed at MAGSAT elevations. Examples

may include the lack of positive anomaly in Africa across the Gulf of Guinea

in western Africa and in northwestern Africa in the vicinity of the Alpine +

age Atlas Mountains. However, the general lack of a significant long- I!'j

wavelength anomaly focused on the continental joins testifies to the lack of

1
broadscale igneous activity on the continental side of the rift which led to

the present oceans.

The close correspondence of most of the radially polarized MAGSAT

51

anomalies across the continental joins verifies the pre-Cretaceous origin of

their magnetic lithospheric sources. However, there are places indicated by

the magnetic data which may involve problems with respect to the continental
I
reconstruction adopted i:i Figure 1. For example, the magnetically positive

source region of Florida could be brought down to the vicinity of the join 4

between northwestern Africa and northeastern South America to better

accommodate the radially polarized anomalies. The worst correspondence of

magnetic anomalies in Figure 1 involves the juxtaposition of the large

1

positive of Madagascar with the pronounced minimum at the join to the

eastceatral African coast. Geologic and bathymethric constraints permit

RI
 b

attaching Madagascar at other points along the east African coast up to

about 15° south of its position in Figure 1 (Powell at al., 1980).

Magnetically, the preferred position is roughly 10° south of its location in

Figure 1 (i.e., close to its present-day position), where the Madagascar

high would correlate with a broad African positive magnetic anomaly

overlying the Cubango Basin and an extensive region of Precambrian Shield

rocks.

Conclusions

Radially polarized KAGSAT anomalies of North and South America, Euro-

Africa, India, Australia and Antarctica demonstrate remarkably detailed

correlation of regional magnetic lithospheric sources across rifted margins

when plotted on a reconstruction of Pangea. These major magnetic features

apparently preserve their integrity until a superimposed metamorphic event

alters the magnitude and pattern of the anomalies. The longevity of

continental scale magnetic anomalies contrasts markedly with that of regional

gravity anomalies which tend to reflect predominantly isostatic adjustments

associated with ne ,3- tectonism. First observed as a result of NASA's magnetic

satellite programs, these anomalies provide new and fundamental constraints

on the geologic evolution and dynamics of the continents and oceans. 1

Accordingly, satellite magnetic. observations provide a further tool for

investigating continental drift to compliment other lines of evidence in

paleoclimatology, paleontology, paleomagnetism, and studies of the

radio=ctric agcs and geometric fi' of the rnnrinonrF.

Acknowledgements

Financial support for this investigation was provided by the Goddard

Space Flight Center under NASA contract NAG5-304.

1
4

,
 ,...._

References Cited

Braile, C. W., W. J. Hinze, R. R. B. von Frese and G. R. Keller, 1984,

Seismic properties of the crust and upper mantle of North America,

Tectonophysics (submitted).

Frey, H., R. A. Langel, G. Mead and K. Brown, 1983, POGO and Pangea,

Tectonophysics, v. 95, p. 181-189.

Hayes, D. E., and F. J. Davey, 1975, A geophysical study of the Ross Sea,

Antarctica, Initial Repts. of the Deep Sea Drilling Project (U. S. Govt.

Printing Off., Wash., P.C.), v. 28, p. 229-240.

Hinze, W. J., R. R. B. von Frese, M. B. Longacre, L. W. Braile and

E. G. Lidiak, 1982, Regional magnetic and gravity anomalies of South

A:^rica, Geophys. Res. Lett., v. 9, p. 314-317.

Langel, R. A., 1980, Near-earth satellite magnetic field measurements: A

prelude to MAGSAT, EOS (Am. Geophys. Union Trans.), v. 60, p. 667-668.

Langel, R. A., 1981 The magnetic earth as seen from MAGSAT, initial

results, Geophys. Res. Lett., v. 9, p. 239-242.

Langel, R. A., J. D. Phillips and R. J. Horner, 1982, Initial scalar magnetic

anomaly map from K AGSAT, Geophys. REs. Lett., v. 9, p. 269-272.

Longacre, M. B., W. J. Hinze and R. R. B. von Frese, 1982, A satellite

magnetic model of nortreastern South American anlacogens, Geophys. Res.

Lett., v. 9, p. 318-321.

Mayhew, M. A., B. D. Johnson and P. .i. Wasilewski, 1985, A review of problems

and progress in studies of satellite magnetic anomalies, J. Geophys. Res.,

v. 90, p. 2511-2522.

Powell, C. McA., B. D. Johnson and J. J. Vicevers, 1980, A revised fit of

East and West Gondwanaland, Tectonophysics, v. 63, p. 13-29.

N
1
 10

:a
Olivier, R., W. J. Hinze and R. R. B. von Frese, 1982, Satellite magnetic

anomalies of Africa and Europe, SEG 52nd Annual International Meeting and

Exposition, Technical Program (Abstracts b Biographies), p. 413-415.

Regan, R. D., and B. D. Marsh, 1982, The Bangui magnetic anomaly: Its

geological origin, J. Geophys. Res., v. 87, p. 1107-1120.

Regan, R. D., J. C. Cain and W. M. Davis, 1975, A global magnetic anomaly

map, J. Geophys. Res., v. 80, p. 794-802.

Ritzvoller, M. H., and C. R. hentley, 1983, Magnetic anomalies over

Antarctica measured from MAGSAT, in (Olivier, R. L., P. R. James and J. B.

Jago, eds.) Antarctic Earth Science-Fourth International Symposium,

Cambridge Univ. Press, N. Y., p. 504-507.

Sass, J. H., and A. H. Lachenbruch, 1978, Thermal regime of the Australian

continental crust, EOS (Am. Geophys. Union Trans.), v. 59, p. 1026.

Smith, A. G., A. M. Hurley and J. C. Briden, 1981, Phanerozoic

Paleocontinental World Maps, Cambridge Univ. Press, N. Y., 102 pp. 1 _^

-. s

Van Schmus, W. R., and M. E. Bickford, 1981, Proterozoic chronology and

evol,tion of the midcontinent region, North America, in (Kroner, A., ed.)

Precambrian Plate Tectonics, Elsevier, N. Y., p. 261-296. ^r

von Frese, R. R. B., 1982, Regional anomalies of the Mississippi River ^ L

aulacogen, SEG 52nd Annual International Meeting and Exposition, Technical

Program (Abstracts 6 Biographies), p. 295-297.

von Frese, R. R. B., W. J. Hinze and L. W. Brai lle, 1981, Spherical earth

gravity and magnetic anomaly analysis by equivalent point source

inversion, Earth Planet. Sci. Lett., v. 53, p. 69-83.

von Frese, R. R. B., W. J. Hinze and L. W. Braile, 1982a, Regional North

American gravity and magnetic anomaly correlations, Geophys. J. R. astr.

Soc., v. 69, p. 745-761.

 i

. • • 11
1^ ryry
I ;i

von Frese, R. R. B., R. Olivier and W. J. Hinze, 1983, Long- w avelength

magnetic and gravity anomcly correlations of Africa and Europe, IAGA

Bulletin, v. 47, p. 140.

WasilewsH , P. J., H. H. Thomas and M. A. Mayhew, 1979, The Moho as a

magnetic boundary, Geophys. Res. Lett., v. 6, p. 541-544.

Zietz, I., G. E. Andreasen and J. C. Cain, 1970, Magnetic anomalies from

satellite magnetometer, J. Geophys. Res., v. 75, p. 4007-4015.

Figure Caption

Figure 1 Pangea and averaged MAGSAT scalar magnetic anomalies reduced to an

elevation ( Z) o f vii- '- cm and to a radial pole strength (AMP) of 60,000 nT.

14
I

^^ _ . ,Rases _ -T •
*,IT
FIGURE 1
ORIGINAL NAM'rS
OF POOR QUALITY `
r^
^ I
vl
r l
w r
^ R^° d s
CL
0
W ° N A
^ N
V
a + L : ^Izz ::: I
V i I y r + V d .^. 7 ^.
+; J :`
C ` ;
zi
W ^ ^
~
r i ^ 1 /) I `r
S
L^^ %
ZM
e
r
0
c 4 / I
s^
1
— "PA, J