Recent reviews on lithospheric studies covering 2005-2009

Chapter 13

Mapping and Interpretation of the LithosphericMagnetic Field

Michael E. Purucker and David A. Clark

Abstract We review some of the controversial andexciting interpretations of the magnetic field of theearth’s lithosphere occurring in the four year periodending with the IAGA meeting in Sopron in 2009.This period corresponds to the end of the Decadeof Geopotential Research, an international effort topromote and coordinate a continuous monitoring ofgeopotential field variability in the near-Earth envi-ronment. One of the products of this effort hasbeen the World Digital Magnetic Anomaly Map,the first edition of which was released in 2007.A second, improved, edition is planned for 2011.Interpretations of the lithospheric magnetic field thatbear on impacts, tectonics, resource exploration, andlower crustal processes are reviewed. Future inter-pretations of the lithospheric field will be enhancedthrough a better understanding of the processesthat create, destroy, and alter magnetic minerals,and via routine measurements of the magnetic fieldgradient.

13.1 Introduction

The magnetic field originating in the earth’s litho-sphere is part of the earth’s magnetic field complex,a dynamic system (Friis-Christensen et al. 2009) dom-inated by the interaction of the earth’s magnetic fielddynamo with that of the sun’s. The lithospheric field

M.E. Purucker (!)Raytheon at Planetary Geodynamics Lab, Goddard Space FlightCenter, Code 698, Greenbelt, MD 20771, USAe-mail: michael.e.purucker@nasa.gov

is dominated by static (on a human time scale) con-tributions that typically represents less than 1% ofthe overall magnitude of the magnetic field complex,and originate from rocks in the crust and locally, theuppermost mantle. Interpretation of the lithosphericmagnetic field is used in (1) structural geology andgeologic mapping, and extrapolation of surface obser-vations of composition and structure, (2) resourceexploration and 3) plate tectonic reconstructions andgeodynamics.

This article is designed as a review describing recentprogress in mapping and interpreting the lithosphericmagnetic field, and also includes some highlights fromthe 2009 IAGA meeting in Sopron, Hungary. SinceIAGA meets every four years, we have designed thisreview to highlight progress in the four year periodfrom 2005 through 2009, although references to ear-lier important works are not neglected, especially inthe area of resource exploration. Several reviews bear-ing on the mapping and interpretation of the litho-spheric magnetic field have appeared between 2005and 2009. Review articles within books and encyclo-pedias have included those within the Encyclopediaof Geomagnetism and Paleomagnetism (Gubbins andHerrero-Bervera 2007) and the Treatise of Geophysics(Schubert 2007). The Encyclopedia included arti-cles on the Crustal Magnetic Field (D. Ravat, pp.140–144), Depth to Curie temperature (M. Rajaram,pp. 157–159), Magnetic anomalies for Geology andResources (C. Reeves and J. Korhonen, pp. 477–481),Magnetic Anomalies, Long Wavelength (M. Purucker,pp. 481–483), Magnetic Anomalies, Marine (J.Heirtzler, pp. 483–485), and Magnetic Anomalies,modeling (J. Arkani-Hamed, pp. 485–490). TheTreatise of Geophysics included articles on ’Crustal

311M. Mandea, M. Korte (eds.), Geomagnetic Observations and Models, IAGA Special Sopron Book Series 5,DOI 10.1007/978-90-481-9858-0_13, © Springer Science+Business Media B.V. 2011

 Thebault (2009) Sopron IAGA Reporter Review

 Purucker and Clark (2010) IAGA Div. 5 Book Chapter – published by Springer

Themes Covered   Regional and global anomaly maps and data sets

for lithospheric fields

  Theoretical anomaly interpretation developments and their applications

  Connecting rock properties with crustal-scale interpretation of magnetic anomalies

  What new have we learned about magnetism of the lithosphere? And what new science could be done using the developments in the last couple of years?

Anomaly maps & Datasets   AWAGS long-line, short-duration aeromagnetic and

radiometric survey (2007)

  AWAGS leveled 5th edition Australian magnetic map (Milligan et al., 2010)

Anomaly maps & Datasets   NURE_NAMAM2008 (Ravat et al., 2009) –

Comprehensive Model CM4 corrected US NURE surveys

Blues and Magentas ±200-300 nT difference between Corrected NURE data set and NAMAM

(2002) – The pattern of differences related to survey boundaries and the

years of surveys (IGRF vs CM4)

Anomaly maps & Datasets   WDMAM 2nd Edition – Available at the Int’l

Geological Congress in Brisbane, August 2012 (Korhonen and the WDMAM group)

6

Anomaly maps & Datasets

  EMAG-2 (Maus et al., 2009) X - 30 MAUS ET AL.:

Total intensity anomaly (nT)

Figure 7. Mercator projection and polar stereographic projections (> 40! latitude) of

the EMAG2 global grid.

D R A F T June 9, 2009, 9:48am D R A F T

Anomaly maps & Datasets   CHAMP MF7 (Maus et al., 2010)

Anomaly maps & Datasets MAUS: DEGREE-720 ELLIPSOIDAL HARMONIC MODEL X - 29

-200

-150

-100

-50

0

50

100

150

200

nT

Bx

By

Bz

Figure 5. Equatorial view in Mollweide projection of the north (Bx), east (By), and

down (Bz) components of the magnetic field at ellipsoid altitude, after applying a Hanning

filter to the ellipsoidal expansion coe!cients.

D R A F T March 25, 2010, 5:05pm D R A F T

Theoretical anomaly interpretation

developments and their applications

  Grauch & Hudson (2007, 2011) …. aeromagnetic expression of faults in sedimentary basins…lessons from the Rio Grande Rift

Sedimentary Magnetic Anomalies

Grauch and Hudson

598 Geosphere, December 2007

Figure 2. Color shaded-relief image of reduced-to-pole (RTP) aeromagnetic data for the Rio Rancho area (Sweeney et al., 2002). The colors primarily refl ect the broad variations in the data, whereas the illumination (from the west) empha-sizes detailed variations, especially linear features associated with faults. Geologic contacts from New Mexico Bureau of Geology and Mineral Resources (2003). QTs—Quaternary and Tertiary sediments (Santa Fe Group and alluvial cover). QTb—Quaternary and Tertiary basaltic and andesitic rocks, undifferentiated. Mz—Mesozoic sedimentary rocks. Aster-isks indicate volcanic vents. See inset for location. Labeled profi les are shown in Figures 6 and 11. Dashed white boxes show areas of Figures 11 and 12.

-2

0

2

500

0 1000 2000Distance (meters)

0

Dep

th (m

eter

s)

500

0

Depth (m

eters)

Mag

netic

fiel

d (n

T)

observation level

0

10RTP

HGM-RTP

HGM-potential

HG

M-R

TP (nT/km)

1 x 10 SI-3

1 x 10 SI-3

-2

0

2

500

0 1000 2000Distance (meters)

0

Dep

th (m

eter

s)

500

0

Depth (m

eters)

Mag

netic

fiel

d (n

T)

observation level

0

10

20RTP

HGM-RTP HGM-potential

HG

M-R

TP (nT/km)

1 x 10 SI-3

-2

0

2

500

0 1000 2000Distance (meters)

0

Dep

th (m

eter

s)

500

0

Depth (m

eters)

Mag

netic

fiel

d (n

T)

observation level

0

20

30

HG

M-R

TP (nT/km)

RTP

HGM-RTP

HGM-potential

-2

0

2

500

0 1000 2000Distance (meters)

0

Dep

th (m

eter

s)

500

0

Depth (m

eters)

Mag

netic

fiel

d (n

T)

observation level

RTP

HGM-RTP0

10

HG

M-R

TP (nT/km)

20HGM-potential

1 x 10 SI-32 x 10 SI-3

2 x 10 SI-3

2 x 10 SI-3

(A) Truncated-layer Model (B) Offset-layers Model

(C) Contrasting-layers Model (D) Thin-thick Layers Model

W E W E

W E W E

Figure 2. Simple 2D geophysical models illustrating the four main types of aeromagnetic signatures associated with intrasedimentary normal faults in the Rio Grande rift. Profiles are computed for the reduced-to-pole (RTP) magnetic anomaly (bold blue lines), the horizontal gradient magnitude of the RTP anomaly and of its potential (HGM-RTP—solid gray lines; HGM-potential—dashed gray lines). The anomalies for the truncated-layer and offset-layers models (A and B) are typical, whereas the contrasting-layers and thin-thick layers models (C and D) are unexpected. Note the multiple peaks in the HGM profiles for C and D, discussed in text. Magnetic susceptibilities are color-coded and labeled.

Sedimentary Thickness (Laske & Masters, 1997) 1° Ave. Tilt-Depths Corr. Coeff. = 0.87

€

θ = tan−1∂M

∂z∂M

∂h

⎛

⎝

⎜ ⎜

⎞

⎠

⎟ ⎟

Tilt-Depth (Salem et al., 2010)

Eastern Officer Basin Drillhole Depths Eastern Officer Basin Tilt-Depths

Variable RTP Australia 4th Ed. Aeromag Deepest Sed. Drilled Depths Tilt-Depths

Tensor Potential Fields

€

θ a = tan−1 aTzz

Tzx2

+ Tzy2

⎡

⎣

⎢ ⎢

⎤

⎦

⎥ ⎥

Inferred Depths of Vinton Salt Dome, Louisiana, USA

Tensor Interpretation Methods   Recognition of dikes, faults, and other thin bodies from Full

Tensor Gradiometer Magnetics (FTGM) system (Fitzgerald et al., 2010)

  IPHT (Institute for Photonic Technology) instrument: 6 cross-line mixed gradients using SQUID

  Decomposition into invariant structural (eigenvalues) and rotational (eigenvectors) part using quaternions

  Robust interpolation of tensor components using spherical linear interpolation of quaternions

Dykes and Full Tensor Magnetic Gradiometry

There are 5 prominent dolerite dykes in this survey area. Using conventional scalar 2D modelling technology on the Tzz component, it appears that all these dykes show an induced response, except for the most Northern one which exhibits evidence of remanence. We have chosen the Southernmost dyke (No 5) as the subject for this development work. Integrating Geology and Geophysics in a Common 3D Model A simplified 3D geology model was built from the surface geology, the digital terrain grid and a section. The shuttle radar (SRTM) terrain is used to provide the surface relief. The observed FTGM signal is compared to the predicted thin-body responses from the model. Faults, dykes and topographic effects associated with the Platinum reef outcrop are the immediate concern. Other magnetic features such as the hornfels, magnetite cumulates and then remanence may follow as the model evolves Discussion In this case it is reasonable to assume that the signal from a fault or thin dyke could be modelled by a thin sheet with no thickness in a geometrical sense but a thickness for the geophysical signature. This simplifies the process of modelling these features. Thicker dykes may require a variable thickness.

Conclusions The methods outlined in this paper have proven to be very useful for signal processing, gridding and interpretation of FTG data. They require the correct treatment of (non-commutating) tensors as compound objects without resorting to processing the components separately. This is a direct result of recognizing the rotational components within the signal. An airborne full tensor magnetic gradiometry survey of the Mogalakwena Platinum mine has been collected and processed using the IPHT system, aided by the Intrepid custom tool. The method of gridding the tensor field is critical in preserving its physical integrity. The amplitude and phase quantities of the tensor grid are directly correlated with geology for those bodies that have a susceptibility contrast. A 3D geology model that captures terrain, faults, dykes and contacts is required to further unravel what new insights the FTGM method has to offer the interpreter. A "Naudy" like procedure has been developed to find by inversion, a set of contacts, dips and thicknesses that define magnetic dykes. An inversion scheme that is formulated to use the tensor invariants as well as the eigenvectors expressed as a rotation matrix (quaternion) is considered the most satisfactory approach once signal compensation is also improved. The work is on-going.

Figure 3: Tensor amplitude product one – cube root invariant. A composite with the geology in the background. Strong signal from the Hornfels bottom right hand side, and the magnetite cumulates in the Upper Zone. All dykes report very clearly.

Figure 4: Tensor phase (0 to 90). The main structural geology posted as vectors. Muted signal from the Hornfels bottom right h and the magnetite cumulates in the Upper Zone. New detail appears here. All the successions in the reef are clearly evident. All dykes report very clearly. Faults could be better mapped with this enhancement. Note: Please ignore line levelling artefacts, this can still be improved.

1165SEG Denver 2010 Annual Meeting© 2010 SEG

Downloaded 11 Jun 2011 to 74.131.181.226. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/

Tensor Amplitude

Tensor Phase

Undercover Lithology Identification   Gettings (2009) Identification of Concealed Lithologies

Using Possibility Theory and Aeromagnetic Data in P. J. Williams et al. (editors) Smart Science for Exploration and Mining

  Pattern Recognition with Possibility Theory

!"#$%!& '"() '*+(,%#! '%"*- './"(!0!"!0,%12' !+ '" '3+4405010!2'

)04!#05/!0+(6 ' ' 78 ' !-% ')04!#05/!0+(4 '"#% ',%#2'40901"#: ' 0! ' 04'

#%"4+("51%'!+'"44%#!'!-%'*+,%#%)'!"#$%!'10!-+1+$2'04'!-"!'+8'

!-%'*"()0)"!%'!%##"(%6'';+)%#('<%+$#"3-0*"1'7(8+#9"!0+('

=24!%9'4+8!>"#%'9"?%4'!-%'4/54%!!0($'+8'$#0))%)'4/#,%2'

)"!"'/40($'"'$%+1+$0*'9"3'1"2%#'%"42'!+'"**+93104-6'@-04'

20%1)4')%#0,"!0,%'$#0)4'!-"!'*"('5%'3#+*%44%)'8+#',"#0+/4'

!%A!/#"1'9%"4/#%46''

=%,%#"1 ' *+9310*"!0+(4 ' "#04% ' 0( '3#"*!0*"1 ' "3310*"!0+('

)/%'!+'!-%'("!/#%'+8'3+!%(!0"1'80%1)46' 'B%*"/4%'9"$(%!0*'

80%1)'0(!%(40!2')%*#%"4%4 '"4' !-%'*/5%'+8' !-%')04!"(*%'!+'

4+/#*%:'0!'04'(%*%44"#2'!+'*+##%*!'!-%'9"$(%!0*'80%1)'+8'!-%'

*"()0)"!%'!%##"(%'!+'!-%'4"9%')04!"(*%'!+'4+/#*%'"4'!-%'

*+,%#%)'!"#$%!6' '7('!-%'*"4%4'4!/)0%)'-%#%:'!-04'-"4'5%%('

"**+93104-%) ' 52 ' /!010C0($ ' $#",0!2 ' "(+9"12 ' )"!":'

4/33+#!%)'52'$%+1+$0*'"()')#011-+1%')"!":'!+'%4!09"!%'!-%'

)%3!-'!+'5%)#+*?'5%(%"!-'!-%'*+,%#6' '@-%'"%#+9"$(%!0*'

)"!"'+,%#'!-%'*"()0)"!%'!%##"(%'04'!-%('/3>"#)'*+(!0(/%)'

52 ' !-% ' )%3!- ' %4!09"!% ' "() ' !-% ' 3+4405010!2 ' 8/(*!0+('

*"1*/1"!%)'8#+9'!-%'/3>"#)'*+(!0(/%)')"!"6 ' 'D'4%*+()'

3#+51%9'!-"!'>011'5%'+5,0+/4'0('!-%'%A"931%'5%1+>'04'!-"!'

3+!%(!0"1'80%1)4'"#%'0(,"#0"(!'!+'!-%'"))0!0+('+8'"'*+(4!"(!:'

0('!-04'*"4%')/%'!+'*-"($%4'0('!-%'#%$0+("1'80%1)6''@-04'04'

%"4012'"**+/(!%)'8+#'52')%!%#90(0($'!-%'"54*044"'+884%!'

!-"!'9"A090C%4'!-%'+,%#1"3'"()'80!'+8'!-%'!>+'3+4405010!2'

8/(*!0+(46

=%,%#"1'!%A!/#"1'9%"4/#%4'*"('5%'*+93/!%)'8#+9'!-%'

"%#+9"$(%!0* ' 80%1)6 ' '@-%4% ' 0(*1/)% ' !-% ' )04!#05/!0+( ' +8'

"(+9"12'"9310!/)%:'!-%'(/95%#'+8'%A!#%9"'3%#'/(0!'"#%":'

!-%'%1+($"!0+('#"!0+:'!-%'4/#8"*%'"#%"'3%#'/(0!'"#%":'"()'

!-%'!#%()')04!#05/!0+(6' '@-%'(/95%#'+8'%A!#%9"'9%"4/#%'

!-%'E(+040(%44F'+8'!-%')"!":'"()'!-%'4/#8"*%'"#%"'9%"4/#%4'

#%1"!0,% ' "9310!/)%46 ' 'B+!- ' "#% ' #+5/4! ' 8#"*!"1 '9%"4/#%4'

!-"! ' -%13 ' *-"#"*!%#0C% ' !%##"(%46 ' ' G1+($"!0+( ' #"!0+ ' 04'

)%!%#90(%)'8#+9'!-%'#"!0+'+8 ' #0)$%4 '"()' !#+/$-4' !+ '"11'

%A!#%9" ' "() ' 04 ' )%!%#90(%) ' 8#+9 ' */#,"!/#% ' "("12404'

HI-011034'%!'"16'JKKL&6' '@#%()')04!#05/!0+('04'"'(+(50"4%)'

9%"4/#%'+8'!#%()4'0('!-%'9"$(%!0*'80%1)'*+93/!%)'8#+9'

!-%'-+#0C+(!"1'$#")0%(!'H<%!!0($4'"()'B/1!9"('JKKM&6''7('

!-04 ' "54!#"*!: ' +(12 ' !>+ ' %A"931%4 ' +8 ' !-% ' 3+4405010!2'

)04!#05/!0+(4 '+8 ' !-%'"(+9"12'"9310!/)% '>011 '5%'4-+>(6'

N!-%#'!%A!/#"1'9%"4/#%4'>+/1)'5%'!#%"!%)'"("1"$+/4126

4 An Example from Southeastern Arizona

O0$/#%'J'4-+>4'!-%'$%+1+$0*'9"3'+8'!-%'I"!"$+(0"';!(46P

="(!"'Q0!"';!(46P@/9"*"*+#0';!(4'"#%"'+8'4+/!-%"4!%#('

D#0C+(":'R=D'H<%!!0($4'JKKJ&:'"('"33#+A09"!%12'SKASK'

?9'"#%"'>-%#%'-0$-'#%4+1/!0+('"%#+9"$(%!0*')"!"'HO0$/#%'

T&'>%#%'+5!"0(%)'>0!-'"'10(%'43"*0($'"()'810$-!'!%##"0('

*1%"#"(*%'+8'JMK96''@-%'*+,%#%)'*%(!#"1'!"#$%!'"#%"'HEU!F'

+('O0$46'J'"()'T&'04'V/4!'>%4!'+8'#+*?4'+8'-0$-'3+#3-2#2'

)%3+40!'3+!%(!0"1'"()'+8'0(!%#%4!'"4'"'3+44051%'%A31+#"!0+('

!"#$%!:'"4'0!'04'+('"'5%)#+*?'-0$-'5%!>%%('!>+'4/55"40(4'

H<%!!0($4'"()'W+/4%#'XYYL&6' '@-%'"#%"'EZ[!F'HO0$46'J'

"() 'T& ' 04 ' " ' !+!"112 ' *+(*%"1%) ' 1"#$% '9"$(%!0* ' "(+9"12'

,04/"112 ' 40901"# ' !+ ' !-% ' "#%"4 ' +8 ' %A3+4%) ' U#%!"*%+/4'

)0+#0!%'H\):'O0$6'J&'"()'"(+!-%#'!"#$%!'+8'0(!%#%4!6' '@-%'

!>+ ' !"#$%! ' "#%"4 '>%#% ' %A!#"*!%) ' "4 ' " ' 4/54%! ' 8#+9 ' !-%'

"%#+9"$(%!0*'4/#,%2')"!"6 ' ']%3!-' !+ '5%)#+*?'8+# 5+!-'*+,%#%)'"#%"4'04'"5+/!'TKK9'H<%!!0($4'"()'W+/4%#'XYYL&'

4+'!-%'"%#+9"$(%!0*'80%1)'+8'O0$6'T'>"4'/3>"#)'

Figure 2. ! <%+1+$0*"1 '9"3 '+8 ' !-% ' 4!/)2 ' "#%" ' 0( ' 4+/!-%"4!'D#0C+("6 ' Z+!% ' !-% ' "#%"4 ' 1"5%11%) ' E@,F ' H@%#!0"#2 ' ,+1*"(0*'

#+*?4&':'E\)F'HU#%!"*%+/4')0+#0!%4&:'"()'!-%'!>+'!"#$%!'"#%"4'+8'

*+,%#'EZ[!F'"()'EU!F6''=*"1%'5"#'04'XK'?9'0('1%($!-6'

Figure 3. !D%#+9"$(%!0*'9"3'+8'!-%'4!/)2'"#%"'0('4+/!-%"4!'D#0C+("6 ' Z+!% ' !-% ' "#%"4 ' 1"5%11%) ' E@,F ' H@%#!0"#2 ' ,+1*"(0*'

#+*?4&':'E\)F'HU#%!"*%+/4')0+#0!%4&:'"()'!-%'!>+'!"#$%!'"#%"4'+8'

*+,%#'EZ[!F'"()'EU!F6''=*"1%'5"#'04'XK'?9'0('1%($!-6'

*+(!0(/%) TKK9 ' 5%8+#% ' %A!#"*!0($ ' " ' 4/54%! ' +8 ' !-%'%A3+4%)'>%4!%#('@%#!0"#2',+1*"(0*'#+*?4'HE@,F:'O0$6'T&'

"()'U#%!"*%+/4')0+#0!%4'HE\)F:'O0$6T&6' 'D'-04!+$#"9'+8'

!-%')04!#05/!0+('+8'"(+9"12'"9310!/)%4'"!'M'(@'0(!%#,"14'

>"4 ' 3#%3"#%) ' 8+# ' %"*- ' +8 ' !-% ' 8+/# ' 4/54%!4: ' "()'

!#"(48+#9%)'!+'"'3+4405010!2')04!#05/!0+(6''O0$/#%'S'4-+>4'

!-"!'!-%'(+#!->%4!'!"#$%!'EZ[!F'04'3#%)+90("(!12:'08'(+!'

!""# $ %&'(!)'*+, $ -.*&.(',/ $ $ $ 01' $ 234 $ 50 $ *66,'( $ .5 $ (1'$

!7,).,,! $ 8!, $ -'('&9.5'- $ 7: $ )1**,.5; $ (1' $ 9.5.9+9$

&'-+)'-$)1.<,=+!&'$7'(8''5$(1'$(8*$-.,(&.7+(.*5,$!,$*5'$

8!, $ 9*>'- $ !)&*,, $ (1' $ *(1'&/ $ $ ?.;+&' $ @ $ ,1*8, $ (1'$

A*,,.7.".(: $-.,(&.7+(.*5, $ 6*& $ (1' $ B0> $!5- $B%(C $ ,+7,'(,/$

01'$7',($6.((.5;$*66,'($.5$(1.,$)!,'$8!,$D@$50/$ $E5$7*(1$

)!,',$(1'$-.,(&.7+(.*5,$7'(8''5$(1'$(!&;'($!5-$)!5-.-!('$

('&&!5' $ !&' $ ,* $ ,.9."!& $ (1!( $ (1' $ ('&&!5' $ .-'5(.(: $ *6 $ (1'$

)*>'&'- $ (!&;'( $ ., $ >.&(+!"": $ )'&(!.5/ $ $ F*(' $ (1' $ "!&;'$

-.66'&'5)',$.5$(1'$A*,,.7.".(:$-.,(&.7+(.*5$,1!A',$7'(8''5$

?.;,/$G$!5-$@/$

Figure 4. !H*".-$".5'$.,$(1'$A*,,.7.".(:$6+5)(.*5$6*&$(1'$(!&;'($!&'! $ BFI(C $ .5 $ (1' $ 6.;+&', $ !7*>'/ $ $J*(('- $ ".5' $ ., $ (1' $ ,!9'$

-.,(&.7+(.*5$,1.6('-$6*&$9!K.9+9$)*&&'"!(.*5$8.(1$(1'$-!,1'-$

".5'#$(1'$A*,,.7.".(:$-.,(&.7+(.*5$6*&$(1'$BL-C$+5.(/

Figure 5. !H*".-$".5'$.,$(1'$A*,,.7.".(:$6+5)(.*5$6*&$(1'$)'5(&!"$(!&;'( $ !&'! $ B%(C $ .5 $ (1' $ 6.;+&', $ !7*>'/ $ $ J!,1'- $ ".5' $ ., $ (1'$

A*,,.7.".(: $-.,(&.7+(.*5 $ 6*& $ (1' $0'&(.!&:$>*")!5.) $ &*)M, $B0>C#$

!5- $ (1' $ -*(('- $ ".5' $ ., $ (1' $ B%(C $ -.,(&.7+(.*5 $ ,1.6('- $ 6*&$

9!K.9+9$)*&&'"!(.*5$8.(1$(1'$B0>C$+5.($-.,(&.7+(.*5/

5 Discussion

N"(1*+;1$(1.,$(',($1!,$8*&M'-$8'""$.5$(1.,$'K!9A"'#$*(1'&$

)!,',$8.""$5*($7'$,*$)"'!&$)+(/$ $E5$9*,($)!,',#$9+"(.A"'$

)&.('&.! $8."" $ 7' $ 5''-'- $ (* $ &!5M $ (1' $ ".M'".1**- $ *6 $ (1'$

('&&!5' $ 7'5'!(1 $ )*>'&/ $ $ ?*& $ 'K!9A"'# $ (1' $ 0'&(.!&:$

>*")!5.)$!&'!$(*$(1'$,*+(1'!,($*6$(1'$(!&;'($B%(C$O?.;/DP$.,$

A&*7!7":$+5-'&"!.5$7:$%&'(!)'*+,$-.*&.('$ .5(&+,.>',# $!,$

8'""$!,$Q+&!,,.)$&'>'&,'":$A*"!&.R'-$;&!5*-.*&.('/$ $S*(1$

&'>'&,'":$!5- $5*&9!"":$A*"!&.R'- $%&'(!)'*+, $!,1 $ 6"*8$

(+66, $ !&' $ !",* $ A&','5( $ .5 $ (1., $ !&'!/ $ $ J'6.5.(.>'$

.-'5(.6.)!(.*5 $ *6 $ ,+7,+&6!)' $ +5.(, $ .5 $ (1., $ )!,' $ 8.""$

A&*7!7": $ 5*( $ 7' $ A*,,.7"'# $ 7+( $ 7: $ )*9A!&.5; $ !5*9!":$

!9A".(+-'#$5+97'&$*6$'K(&'9!#$'"*5;!(.*5$&!(.*#$,+&6!)'$

!&'!#$!5-$(&'5-$ $-.,(&.7+(.*5,#$ $*5'$9!:$7'$!7"'$(*$&!5M$

(1' $ A*,,.7.".(.', $ .5 $ !5 $ *7T')(.>'# $ -'6'5,.7"' $ !5-$

&'A'!(!7"'$8!:/

E6 $ (1'$+5.(,$-'6.5.5;$(1'$('&&!5',$-*$5*($1!>'$"!&;'$

)*5(&!,(,$.5$A1:,.)!"$A&*A'&(.',#$(1'$A*('5(.!"$6.'"-,$8.""$

7'$*6 $ ".9.('-$+(.".(:$ .5$ .-'5(.6:.5;$)*>'&'-$".(1*"*;.',/$

U*8'>'&# $ (1' $ 9'(1*- $ )!5 $ 7' $ !AA".'- $ (* $ *(1'&$

=+!5(.(!(.>'# $ ,'9.<=+!5(.(!(.>'# $ !5- $ =+!".(!(.>' $ -!(!/$

V'*"*;.)!"$!5-$;'*)1'9.)!"$-!(!$)!5$7'$'9A"*:'-$.5$(1'$

,!9' $ 9!55'&# $ !5- $ +"(.9!('": $ (1' $ ',(.9!(.*5 $ *6$

,+7,+&6!)' $ ".(1*"*;: $ ,1*+"- $ .5)"+-' $ !"" $ &'"'>!5( $ -!(!$

"!:'&,/$$01'$6"'K.7.".(:$*6$(1'$"*;.)!"$)*97.5!(.*5,$*6$(1'$

,A!(.!"":$&';.,('&'-$A*,,.7.".(:$6+5)(.*5,$'5!7"',$*5'$(*$

.5)"+-' $ !;'# $ )*9A*,.(.*5!" $ !5- $ *(1'& $ 9*-'" $ &'"!('-$

)*5,(&!.5(, $ .5 $ (1' $ '>!"+!(.*5 $ (* $ !&&.>' $ !( $ !5 $ *A(.9!"$

',(.9!('$*6$(1'$)*5)'!"'-$".(1*"*;:/

Acknowledgements

01., $ 8*&M $ 8!, $ ,+AA*&('- $ 7: $ (1' $ W/H/ $ V'*"*;.)!" $ H+&>':$

A&*T')( $ N,,',,9'5( $ 0')15.=+', $ 6*& $ %*5)'!"'- $ X.5'&!"$

Y',*+&)',/$$

References

S!&-*,,:$V#$?*-*&$Q$OD442P$0&!-.(.*5!"$!5-$F'8$I!:,$(*$U!5-"'$

W5)'&(!.5(:$.5$V'*"*;:Z$F!(+&!"$Y',*+&)',$Y','!&)1$#$24Z2[3\

2][

S*51!9<%!&('&$V?$O233GP$V'*;&!A1.)$.56*&9!(.*5$,:,('9,$6*&$

;'*,).'5(.,(,Z$9*-'"".5;$8.(1$VEHZ$^'&;!9*5#$$_3]$A

%!:&!)$J#$J+7*.,$J#$^&!-'$U$O233`P!U!5-".5;$W5)'&(!.5(:$8.(1$

^*,,.7.".(:$01'*&:$!5-$?+RR:$H'(,$.5$!$H!('"".('$?!+"($

J.!;5*,.,$NAA".)!(.*5Z$Eaaa$0&!5,!)(.*5,$*5$?+RR:$H:,('9,$#$

GO_PZD@2\D`3

J!&7:$Qb$OD44GP$a,(.9!(.5;$0'&&*&.,($Y.,M$8.(1$^*,,.7.".(:$01'*&:Z$

0')15.)!"$Y'A*&($bN<2G2[3c$0YFZ$WHD44@4G#$b*,$N"!9*,$

F!(.*5!"$b!7*&!(*&:

J'9.))*$Y#$L".&$V#$'-,$OD44GP$?+RR:$b*;.)$.5$V'*"*;:Z$a",'>.'&$

N)!-'9.)$^&',,#$_G[A/

J+7*., $J# $^&!-' $U$ O23]]P $^*,,.7.".(: $01'*&:Z $N5 $NAA&*!)1 $ (*$

%*9A+('&.R'-$^&*)',,.5;$*6$W5)'&(!.5(:Z$^"'5+9$^&',,#$D`_A

V'((.5;,$Xa$OD44DP$N5$.5('&A&'(!(.*5$*6$(1'$233`$!'&*9!;5'(.)$

-!(!$6*&$(1'$H!5(!$%&+R$7!,.5#$0+9!)!)*&.$X*+5(!.5,#$H!5(!$

Y.(!$X*+5(!.5,#$!5-$^!(!;*5.!$X*+5(!.5,#$,*+(1<)'5(&!"$

N&.R*5!Z$W/H#$V'*"*;.)!"$H+&>':$dA'5<?."'$Y'A*&($4D<33

V'((.5;,$Xa#$S+"(9!5$XI$O233_P#$e+!5(.6:.5;$6!>*&!7"'5',,$6*&$

*))+&&'5)'$*6$!$9.5'&!"$-'A*,.($(:A'$+,.5;$6+RR:$"*;.)$<$!5$

'K!9A"'$6&*9$N&.R*5!Z$W/H/$V'*"*;.)!"$H+&>':$dA'5<?."'$

Y'A*&($3_<_3D#$2_$A/

V'((.5;,$Xa#$S+"(9!5$XI$OD44@P$N$A&'-.)(.>'$A'5'(&!(.>'$

6&!)(+&'$9!AA.5;$9'(1*-$6&*9$&';.*5!"$A*('5(.!"$6.'"-$!5-$

;'*"*;.)$-!(!,'(,#$,*+(18',($%*"*&!-*$^"!('!+#$W/H/N/Z$a!&(1$

^"!5'(,$HA!)'#$@[$O$]PZ[42<[2@

V'((.5;,$Xa#$U*+,'&$SS$O233[P$S!,.5$;'*"*;:$*6$(1'$+AA'&$H!5(!$

%&+R$f!""':#$^.9!$!5-$H!5(!$%&+R$%*+5(.',#$,*+(1'!,('&5$

N&.R*5!Z$W/H/V'*"*;.)!"$H+&>':$dA'5$?."'$Y'A*&($3[<`[`#$G4A

U*A;**-$N$OD442P$E5('"".;'5($H:,('9,$6*&$a5;.5''&,$!5-$

H).'5(.,(,Z$%Y%$^&',,#$D$'-/#$G`[A

^1."".A,$QJ#$U!5,'5$Yd#$S"!M'":$YQ$OD44[P$$01'$+,'$*6$)+&>!(+&'$

.5$A*('5(.!"<6.'"-$.5('&A&'(!(.*5"!aKA"*&!(.*5$V'*A1:,.),#$_]Z$

222<223

H'-"!M$d#$%.&.)$g$OD44@P$X*-'".5;$J').,.*5$X!M.5;$+5-'&$

W5)'&(!.5(:$!5-$f!;+'5',,Z$E5$^&*)''-.5;,$HEHhD44@#$01.&-$

H'&7.!5$U+5;!&.!5$H:9A*,.+9$*5$E5('"".;'5($H:,('9,#$A22

0+(9'R$S#$0'&)!5$N#$L!:9!M$W$OD44[P$?+RR:$X*-'".5;$6*&$

Y','&>'$a,(.9!(.*5$S!,'-$*5$HA!(.!"$f!&.!7.".(:Z$X!(1'9!(.)!"$

V'*"*;:#$_3Z][\222

Theoretical Anomaly Interpretation Methods

  MagSoundDST (Gerovska et al., 2010)

Anomaly at all locations w.r.t. the central point of the similarity transform (probing point) becomes linear when the probing point and the center of the source coincide

the inversion of the field of a magnetized sphere; it shows that MaG-SoundDST works successfully on magnetic field data, is robust torandom noise, and can correctly estimate source coordinates whenthe source does not coincide with a CPS grid point. The second ex-ample is the inversion of magnetic data from five sources with inter-fering fields; it demonstrates that the technique works in the pres-ence of a linear background, that one solution is obtained per singu-lar point, and that the three newly introduced maps help to relate thesolutions for singular points to real sources. The third simple exam-ple, the inversion of the gravity field of a spherical mass, demon-strates that MaGSoundDST works well with gravity data from sim-ple sources. The fourth example illustrates an application to gravityanomalies of transition type, i.e., from sources characterized by anoninteger negative structural index.

The first example is noise-free magnetic model data of 40!40points, generated by a spherical source !Figure 2a" with center at!x0!5,y0!5,z0!1" km and having an induced magnetizationwith inclination of 45° and declination of 0°. The CPS !a,b,c" set hasa 3D grid spacing of 0.25!0.25!0.25 km. The moving window is21!21 points !5!5 km". The lower half-space was probed withsix points along a vertical line under each window center. The mini-mal Qmin obtained from the four tested structural indices !Figure2c-f" is for N!3 !Qmin!0.00", which corresponds to a sphericalsource with center estimated at !a!5,b!5,c!1" km.

For N!0 !Figure 2c" and N!1 !Figure 2d", no local minima ofQ were detected. Note that for inappropriate values of N, Q!N" may

not have any minima. A tentative value of N!2 resulted in a Qmin of0.44 for a CPS at !a!5,b!5,c!0.6" km. Figure 2b shows thenewly introduced map Q!Qmin", which combines the four 3D mapsQ!N", namely, Q!0", Q!1", Q!2", and Q!3" as described in step 4 ofthe process and illustrated in Figure 1a, correctly estimating thesource position.

To check the performance of the local minimum refinement, wetested the technique on the magnetic field caused by the same spheri-cal source but located at !x0!4.85,y0!5.15,z0!0.85" km, i.e.,the source does not coincide with a point of the CPS grid. If weuse the discrete minima option, the estimated sphere location is!a!4.75,b!5.25,c!0.75" km for !Q!Qmin""min!0.38. If we usethe option for refined minima !step 8", the refined location of thesource is !a!4.85,b!5.14,c!0.85" km for !Q!Qmin""min!0.38,which is virtually coincident with the true source position.

Hypothesizing that each magnetic sensor for the signals A, Ax, Ay,and Az could be affected by independent Gaussian noise, we per-formed a perturbation analysis. We contaminated the four inputchannels A, Ax, Ay, and Az for the analytical spherical magnetic casewith zero mean, Gaussian-distributed random noise in the range10–15-dB S/N in 1-dB increments. We simulated 100 replicates foreach S/N; the perturbations of !Q!Qmin""min and its respective param-eters !a,b,c,N" determined by MaGSoundDST are plotted in Figure3. MaGSoundDST is robust for S/N of 11 dB and higher. The maxi-mum amplitude of the noise for S/N of 11 dB is about 40 nT !Figure3", and the amplitude of the anomaly is about 300 nT !Figure 2a".For modern surveys, the noise level is less than 1 nT, which means

N = 3

j = 3 i = 3

z

5

4

3

2

1

5

4

3

2

x y

1

2

3

4

5 1

54

32

1

5

4

3

2

15

43

21x y

N = 0

j = 3 i = 3

54

32

1

5

4

3

2

15

43

21x y

z

N = 1

j = 3 i = 3

54

32

1

5

4

3

2

15

43

21x y

N = 2

j = 3 i = 3

5 14

32

1

5

4

3

2

15

43

2x y

z

N = 3

j = 3 i = 3

a) b)

Figure 1. Method of implementation. !a" Vertical trace line of the sounding method across the 3D Q matrices associated with all tested structuralindices. The blue spheres are the tentative probe points aligned on the vertical probing track line. The red sphere corresponds to the calculated lo-cal minimum grid point. The Q!Qmin" map takes at point !i,j" the Q at the red sphere location, N!Qmin"!i,j"!3, Z!Qmin"!i,j"!3 length units.!b" Set of grid points used to interpolate the local minimum around the calculated local minimum grid point. The interpolating grid points, shownwith blue spheres, define a cuboctahedron, shown in green.

DST magnetic and gravity sounding L29

Downloaded 08 Jan 2011 to 74.131.181.226. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/

7

6

5

4

3

6

5

4

3

2

1

3 4 5 6 7

Q(Q )min

9

8

7

6

5

4

3

2

1

00 2 4 6 8

y (km)

x(k

m)

A

200

150

100

50

0

–50

–100(nT)

y (km)

x(k

m)

a) b)

0.5

1

1.5

3

45

6

7 34

5

67

7

6

5

4x (km)y (km)

z(k

m)

Q (N = 0)

0.5

1

1.5

3

45

6

7 34

5

67

5

3

x (km)y (km)

z(k

m)

Q (N = 1)

0.5

1

1.5

3

45

6

7 34

5

67

5

3

x (km)y (km)

z(k

m)

Q (N = 2)

0.5

1

1.5

3

45

6

7 34

5

67

x (km)y (km)

z(k

m)

Q (N = 3) 7

5

3

c) d)

e) f)

Figure 2. Model data set caused by a spherical magnetic source. !a"Anomalous magnetic field !T, in nT; !b" Q!Qmin" with minimum correctly lo-cating the source position; !c-f" 3D maps of the estimator Q!a,b,c;N" of DST for structural indices N!0, 1, 2, and 3. The red spheres denote thelocations of local minima of Q and thus of the singular point !center" of the source. The view of the 3D contour maps Q has an azimuth of 48° andelevation of 30°.

L30 Gerovska et al.

Downloaded 08 Jan 2011 to 74.131.181.226. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/

  Cooper (2008) Wavelet-based semblance filtering, Computers & Geosciences — to allow phase comparison of two data sets as a function of both time and wavelength

  Cooper & Cowan (2009) Terracing potential field data, Geophy. Prosp. — Improvement using zero contour of profile curvature

  Cooper (2010) Enhancing ridges in potential field data, Expl. Geophys. — a method of extracting ridges and valleys based on maxima and minima of a balanced plan curvature data set

  Cooper (2010) Enhancing circular features in potential field data, Expl. Geophys. — use of the generalized radial derivative filter

  Cooper & Cowan (2011) A generalized derivative operator for potential field data, Geophys. Prosp. — a new thoeretical derivative operator based on balancing horizontal and vertical derivatives leading to no directional bias

Theoretical Anomaly Interpretation Methods

Tectonic Interpretations

  Integration of Magnetic, Paleomagnetic, Gravity, GPS, LIDAR, Geologic, and Paleo-seismic data to constrain Holocene earthquake activity in Puget Lowland in NW United States (Blakely and co-workers, 2009)

Tectonics & Earthquakes

Blakely et al.

116 Geosphere, April 2009

the Saddle Mountain fault northeastward, well beyond mapped LIDAR scarps.

The Frigid Creek fault (Figs. 2 and 13) is parallel to and 4 km south of the Saddle Mountain fault and exhibits a well-defi ned northwest-side-up scarp 2.7 m in height. A single trench excavated across the Frigid Creek scarp (Fig. 14) revealed conformable strata con-sisting of oxidized sandy gravels, sandy loams, and sandy silts. Radiocarbon analyses by Law-rence Livermore National Laboratory showed that charcoal clasts collected from units 2 and 3 ranged in age from 5657 to 5476 calendar yr B.P. (4850 ± 40 14C yr B.P.) to 4513–4220 cal yr B.P. (3925 ± 40 14C yr B.P.). A single clast of charcoal from unit 6 (part of the surface soil profi le on the scarp) yielded an age of 526–319 cal yr B.P. (415 ± 40 14C yr B.P.). These strata resemble deposits observed throughout the southeast Olympic Mountains that consist of intercalated Holocene debris fl ows and soils.

Beneath the scarp, the strata are offset by a normal master fault and several smaller anti-thetic normal faults, forming a small graben along the scarp. We interpret the offset strata as the result of movement along the normal master fault during an earthquake between 5657 and 319 cal yr B.P. The deformation is best interpreted as downward movement of two hanging-wall blocks (blocks 2 and 3) rela-tive to the footwall block (block 1; Fig. 14B). Downward movement and clockwise rotation of block 2 formed a small graben adjacent to the scarp. Piercing points observed in the excava-tion show that 2.5 m of vertical separation can be accounted for in one event (93% of total 2.7 m scarp height). The unaccounted 0.2 m of scarp height suggests either an earlier and much smaller earthquake, or differential erosion and deposition along the scarp after the earthquake.

We interpret the Frigid Creek fault as a bending-moment fault in the hanging wall of a

large thrust sheet, or as a normal fault associ-ated with a bend or stepover in a lateral fault system. The Frigid Creek fault is astride a high- amplitude, sinuous magnetic anomaly (Fig. 12) that we interpret as a fold in Crescent Formation. A magnetic contact is not observed directly along the Frigid Creek fault scarp, suggesting that the fault is entirely above or roots into underlying Crescent Formation basalts. The Frigid Creek fault, with southeast side down, may be respond-ing to subsidence of the Dewatto and Tacoma basins directly to the east and midway between the Olympia and Seattle uplifts (Fig. 1B).

A narrow but pronounced magnetic anomaly (Fig. 12, label OF, Olympia fault) and coinci-dent gravity anomaly (Fig. 6B) extend south-eastward from Crescent Formation exposures in the Olympic Mountains to the southern edge of our study area, where the source of the anomaly is entirely concealed beneath Pleistocene glacial deposits. South of our study area, anomaly OF

5257000

5257500

5258000

486500 487000 487500 488000

Wes

t fault

East fa

ult

Saddle Mt.

300 m 538025383953868539095398354066541415421054410545445460254722549705527255480

nT

Figure 9. Magnetic anomalies over Price Lake, measured from nonmagnetic canoe. See Figure 2 for map location. Base map is LIDAR (light detection and ranging) image of Figure 2. Black dotted lines show location of canoe transects. Tick marks are Universal Transverse Mercator coordinates in meters. White dashed lines indicate magnetic trough on strike with LIDAR scarps (red lines). Dashed yellow line shows location of cross section shown in Figure 10.

Canoe-borne Magnetic Surveying

Blakely et al.

124 Geosphere, April 2009

Division of Geology and Earth Resources, v. 1, no. 3, p. 1–2.

Carson, R.J., and Wilson, J.R., 1974, Quaternary faulting on Dow Mountain, Mason County: Washington Divi-sion of Geology and Earth Resources Open-File Report 76–2, scale 1:62,500.

Glassley, W., 1974, Geochemistry and tectonics of the Crescent volcanic rocks, Olympic Peninsula, Washington: Geo-logical Society of America Bulletin, v. 85, p. 785–794, doi: 10.1130/0016-7606(1974)85<785:GATOTC>2.0.CO;2.

Haug, B.J., 1998, High-resolution seismic refl ection inter-pretations of the Hood Canal–Discovery Bay fault zone; Puget Sound, Washington [M.S. thesis]: Portland, Oregon, Portland State University, http://nwdata.geol.pdx.edu/Thesis/FullText/1998/Haug/

Haugerud, R.A., and Sherrod, B.L., 2007, Geomorphic evi-dence for Holocene tectonism in western Washing-ton [abs.]: 103rd Annual Meeting of the Cordilleran

Section, Geological Society of America, Belling-ham, Washington, May 2007, Paper 22–3.

Haugerud, R.A., Harding, D.J., Johnson, S.Y., Harless, J.L., Weaver, C.S., and Sherrod, B.L., 2003, High-resolution lidar topography of the Puget Lowland, Washington—A bonanza for earth science: GSA Today, v. 13, no. 6, p. 4–10.

Hirsch, D.M., and Babcock, S., 2006, Unexpected vertical variations in metamorphism within the Crescent basalt, Dosewallips River valley, Olympic Penin-sula, Washington: Geological Society of America Abstracts with Programs, v. 38, no. 5, p. 18.

Hughes, J.F., 2005, Meters of synchronous Holocene slip on two strands of a fault in the western Puget Sound lowland, Washington: Eos (Transactions, American Geophysical Union), Fall Meeting, v. 86, no. 52, abs. S51C-1020.

Johnson, S.Y., Potter, C.J., and Armentrout, J.M., 1994, Origin and evolution of the Seattle fault and Seattle

basin, Washington: Geology, v. 22, p. 71–74, doi: 10.1130/0091-7613(1994)022<0071:OAEOTS>2.3.CO;2.

Johnson, S.Y., Potter, C.J., Armentrout, J.M., Miller, J.J., Finn, C.A., and Weaver, C.S., 1996, The southern Whidbey Island fault, an active structure in the Puget Lowland, Washington: Geological Soci-ety of America Bulletin, v. 108, p. 334–354, doi: 10.1130/0016-7606(1996)108<0334:TSWIFA>2.3.CO;2.

Johnson, S.Y., Blakely, R.J., Stephenson, W.J., Dadisman, S.V., and Fisher, M.A., 2004a, Active shorten-ing of the Cascadia forearc and implications for seismic hazards of the Puget Lowland: Tectonics, v. 23, TC1011, doi: 10.1029/2003TC001507.

Johnson, S.Y., Nelson, A.R., Personius, S.F., Wells, R.E., Kelsey, H.M., Sherrod, B.L., Okumura, K., Koehler, R., III, Witter, R.C., Bradley, L., and Harding, D.J., 2004b, Evidence for late Holocene earthquakes

Figure 16. Tectonic setting of the Saddle Mountain fault. Red and blue stipple indicates Puget Sound uplift and sedimentary basins, respectively, as defi ned by regional gravity anomalies. Red lines are faults of the Saddle Mountain deformation zone. Yellow arrow indicates regional strain direction (McCaffrey et al., 2007). LRF—Leech River fault; RMF—Rattlesnake Mountain fault; WRF—White River fault; SCF—Straight Creek fault; OF—Olympia fault; OU—Olympia uplift; SU—Seattle uplift; KA—Kingston arch. Other labels are explained in Figure 1 caption.

47°

48°

-124° -123° -122° -121°

SWIF

SWIF

WRF

RM

F

SFSF

TFTF

OF

OU

TB

OF

DMFDMF

SC

F

OU

TB

DBDBSUSU

SBSB

KAKA

EBEB

Pacific Ocean

LRF

Olympic

Massif

Olympic

Massif

Canada

USA

FCFSMF

FCF

CRF

Saddle Mt Deform

atio

n Zo

ne

Saddle Mt Deform

atio

n Zo

ne

SMF

40 km

Region of upliftBasin

Tectonics & Lithospheric Temperatures

  Magnetic Bottom / Curie isotherm using fractal magnetization — Bouligand et al. (2009 ) found an analytical expression for Maus et al. (1997) integral

  Spectra are modeled with 3 parameters: Depth-to-Top (Zt), Thickness (ΔZ), Fractal parameter (β) that the integral in expression (3) can be resolved analytically,which leads to the following expression:

FB1D kH! " # C $ 2kHzt $ b $ 1! " ln kH! "

% $kHDz% ln

!!!p

p

G 1% b2

" #cosh kHDz! "

2G

1% b2

$ %$ "

$K1%b2

kHDz! " kHDz

2

$ %1%b2

%!#

!4"

where G(u) is the gamma function and Ka(u) is the modifiedBessel function of the second kind.[9] The final term in brackets in equation (4) vanishes as

Dz becomes large, so that the radial power spectrum for ahalf-space of magnetic sources is given by [Maus andDimri, 1995]

FB1D kH! " # C $ 2kHzt $ b $ 1! " ln kH! " !5"

Thus, the effect of finite Dz is contained entirely in the finalbracketed term of equation (4). This term, which depends onDz and b, significantly influences the shape of the radialpower spectrum at low wave numbers (Figure 2b). Theradial power spectrum of a finite layer is similar to those ofa half-space at high wave numbers, but diverges from thiscurve at low wave numbers. This divergence occurs atlower wave numbers, as the layer thickness increases. Thisreveals that the deeper the bottom of magnetic sources, thelarger the window will need to be in order to recover lowwave numbers in the radial power spectrum and toaccurately estimate the depth to the bottom of magneticsources. In contrast, the depth to the top of magnetic sourceszt influences the slope of the radial power spectrum at highwave numbers (Figure 2a). Specifically, the slope of thespectrum at high wave numbers increases with increasing zt.Finally, the fractal parameter b influences the general shapeof the radial power spectrum over the whole range of wavenumbers. The radial power spectrum displays a peak forsmall b (&2) but not for large b ('3) (Figure 2c). As bincreases, the slope of the radial power spectrum at high

Figure 2. Theoretical radial power spectra (black curves) of magnetic anomalies FB1D(kH) defined byexpression (4) and predicted for different values of the (a) depth to the top zt, (b) thickness Dz, and(c) fractal parameter b of magnetic sources. Values of the two other fixed parameters are zt = 1 km, Dz =20 km, b = 3. The horizontal scale has been set to a logarithmic scale in order to highlight the long-wavelength part of the power spectrum. The plot on the top right corner of Figure 2c shows the shape oftheoretical radial power spectra for a linear horizontal scale. The red curve in Figure 2b is the theoreticalradial power spectrum predicted for a half-space as defined by expression (5). The red curve in Figure 2cis the theoretical radial power spectrum defined by expression (9) and used by early studies [e.g.,Connard et al., 1983; Okubo et al., 1985; Blakely, 1988; Tanaka et al., 1999; Ross et al., 2006].

B11104 BOULIGAND ET AL.: MAPPING CURIE TEMPERATURE DEPTH

4 of 25

B11104

  Bouligand et al. (2009) continued: Based on model studies suggested an approach of automated non-linear fit to the spectra, fixing β to 3, and using variable window sizes

Ravat et al. (2011, in revision, Tectonophysics) - an alternative procedure in Egypt and the Red Sea Spectral Depth Study

Residual L1 norm minimum Residual L2 norm minimum Knee of the minimum of the tradeoff space

Fractal Magnetization based Depth to Bottom ~ 13 km

Centroid based Depth to Bottom ~ 14.8 km Area 2

Top to Bottom Lithospheric Structure Magnetics, Gravity & Seismic Tomography

  Musgrave & Rawlinson (2010) Linking the upper crust to the upper mantle…SE Australia, Expl. Geophys.

P-wave velocity anomalies at 150 km

Purple – geologic; blue – Magn. 1VD Dashed blue – Magn. Tilt, Red – division strong magn granite to W & weak to E

TMI upward continued to 20 km

  Suzanne McEnroe and co-workers have published 22 reviewed articles related to relevance of nm-size lamellar hemo-ilmenite remanent magnetization to crustal anomalies and also the mineralogy of lamellar hemo-ilmenite magnetism

ELEMENTS AUGUST 2009243

CONTINENTAL CRUSTThe continental crust is mineralogically diverse, and rema-nence-dominated anomalies have various origins. Here we focus on a newly discovered source of anomalies related to the hematite–ilmenite series. We discuss the mineralogical background of this series and present three case studies to demonstrate its abundance and magnetic properties.

Hematite–Ilmenite SeriesAll minerals in the hematite–ilmenite series have rhombohedral structure involving layers of octahedrally coordinated cations. Hematite is a canted antiferromagnet (CAF) (see Harrison and Feinberg 2009 this issue) with a room-temperature magnetization of 2.1 kA/m (~0.4% that of magnetite) parallel to the (0001) basal plane. Ilmenite is paramagnetic (PM) above 57 K and therefore carries no magnetization on Earth at geologically relevant temperatures. The complexity of the phase diagram (FIG. 4) results from the interplay between cation ordering, magnetic ordering and exsolution processes.

Exsolution occurs at intermediate to low temperature (Burton 1991; Harrison 2006; Ghiorso and Evans 2008) and leads to the development of fine-scale intergrowths of CAF hematite and PM ilmenite (“hemo-ilmenite”) in slowly cooled rocks. Remarkably, these intergrowths are several times more magnetic than can be explained by the CAF moment of the hematite contained within them. Examination of the intergrowths using scanning electron and transmission elec-tron microscopy (FIG. 5) shows that exsolution has taken place in diffusion-controlled steps. Commonly in the last stages, under slow cooling, final crops of lamellae are produced as thin as 1–3 nm, equivalent in thickness to just 1–2 unit cells. Recent work, described below, has demonstrated that the excess magnetization of these fine-scale intergrowths comes from the phase interfaces between nanoscale CAF hematite and PM ilmenite lamellae.

Lamellar MagnetismSeveral case studies indicate that rocks containing finely exsolved hematite–ilmenite microstructures, with abundant (0001) nanoscale phase interfaces, create large remanence-dominated anomalies (McEnroe and Brown 2000; McEnroe et al. 2001, 2002). Based on these observations, the theory of “lamellar magnetism” or chemical interface magnetization has been developed (Harrison and Becker 2001; Robinson et al. 2002, 2004). This theory claims that mixed-valence contact layers form at the interface between CAF hematite and PM ilmenite to reduce local ionic charge imbalance (Robinson et al. 2006) (FIG. 4). This hypothesis was further confirmed by ab initio calculations (Pentcheva and Nabi 2008). A contact layer carries an average magnetization of ~4.5 µB (where µB is the Bohr magneton, a measure of the electron magnetic dipole moment). The two contact layers on either side of a lamella are magnetized parallel to each other, yielding a total of 2 ! 4.5 ! 9 µB. This moment is counterbalanced by the opposite moment of one hematite layer at ~5 µB, giving a net unbalanced “lamellar” moment of 2 ! 4.5 – 5 ! 4 µB. When the moments of individual lamellae are magnetically aligned with each other, the inten-sity of lamellar magnetism is proportional to the density of lamellae and the surface area of contact layers. Parallel alignment is most strongly favoured in crystals with (0001) parallel to the magnetizing field. Lamellar magnetism is thermally very stable and extremely resistant to demagne-tization in alternating fields commonly 1000 times that of the Earth’s field. This makes it an ideal candidate for the generation of ancient crustal magnetic anomalies.

Exchange Bias Proves Lamellar Magnetism Areas of negative remanent magnetization near Modum, Norway, contain metamorphic titanohematite with abundant nanoscale exsolution of ilmenite ("ilmeno-hematite"). Low-temperature magnetic measurements provide direct proof that the magnetic anomalies result from lamellar magne-

FIGURE 4 Hematite–ilmenite phase diagram at 1 atmosphere showing composition–temperature relations involving

Fe–Ti (long-dashed line) and magnetic (short-dashed line) ordering, miscibility gaps, and two ‘tricritical points’. A tricritical point on the Fe–Ti ordering curve at !57 mol% FeTiO3 and 1050 K creates the two-phase region disordered paramagnetic PM R 3̄c + ordered PM R 3̄. The second tricritical point, on the magnetic ordering curve at !15 mol% FeTiO3 and 815 K, creates the two-phase region CAF R 3̄c + PM R 3̄c. Limbs of the PM R 3̄c region converge at the eutectoid (E, !800 K) and the stable low-T assemblage CAF R3̄c hematite + PM R3̄ ilmenite. For compositions richer in hematite than E, titanohematite can magnetize at T > 800 K. For compositions richer in ilmenite than E, magnetization occurs only at <800 K, with exsolution of CAF hematite (a chemical reaction) from a host richer in ilmenite. Insets illustrate atomic configurations for !80 mol% FeTiO3 above (A) and below (B) the Fe–Ti ordering reaction, and for the fully exsolved case (C) with a CAF hematite lamella (HEM) and two contact layers(CL) in PM ilmenite (ILM).

FIGURE 5 (A) Electron backscatter image of an intergrowth of exsolved hematite and ilmenite, and minor magnetite, from pyroxene granulite, SW Sweden. (B) Electron backscatter image of hemo-ilmenite; hematite lamellae are white and ilmenite gray. (C) High-resolution TEM

image of hemo-ilmenite, COURTESY OF FALKO LANGENHORST

A

A

B

C

B C

  Pilkington and Saltus (2009) - McKenzie River Early Proterozoic Magmatic Arc Anomaly

  Basement depth constrained from seismic data and Euler solutions

  Magnetization: average 2.5A/m (from 15 to 35 km depth)

Author's personal copy

data, to recent (post-2000) digitally-acquired, GPS-located, high-resolution survey data !own under the auspices of the GeologicalSurvey of Canada (GSC). Details on individual surveys are available fromwww.gdcinfo.agg.nrcan.gc.ca. The combined Mackenzie River (MRA,Fig. 2) and Fort Simpson (FSA, Fig. 2) anomalies dominate the magnetic"eld of the studyarea. Previous studies have included theMRA as part ofthe Fort Simpson anomaly trend (e.g., Hoffman, 1989; Ross et al., 1991).Here, we distinguish the MRA as the northernmost part (from 65°N) ofthe FSA trend to simplify the following discussion. This does not implyany separation or differences in the two anomaly trends in terms oftheir source or origin. South of 65°N, the FSA is generally a northerly-trending, !100-km wide, positive magnetic anomaly with amplitudes

of up to several hundred nanoteslas (nT). Superimposed on this longer-wavelength anomaly are short-wavelength (tens of km), northerly-trending, linear features,most likely due to sources shallower than thoseproducing themain FSA feature. North of 65°N, the FSA turns fromaN–Sto an E–Worientation,mimicking Cordilleran trends to the southwest. Itincreases in width up to 250 km and appears to continue to the TintinaFault (Fig. 1; Cook et al., 1992). Within the broad positive anomaly ofthe MRA, sinuous, long-wavelength (50–100 km), E–W striking trendsoccur to the south. To the north, these trends are replaced by similarwavelength, ovoid-shaped anomalies with no single preferred orienta-tion. Shortwavelength features (b20km) are superimposed on theMRAat the northern edge of the Cordillera (65°N, 138°W) and the Colville

Fig. 1. General geology of northern Yukon and western Northwest Territories (after Aspler et al., 2003).

Fig. 2. Low-altitude (300 m) Geological Survey of Canada (GSC) aeromagnetic data. White boxes delineate areas used in power spectral analysis. Numbers in each box indicateestimates of the average depth (km) to the tops of themagnetic sources in that area.MRA=Mackenzie Rivermagnetic anomaly. FSA= Fort Simpsonmagnetic anomaly. TF= TintinaFault. CDF = Cordilleran deformation front.

79M. Pilkington, R.W. Saltus / Tectonophysics 478 (2009) 78–86

Author's personal copy

and to minimize the contribution of shallower sources. Spectral depthestimates aremore reliable where the !eld due to a single ensemble ofsources located at approximately the same depth can be isolated. Thespectral method also assumes that the source magnetizations arespatially uncorrelated, and as such, the estimated depth provides anupper (deeper) bound.

More detailed spatial information of the distribution of magneticsource depths can be gained from automated interpretation methodssuch as Euler deconvolution (Reid et al., 1990). This approach respondsprimarily to the gradients of the magnetic !eld which are generallycaused by lateral changes in magnetization or, equivalently, magneticsource edges, assuming homogeneously magnetized volumes. Themethod also requires speci!cation of the structural index (SI), whichtunes the algorithm to a particular source geometry. An SI of zerocorresponds to a geologic contact, whereas a value of unity is appro-

priate for dyke-like bodies. We assign a value of zero which permitsconsistent evaluation of magnetic contacts within regional sourcesthat have diverse shapes. Furthermore, setting SI to zero produces con-servative (minimum) depth solutions.

Fig. 5 shows the estimated depth solutions for the study area basedon the low-altitude GSC data. The higher-altitude EPB data have large"ightline (!35 km) and along-line data spacings (!3.5 km) whichseverely limits their resolving power. The GSC surveys in the studyarea have line spacings ranging from 400 to 3000 mwith the majorityalong the MRA having an 800 m spacing. Unfortunately, the GSC datahave only limited coverage leading to some data gaps, which mostlyoccur over the Cordillera, to the south of the MRA. Other gaps in theEuler solution plot coincide with either nonmagnetic areas betweensources or the interiors of homogeneously magnetized bodies. Thevertical spread of Euler depths at a given body edge arises from

Fig. 5. Euler deconvolution depth estimates from aeromagnetic data in Fig. 2. MRA=Mackenzie River magnetic anomaly. FSA= Fort Simpson magnetic anomaly. TF = Tintina Fault.CDF = Cordilleran deformation front.

Fig. 4. Satellite magnetic data from the CHAMP mission (Maus et al., 2002). Data were collected at an average 400 km elevation. Dashed line shows study area. VI = Victoria Island.MRA = Mackenzie River magnetic anomaly. FSA = Fort Simpson magnetic anomaly. CDF = Cordilleran deformation front.

81M. Pilkington, R.W. Saltus / Tectonophysics 478 (2009) 78–86

Aeromagnetic

CHAMP Satellite - MF1

  Rajagopalan, Schmidt, and Clark (2010) Magnetic overprinting of the Brachina Formation/Ulupa Siltstone, Southern Adelaide Foldbelt, prior to Delamerian deformation, Aus. J. Earth Sci.

  Schmidt and Clark (2011) Magnetic characteristics of the Hiltaba Suite Granitoids and Volcanics: Late Devonian overprinting and related thermal history of the Gawler Craton, Aus. J. Earth Sci.

  Greenfield et al. (2011) The Mount Wright Arc: A Cambrian subduction system developed on the continental margin of East Gondwana, Koonenberry Belt, eastern Australia, Gondwana Research.

  Dunlop et al. (2010) Magnetic properties of rocks of the Kapuskasing uplift (Ontario, Canada) and origin of long-wavelength magnetic anomalies, Geophys. J. Int.

  210 samples of anorthosite (NRM 0.001-0.3 A/m), tonalite (bimodal: strong ones NRM & IM 0.1–5 A/m), mafic gneiss (NRM 0.01-2 A/m and IM 0.01-0.6 A/m – resistant to thermal & AF demagnetization)

  Thermoviscous magnetization acquired during Brunhes chron (since the past geomagnetic reversal)

  Remanence - a larger role for mid-crustal sources where single-domain grains are well below their blocking temperatures

What new have we learned in the lithospheric studies in the last couple of years?

  Rock Magnetism

  Mineralogy of lamellar hemo-ilmenite magnetism and further relevance of hemo-ilmenite magnetism to intense remanent crustal magnetic features on Earth and Mars

  Crustal Scale Studies

  The first magnetic anomaly based evidence for “The Moho as a magnetic boundary” concept (except where the mantle may be serpentinized)

  No “missing” magnetization in the areas of uplifted continental lower crustal cross-sections

What new science could we do based on developments in the last couple of years?

  From High Resolution Datasets

  Three dimensional near surface earth structure

  Australian AWAGS-leveled 80 m grid is roughly similar in resolution to satellite-borne multispectral and InSAR-type pixels and can now be integrated for extending depth interpretation of satellite imagery and also LIDAR data

  Blakely and co-workers (2009, 2010, 2011) Integration and detailed analyses of geology, paleoseismic, GPS, LIDAR, etc. to the high resolution magnetic surveys will lead to better earthquake hazard evaluation