Paleomagnetism• The magnetic axis wobblesaround the Earth’s rotationaxis• 2500-8000 years to completea loop (at current rates)• When averaged over 10,000years or more, the directionof the magnetic fieldcoincides with the true north= paleomagnetic polePaleomagnetism• Suppose we are able to measurepaleomagnetic direction in asuccession of lava flows (e.g.Deccan Traps, India)• Each lava flow cooled down in afew years and recorded themagnetic field of the time• Each lava flow can be dated• One can reconstruct the history ofpaleomagnetic directions• Using tan I = 2 tan λ, one canreconstruct the history of paleo-latitudesPaleomagnetism• In the field:– Oriented sample (short core or blockof rock): azimuth and dip (using a suncompass…)– Geometry of the layer (for fold tests)– Several sample at each site (5-10)within a few meters• In the lab:– Spin sample => produce current in acoil– Reproduce in 3 directions => get 3Dmagnetic vectorPaleomagnetism• Declination => direction ofnorth pole• Inclination => distance to thenorth pole• Example of a rock now at 10oN:– Inclination = 49o => paleo-latitude = 30o (= pole was 60onorth of the rock)– Declination = 20o• Apparent pole (= relative to therock) can be reconstructed =virtual geomagnetic pole(VGP)Paleomagnetism• Inclination:– How far away the pole was located= angular distance p– p = radius of a small circle centeredon the sampling site (latitude λs,longitude φs)– Small circle = locus of all possibleVGPs• Declination– Defines a meridian (= great circle)passing through the sampling site– Angle D with the NS meridian– latitude λp, longitude φp, of the polecan be calculated fromtrigonometric formulassin !p = sin !s cos p + cos !s sin p cos D"p = "s + #, for cos p sin !s sin !p "p = "s + 180 $ #, for cos p < sin !s sin !p with sin !p = sin !s cos p + cos !s sin p cos D"p = "s + #, for cos p sin !s sin !p "p = "s + 180 $ #, for cos p < sin !s sin !p with sinsin sincos!"=p DpPaleomagnetism• Paleomagnetic poles forPlio-Pleistocene andPermian rocks:– Clustered => axialgeocentric dipole hypothesisok– Permian pole 45o from PPpole => something hasmoved…• Geographic pole?• Continent?Paleomagnetism• Trace of the paleo-apparent polesthrough time for a given location= apparent polar wander(APW) path• It is NOT the motion of theEarth’s pole, but the position ofthe continent relative to the pole• As if the paleomagnetic pole hadmoved slowly along this pathtoward the present pole• APW path can be determined foreach continentThe blue line shows the apparent polar wander pathfor Australia from the mid- to the late Paleozoicperiod.Paleomagnetism• APW paths for Europe and North America, Ordovician to Jurassic• As measured: different APW paths, but similar shapes => continentsmove, paleomagnetic poles don’t• After closing the Atlantic (rigid rotation on a sphere by 38o clockwiseabout a pole at 88.5oN/27.7oE):– Coincide until upper Triassic– Then: relative motion of the 2 continentsPaleomagnetism• What can we learn fromapparent paleomagneticpoles?– If different from present => therock has moved– Declination not due to north =>the rock has rotated– Inclination different frompresent => the rock has movednorth or south, or has been tilted• What is not possible to learn?– Changes in longitudeMean paleolatitude for the Caribbean plate (from ODP cores)Paleomagnetism• Effects of large scaletectonics• Late Paleozoic to LateCretaceous declinations:– Large difference betweenEurope and Africa =>relative rotation– Similar declinations forcentral and southern Europe– Slight difference in Corsica-Sardinia…Paleomagnetism• 19th century: E. Seuss => great LatePaleozoic super-continent,Gondwanaland• 1912: Wegener– All the present continents laid closetogether during the Late Paleozoic =Pangea– Paleoclimatic, paleoenvironments,paleontology, tectonic evidence– Continental drift: “dispersion” ofPangea– No mechanism, no direct evidence• 1950’s: development ofpaleomagnetism– Irving: Paleomagnetic poles of southerncontinents:• Incompatible with their present-dayarrangement• Agree better when the continents arerearranged in a Gondwanaland cluster• Using APW path for each continent,one can reconstruct their relativemotionshttp://jan.ucc.nau.edu/~rcb7/paleogeographic.htmlhttp://jan.ucc.nau.edu/~rcb7/paleogeographic.htmlPaleomagnetism• David and Bruhnes, 1904-1906:– Clays baked in lavas, MassifCentral, France– Same direction of TRM as lava– Changes in magnetizationdirection => polarity change• Matyuama, 1929:– Quaternary lava = normal polarity– Pleistocene lava = opposite polarity– Miocene = 3 different polarities– Link between polarity and ageRecord of a reverse-to-normal MiocenePolarity transition, OregonPaleomagnetism• Secular variations of the Earth’smagnetic field– Slow and irregular variations– When average over several 1000 years =>coincides with axis of rotation• Polarity reversals– Inversion of the Earth’s magnetic field– Occurred at regualr intervals– Inversion occurs in a few 1000 years• Excursions– Magnetic pole moves far from thegeographic pole– Returns within a few 1000 years• Definitions: chrons, sub-chrons, excursions• MagnetostratygraphyPaleomagnetism• 1950’s: improvement of radiometric datingtechniques => correlation between polarity and age isglobal• Magnetic time scale:– Established progressively as number and distribution ofsamples improve– Epochs named after investigators or locations– Last 5 million years particularly well documentedPaleomagnetism• Since the 1960’s, additional datafrom deep-sea sediments:– Quiet sedimentary environment– Rather uniform sedimentation rates– Continuous sedimentary series– Dating: microfossils– Reversal occur at different depthsbecause of different sedimentationrates– Perfect correlation with on-landstudies in lavas• Since magnetizatin process isdifferent (TRM vs. DRM) => self-reversal unlikely• Same pattern of reversals foundeverywhere => global phenomenonGeomagnetic time scale• Based on compilation of magneticanomalies in all major oceans– Cover the last 155-160 Ma– Older oceanic crust ~180 Ma, but noclear anomaly• 1 sequence of polarity reversals:– Current chron = Bruhnes (=C1N)– Chrons C1N to C33R• 1 long interval of constantpolarity = Cretaceous