Seismic Attenuation and Earth Structure

Seismic Attenuation and Earth StructureDouglas WiensDepartment of Earth & Planetary SciencesWashington University in Saint Louis

Gung & Romanowicz [2004]Pozgay, Wiens, et al. [2009]1OutlineWhat is seismic attenuation?How is it measured?How is it related to earth properties such as temperature and composition ?Global attenuation structureAttenuation and geodynamic processesSeismic AttenuationIn a perfectly elastic medium, the total energy of the wavefield is conservedSeismic attenuation is the absorption of seismic energy, or the deviation from perfect elasticity

Surface wavesWidmer & Laske [2007]

Coutier & Revenaugh [2006]Body wavesQ Quality FactorAttenuation is quantified by 1/Q, in analogy to the damped harmonic oscillator (underdamped)

Smaller Q results in faster damping (greater deviation from elastic case)Frequency-independent Q damps high frequencies more than low frequencies Q = 2 (total energy/energy lost during one cycle)Shear and Bulk QShear wave attenuation results from relaxation of the shear modulus ()P wave attenuation results from the relaxation of both the shear () and bulk () moduliIn general bulk attenuation is thought to be very small in the earth (Q > 1000)If Q ~ and assuming a Poisson Solid ( = ), QP = 2.25 QSAnelasticity

Absorption Band & Velocity DispersionA single relaxation time gives an absorption peak at = 1/Velocity increases from relaxed to unrelaxed values at about the same frequencyA spectrum of relaxation times superposes these effects

Frequency Dependence of Attenuation

Lekic et al. [2009]Q is observed to be weakly frequency dependent in the seismic bandDescribed as Q = Q0 -Interpreted as a broad spectrum of relaxation timesHow do we measure seismic attenuation?Use many of the same data types as for seismic velocity studiesFree Oscillations (Normal Modes)Surface WavesGlobal and RegionalBody Waves (P & S)Global and RegionalNormal Modes

Different Modes show different rates of amplitude decaySo we can determine a Q for each modeDifferent Qs result from how each mode samples the earthNormal Mode Attenuation Measurements

Surface wave attenuation measurements

Amplitude Decay with DistanceRayleigh Wave Sensitivity vs DepthYang and Forsyth [2008]Dalton and Ekstrom [2006]Differential AttenuationDetermine attenuation of a region by comparing two similar phasesAdvantage is eliminating source, receiver effects

Flanagan & Wiens [1998]Inner Core AttenuationInner core attenuation determined by comparing PKPdf to PKPab and PKPbc

Waveform fromVern CormierRegional Body Wave Attenuation MeasurementsSolve for earthquake corner frequency and attenuation along each pathTomographic inversion of attenuation along raypaths for 3D structure

Low Attenuation PathHigh Attenuation PathSolve for Corner FrequencyAttenuation of Earth Materials: Experiments

Difficult experiments at seismic frequencies and mantle conditionsTorsion apparatus: measures shear modulus as a function of frequency and temperature at mantle pressures Major questions: the effect of melt and water

Figure from Ian Jackson

Possible Attenuation MechanismsAnother Mechanism: Dislocation Damping (Farla et al., 2012)Identification of mechanism is necessary to scale results from lab to earthScaling in grain size, temperature, pressure, etc.17Attenuation vs Frequency, Temperature, Grainsize

Jackson & Faul [2010]Stong grainsize dependence interpreted in terms of grain boundary sliding

Different Extrapolation: Master Curve analysis(McCarthy et al., 2011)

Many studies, materials scaled by Maxwell Frequency fM

= f (T,d)/E(T)E = Youngs (elastic) Modulus = steady-state viscosityLab 5-20 um mantle 1-10 mm about 10^3 larger19Models agree at lab-sample sizes but vary at mantle grain sizes Blue: Jackson&Faul 2010; Red: McCarthy et al. 2011All 2.5 GPa, constant d, dry olivine

Lab:d = 5 mMantle(?):d = 5 mmSlide from Geoff Abers

Melt-bearingMelt-freeExperimental Results: Effect of Melt on AttenuationFaul et al. [2004]21Melt has a large effect on attenuation

Faul et al. [2004]Line thickness gives melt content; line color gives grain size For a given grainsize, 1% melt gives nearly an order of magnitude increase at 1 HzAttenuation from Water in Olivine

Experiments: Aizawa et al [2008] shows effects but no quantitative relationship Karato [2003] extrapolates from the rheological effect MORB source is 0.005 wt % water (810 ppm H/Si) At 100 km depth water < 0.02 wt % [Hirschmann, 2006] Lowers Qs from 80 to 60; 2% decrease in seismic velocity Water dissolved in anhydrous minerals has limited effect on velocity and attenuation at depths < 100 km due to low solubility; water is mostly present as an aqueous meltKarato, 200323Global Attenuation StructureAverage 1-D attenuation model from modes and surface waves

Durek & Ekstrom [1996]Highest Attenuation is in the asthenosphere and inner coreLow attenuation in the lithosphere and lower mantleGlobal Attenuation Structure from Surface Waves

Dalton et al [2008]Attenuation and Velocity Anomalies are Highly Correlated

Dalton et al. [2009]Q modelS Velocity ModelRegional Studies Subduction Zone Magmatism

Joint US-Japan 11 month deployment of 58 OBSs and 20 Land SeismographsPrimary goal was to image mantle wedge processes associated with magmatism

Deployment of US LT OBS from R/V Kaiyo, June, 2003Deployment Map27Attenuation tomography - Mariana Arc and Backarc

Slow velocity high attenuation beneath the arc at 30-100 km depth Sheet-like high attenuation anomaly beneath spreading center75 km wide, extends to 100 km depth Arc and spreading center anomalies separated at shallow depths

Pozgay et al. [2009]28Comparison with Basalt Geobarometry

Si & Mg thermobarometry suggest a final equilibration depth of magma:21-34 km for the backarc spreading center34-87 km for the volcanic arc [Kelley et al., 2010] Corresponds perfectly with the strongest anomalies beneath the backarc and arc Shows that velocity and Q anomalies delineate the melt production region Anomalies result from small fractions of in-situ melt

= maximum anomalyAttenuation structure comparison - Mariana vs Lau (Tonga) backarc spreading centers

Pozgay et al [2009]Roth et al, 1999;Reprocessed by J. Conder Tonga shows much higher attenuation, greater depth extent Consistent with higher temperatures and greater melt productivity Much broader anomaly - passive vs active upwelling at ridge?Modeling Attenuation Structure Tonga/Fiji

Calculated Q model[Faul and Jackson, 2005]Temperature modelQ tomographyWiens, Conder & Faul [2008]31Central America Attenuation

(Rychert et al., 2008 G-Cubed)Broad, high attenuation (low Q) region in wedge. At 1 Hz (assume Q=Qs1 f0.27):Qs1 = 62 84 (Costa Rica)Qs1 = 38-73 (Nicaragua) 32Get PEAK values from Karens emailhigh 1/Q wedge, low-1/Q forearc; much moreso in Nicaragua than CR#S ASSUME FREQ DEP / ARE 1 HZBackarc Melt Porosity inferred from Attenuation

MethodUse Jackson et al [2004] experimental relations Assume 1 cm grain size Try to match maximum attenuation of Qs ~ 20 for Lau and Qs ~ 40 for Mariana Use temperature constraint from basalt geochemistry [Wiens, Kelley & Plank, 2006]

Result Large difference between melt-free and even small melt fraction Large melt fractions predict greater attenuation than observed Constraints are fit with small melt fraction (0.1%) for both Lau and Mariana

ConclusionsAttenuation furnishes an important constraint on material properties in the earth, somewhat independent of seismic velocitiesGlobal surface wave attenuation models correspond closely with shear velocity, suggesting that temperature is the primary controlling factor in bothIn the Mariana islands, the melt source region is imaged at 40-90 km depth beneath the arc and 20-50 km depth beneath the backarc spreading center, in accord with geochemical constraintsAttenuation anomalies between different arcs/backarcs correspond well with temperatures inferred from geochemistry, and suggest melt porosities of less than 0.2 %.