Beruflich Dokumente
Kultur Dokumente
edu
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.
r r (e) Calculate B ( B U ) .
(2) (a) Using cijkl = ij kl + ( ik jl + il jk ) , = u k ,k , ij =
1 (ui , j + u j ,i ) , and 2
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.
1. For (i) SH wave incidence on the solid-solid interface and (ii) P wave incidence on the free surface: (a) Give the potentials, boundary conditions, Zoeppritz equations and all the elements of the scattering matrix. (20 points) (b) In some cases an angle exists where the reflection coefficient is zero; the angle for which that happens is called the intramission angle: compute for the wavespeeds given the intramission angle. (10 points) (c) Make plots of reflection and transmission coefficients as function of incidence angle (from i = 0 o through i = ic to i = 90 o ). (10 points) (d) Explain the plots. (10 points)
SH incidence ( 1 = 2.5 g cm 3 2 = 3g cm 3 1 = 3km s 1 2 = 4km s 1 ) P incidence ( = 5.5km s 1 = 3.2km s 1 )
u1
u 1
1 , 1, 1
i j z
\
Free surface x
,
x i z j
1
\
2 , 2, 2
u2
SH incidence
P incidence
2. Give the Zoeppritz equations of the P-SV system (P incidence) across a solid-solid interface. (20 points)
1 1 , 1 , 1 , 1 2 , 2 , 2 , 2
x
2
\
P-SV system (P incidence) 3. Suppose we have an incidence plane wave S H as depicted in Figure below: prove that the total response has the same amplitude as that of the incidence wave. (30 points)
Free surface
1 , 1, 1
2 , 2, 2
SH
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.
1. For spherically symmetric Earth models, P-SV wave motion separates completely form SH motion. Despite this, P-wave are often observed (weakly) on traverse component (e.g., Figures 4.16 and 4.22). Give several reasons why this might occur. (10 points) 2. The stacked images in Figures 4.20-4.22 appear grainier at source-receiver distances near 00 and 1800 than at 900, owing to a smaller number of seismograms available at these distances. Why might one expect the number of seismograms to diminish at small and long ranges? (10 points) 3. Some of the seismic phases in Figures 4.18-4.23 are plotted with negative slopes, that is, they arrive sooner at long distances. How can this be? (10 points) 4. You will need to run a M atlab script called runrays.m for this problem. Runrays uses the eikonal equation to calculate traveltimes in a layered earth model that you define. The default model provided in runrays.m has 2 discontinuities. The first is at 10 km depth and the second is at 30 km depth. (a) The discontinuities give rise to two major triplications at which distances? (5 points) (b) On the blowup of traveltimes that is plotted (figure 3 that is output by the code) connect the appropriate dots to outline the two triplications. (10 points) (c) Label the traveltime branch that represents the rays that turn above, below and at the first discontinuity. Do the same for the rays that correspond to the second discontinuity. (5 points) (d) Create a velocity model with a low-velocity zone. Set the velocity nodes at
[0 5 10 15 20 25 30 35 40]
and set the velocities to be
[5 8 11 8 11 15 20 25 30] the low-velocity zone gives rise to a shadow zone starting at which distance? Make a comparison between an estimate from figures generated by the code and your calculation. (10 points) 5. Assume that a half-space medium has a linear P wave velocity gradient of the form v p ( z ) = az + b (z>=0), where a and b are constants. (a) Show that rays propagating in this kind of medium have ray paths that are circular. Derive an expression for the radius of the circle in terms of a and the ray parameter p. (10 points) (b) For a source on the surface, derive the travel time T of P wave in terms of a, b, and the epicentral distance X. (15 points) (c) Assume that a=0.001 s-1 and b=3 kms-1. Plot the travel time curve and explain it. (5 points) (d) Give graphical as well as analytical arguments that show the PP midpoint must be a maximum time point. (10 points) (e) (BONUS) For a source at the depth of c, derive the travel time of P and pP in terms of a, b, c and the epicentral distance X. (30 points)
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.
Problem Set 4
1. Receiver Function
A radial receiver function is generated from the P-coda of a single event recorded at a three-component, broadband seismometer located at the surface. The background velocity model used to analyse the receiver function is isotropic and made up of a horizontal layer over a half space. a) Show that for a 1-D, horizontal discontinuity located at a depth h within the top layer, the delay time t = t s t p between the Ps and P arrivals is given by: t = h( s p ) , with the vertical slowness. (10 points) b) Now consider a background velocity model with the following parameters: Thickness(km) Vp(km/s) Vs(km/s)
(g/cm3)
top layer 25 6.0 3.2 2.8 1/2 space -8.0 4.5 3.4 and an incident P wave with a ray parameter of 0.06 s/km. Calculate the delay time t for a Ps wave converted at a horizontal discontinuity located exactly at the interface between the top layer and the half-space (i.e., h=25km). (5 points) c) For the same background velocity model and incident P wave as in (b), what would be the depth of a horizontal discontinuity if the receiver function displayed a Ps phase arriving i) 2s, ii) 6 s after the P wave? (15 points)
2. Surface Waves
1) On a seismogram of dispersed Rayleigh waves, the times in seconds of the peaks and troughs of the waves (starting with a peak) are 30, 86, 134, 174, 200, 226, 244, 262, 276, 290, 302, 312, 320, 330, 338, 346, and 352. The starting time (t=0) of the seismogram is 10h 21m 55s. The station is a distance of 5000 km from the epicenter and the origin time of the earthquake is 10 h 10 m 0s. (a) Reconstruct approximately the seismogram. (5 points) The same earthquake is observed at a second station along a great circle path at a distance of 400 km from the first. The trend of dispersed Rayleigh waves starts from t=0 at 10h 23m 35s. Peaks and troughs are at times in seconds (the first is a peak) of 36, 100, 146, 185, 215, 240, 270, 284, 304, 318, 330, 334, 355, 364, 375, 384, and 394. (b) Reconstruct approximately the seismogram. (5 points) (c) Calculate and draw the dispersion curve of the group velocity U versus the period. (10 points) (d) Calculate and draw the dispersion curve of the phase velocity c versus the period. (10 points) Hints: you can refer to 12.5 section of Principles of Seismology by Agustin Udias for details of measurement of group and phase velocity.
2) For a Rayleigh wave in a half-space, if the P-wave velocity is 6 km/s and Poissons ratio is 1/4, determine for a wave of Period T=20s (a) the depth at which the amplitude of u1 equals to zero (5 points) (b) the depth at which the amplitude of u1 is larger than that of u3, (10 points)
where u1 and u3 are the x, and z displacements respectively.
3. Anisotropy
1) In shear wave splitting studies of upper mantle anisotropy, recordings of SKS or SKKS phases are most commonly used. What are the advantages and disadvantages of using SK(K)S? What are the advantages and disadvantages of using direct S phases? (5 points) 2) Olivine, the principal upper mantle constituent, has an orthorhombic symmetry system. Why, then, are splitting measurements due to mantle anisotropy usually interpreted with hexagonal symmetry in mind? (5 points) 3) The figure below (from Savage, 1999) shows two synthetic seismograms, with corresponding particle motion diagrams, for an incoming shear wave passing through an anisotropic region. Explain why the particle motion diagrams look so different. What is the difference between the anisotropic models used to create the synthetic seismograms? Which situation is more relevant to studies of upper mantle anisotropy using shear wave splitting? (10 points)