Exam 2 Review Condensed Flashcards
Velocity of p-wave:
Vp=sqrt(k + (4/3)u) / p)
Velocity of s-wave:
Vs=sqrt(u/p)
Seismic impedance equation:
I=pv
reflection coefficient equation:
R=I2-I1 / I2+I1
Relationship between wavelength, frequency, wave speed:
f=w/2pi=1/T=c/lamda
DRAW AND LABEL A TRAVEL TIME CURVE INCLUDING reflected arrival, t0, t, refracted, direct, and slopes for each line
See diagram in notes
Where is the source and receiver for zero offset experiments?
Some location
What is the first potential challenge of the zero offset experiment? Draw a diagram and present a solution.
1) Single records can be noisy: reverberations of energy from near source and near receiver layers can contaminate observations, or, the reflector you wish to image may simply be too weak to generate an observed reflection
Solution: use data from non-zero offset geometries and stack (sum up) the data
What is the second potential challenge of the zero offset experiment? Draw a diagram and present a solution.
2) The time to propagate over layer thickness and back (t1) is less than the time length of source (ts)
Solution: deconvolution: strip the source signal off the seismogram
What is the third potential challenge of the zero offset experiment? Draw a diagram and present a solution.
3) Heterogeneity may scatter energy to different (non-vertical directions, biasing results that assume only vertical propagation paths.
Solution: forward model: do a grid search of all possible sources of scatterers, summing the data for each possibility (migration)
What is the fourth potential challenge of the zero offset experiment? Draw a diagram and present a solution.
4) uncertainties in velocity structure make it difficult to establish depth to reflector
Solution: get help from other means, e.g., drill holes and conduct refraction studies
What are zero offset experiments used for?
Used to map out structure
What are the four steps of the common midpoint method?
1) collect data w/common bounce point location
2) collect t-times, have hyperbolic moveout
3) normal moveout correction: calculate delta(t), move all
4) stack into one signal trace, results in one strong arrival, no noise
What is the point of stacking?
to increase signal to noise ration (where SNR is proportional to sqrt(N) and N is the number of signals in stack)
signal to noise ratio =
amplitude of signal/amplitude of noise
timing correction t(x) is approximately equal to
t0*[1 + 0.5(x/vt0)^2]
Define moveout:
The time difference between arrivals at two different distances
Equation for moveout (delta t):
delta t = x2^2 - x1^2 / 2v^2t0
define normal moveout:
a travel time moveout with respect to the station at a distance x1=0
Equation for normal moveout:
delta(t)NMO = x^2/2v^2t0
Why is stacking useful?
a) to beat down noise that should add incoherently, if random
b) to enhance coherent energy, such as the reflections of interest, which should add coherently
What needs to be muted out before stacking?
A lot of arrivals on the individual seismograms such as surface waves, shallow layer head waves, etc.
If not on flat land for stacking, one must apply…
…elevation corrections, otherwise, you are stacking seismograms that have unaccounted for time shifts in them.
How do you know which arrivals are due to multiples?
Look at intercept times
- the intercept time for the 1st multiple will be 2x the intercept time of the primary arrival
- the intercept time for the 2nd multiple will be 4x the intercept time of the primary arrival
Look at NMO correction as a best fit for the primary arrival
-clean stack at primary, secondary looks noisy & broadened.
What do you need for seismic impedance and reflection coefficient?
Need a physical contrast between layers
R = reflection coefficient = (in words)
amplitude of reflected wave/amplitude of incident wave
The reflection coefficient only works for….
…vertically incident waves
What is the reflection coefficient good for?
We can directly quantify how much energy is reflected vs. transmitted.
Define seismic impedance:
A measure of how sharp the contrast between two layers is and thus how well we can detect a layer boundary using the reflection technique
First case of impedance:
I(layer 2)»_space; I(layer1)
impedance is large, and very little seismic energy will transmit through to the bottom layer
good for detecting layer boundary since most energy will get reflected back
very bad for having any chance of detecting lower layers
Second case of impedance:
I(layer 2) ~> or ~< I(layer 1)
weak reflection from the layer boundary, which may be difficult to detect
Challenges associated with detecting thin layers:
We cant resolve structure that is less than the wavelength of the seismic source
If the structure has a thickness h
How to see a fault in reflection profiles?
Faults seen be offset
What is Mb?
body wave magnitude
What does Mb measure?
measures p-wave amp
What is the period of Mb?
1 s
Why is Mb useful?
Quickest to measure, based on p-waves, first arriving, rapid hazard response
What is ML?
local wave magnitude
What does ML measure?
largest amp on seismogram
What is the period of ML?
Wood anderson central period, 0.8s
Why is ML useful?
Easy to calculate, works well in some environments (S. Cal, UT), accurate for local distances, relatively smaller EQ’s
What is Ms?
Surface wave mag
What does Ms measure?
Rayleigh amp
What is the period of Ms?
20 s
Why is Ms useful?
does a better job estimating magnitude for large mag EQ
What is Mw?
moment wave mag
What does Mw measure?
moment of energy over entire trace, related to Mo
What is the period of Mw?
XXXX
Why is Mw useful?
Takes time to calculate, most accurate estimate
Draw and label each of the beachballs
See notes
What fault is divergent and where is it?
normal, mid ocean ridges
What fault is convergent and where is it?
reverse, subduction zones
What fault is transform and where is it?
strike slip, plate boundaries
What are the steps of t^2 - x^2 technique?
w/ two points calculate slope slope=1/v1^2, v1=1/sqrt(m) t0^2=intercept t0=2h/v1 = sqrt(intercept) h=t0v1/2
two way travel time for vertical incidence t0=
t0=2h/v
why does the hyperbola asymptotically approach direct arrival?
Asymptotically approaches a line with the slope 1/v1, and it does this because increasing the distance between the source and receiver results in not much difference between the direct & reflected wave (draw diagrams that we drew in class review)
