4
$\begingroup$

I have generated a synthetic seismogram using FDTD elastic velocity-stress formulation on a staggered grid. I can't understand origin of oscillations modulating on seismogram. enter image description here

Is it a result of dispersion? I have satisfied the grid dispersion criteria (10 grid per minimum wavelength) which keeps it below 5% but still it is very prominent.

Model description: Velocities: Vp = 2000 m/s; Vs = 1500 m/s; Source: Ricker (20Hz). Grid spacing: dx = dv = 1m; time sampling: dt = 0.1msec

Edit 1

I have corrected one problem related to source (now I am using soft source) and it made initial pulse smooth. However its tail still have oscillations. Pressure corrected

Improve this question
$\endgroup$
6
  • 1
    $\begingroup$ 0.1 ms seems like a very short interval, given the other parameters. Does it help at all if you use 1 ms? Make sure you are handling units properly (I find it best to do everything in seconds). $\endgroup$
    – Matt Hall
    Apr 14, 2016 at 23:06
  • $\begingroup$ @kwinkunks I took care of units in first run itself. I carried out experiment with various values of 'dt' but all these oscillations increase with increase in time. I am using a Ricker wavelet of 30Hz frequency. Any suggestions about origin or cause of these oscillations. $\endgroup$
    – Amartya
    Apr 15, 2016 at 17:25
  • $\begingroup$ Do you have any sharp edges that introduce oscillations after transformation? $\endgroup$
    – Way of the Geophysicist
    Apr 21, 2016 at 15:20
  • $\begingroup$ @WayoftheGeophysicist I have taken a homogeneous model so it doesn't have any discontinuity. Additionally, all the edges have absorbing boundary layers so they won't be producing any reflections. $\endgroup$
    – Amartya
    Apr 22, 2016 at 16:39
  • $\begingroup$ Have you considered taking a FFT of your data and identifying the wavelengths corresponding to the oscillation? $\endgroup$
    – BarocliniCplusplus
    Apr 26, 2016 at 1:04

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.