NMO Inversion Example


    NMO with long offsets (distance from source to receiver) may cause
some effects that are not obvious.
    Recall that NMO corrects a trace with offset x to a trace with
zero offset.  The amount of NMO is the time difference between the
Tx time and the T0 time for each sample.
    If the trace at offset x is labelled Tx and the zero offset trace
is labeled T0.
    Tx**2 = t0**2 + x**2/v**2

    NMO = Tx - T0

Plot  of shot without nmo. (130KB)

    With an NMO velocity of:
   vtp 1500 3.54 1524 3.629 1546 3.65 1629 3.716 1928 3.87 2108 3.97 end
    SIOSEIS gave a warning:
  ***  WARNING  ***  Data zeroed - NMO caused wavelet inversion.

    Using   Tx = sqrt(t0**2 + x**2/v**2), then for x = 6184;
    vtp 1524 3.629, Tx = 5.44379
        1546 3.650, Tx = 5.415
        1629 3.716, Tx = 5.3122
Tx "normally" increases with an increase in T0.  Inversion occurs when
x**2/v**2 decreases faster than t0**2 increases.   Inversion is more likely
to occur at large offsets when the velocity increases "quickly".

Plot  of shot with nmo inversion not killed. (114KB)

    SIOSEIS normally kills the inverted part.
Plot  of shot with inversion partially muted (stretc 1).(117KB)

    The SIOSEIS NMO parameter stretc is the amount of NMO (Tx-T0) and 
the time sample is zeroed if the NMO exceeds STRETC.

Plot  of shot with nmo .5 sec nmo stretch. (97KB)

    A simpler method of getting rid pf the inverted portion is to mute it.
(mute parameters:    addwb yes xtp 3000 -.2 6000 1.8 )

Plot  of shot with nmo mute. (106KB)

    With an NMO velocity of:
   vtp 1500 3.54 end
     There is no NMO inversion because x**2/v**2 is a constant and Tx 
always increases when T0 increases.

Plot  of shot with constant velocity
(nmo parameters:   stretc 2 VTP 1500 3.54 end