Rit (Vísindafélag Íslendinga) - 01.06.1971, Blaðsíða 21

CRUSTAL STRUCTURE OF ICELAND
21
Variations due to shot point variability, seismometer-ground
coupling, and attenuation have then not been considered. The first
two are of a rather random nature and introduce noise into the re-
sults. The attenuation can he expected to affect the amplitudes in a
systematic way, and the reduced amplitudes (4.3.2) should provide
an indication of the amount of attenuation taking place.
For a thick refractor the reduced amplitude Ar obeys the equation:
A,. = Ar • e-ftL (4.3.3)
1 1 o
where /3 is an attenuation coefficient.
The reduced amplitudes Ar have, in the cases where amplitude
studies have been made, been plotted against the distance r. The
amplitudes have been given in decibels according to the formula
A(dB) = 20.1ogl o(^-) (4.3.4)
The reference value Aro has been taken as unity.
The attenuation coefficient may be obtained from the A(dB) vs. r
diagram:
dA(dB) dA(dB)
dr dL
—20 • /3 • log10 e
(4.3.5)
The attenuation may also be given directly by
dA(dB)
dr
in decibel/
km, or alternatively in decibel/wavelength.
In a few cases amplitude studies have been made of reflected
Waves from the boundary between layers 3 and 4. This has been
done mainly to assist in the identification of these waves. Only
corrections for geometrical spreading and variations in charge size
have then been made, as the effects of transmission coefficients and
the free surface can be considered to be constant over the range of
distances considered. The amplitudes thus reduced contain the re-
flection coefficient of the reflecting boundary and would therefore
be expected to show an increase in strength beyond the distance cor-
responding to the critical angle, if the arrivals are correctly identi-
fied. An effect of a possible nonisotropic radiation pattern would also
remain, if such a pattern exists.
When the amplitudes have heen reduced as described above for
the head waves and the reflected waves, there remains a considerable