1308 MICROSEISMS 
ances near the surface (but not from explosions) the 
surface waves are usually more prominent than the 
body waves. 
The first attempt to find the type of movement which 
prevails in microseismic waves was made by Geussen- 
hainer [10]. He found that in G6ttingen the period of 
the vertical component of the microseisms normally 
changed gradually with time while in the horizontal 
component periods of 6, 744, and 9 sec occurred more 
often than periods with intermediate values. If waves 
with one of these preferred periods existed in the ver- 
tical component, the corresponding waves had espe- 
cially large amplitudes in the horizontal components; 
if the period of the microseisms in the vertical com- 
ponent was between two fundamental periods of 
the horizontal, both fundamental periods, sometimes 
changing repeatedly from one to the other, were re- 
corded in the horizontal components. Geussenhainer 
believed that the vertical pendulums record mainly 
free vibrations of the earth’s crust. Such free vibra- 
tions would require additional theoretical considera- 
tions since the resulting movements may be very dif- 
ferent from those calculated for forced vibrations, 
especially if resonance phenomena are involved. 
The observed velocities of microseisms are usually 
between about 2 and 4 km sec~; that is, in the range 
of velocities of surface waves with such periods. It is 
generally believed that the microseisms consist of sur- 
face waves. Unfortunately, measurements of the ampli- 
tude of microseisms as a function of depth in order to 
find the expected decrease are inconclusive [15, 19, 20]. 
The few authors who have studied the question of the 
Wave type in microseisms agree that Rayleigh waves 
prevail [16, 22, 32]. On the other hand, waves of the 
Love type are also involved, although their amplitudes 
are relatively smaller. 
Increase in Period with Distance. It has been found 
that the period of elastic waves usually mecreases as 
they are propagated. For example, in Eurasia the 
periods of microseisms frequently lengthen by several 
seconds as the waves are propagated from the Atlantic 
Coast into Asia. This increase is not due necessarily to 
a stronger absorption of the shorter periods but may 
be due to an actual increasing of the wave lengths with 
distance. Among attempts at a theoretical explanation 
are the wavelet theory by Ricker [83, 34], a theory 
developed by Munk [28], and a theory by Sezawa [87] 
(based on the assumption that the increase in period is 
due to internal friction) which has been extended by 
Gutenberg [13] and by Gutenberg and Schlechtweg [17]. 
This theory leads to the following equation for the 
period JT at a distance D from the source: 
(8) 
where 7 = coefficient of imternal friction (about 10° 
poises), p = density, VY = wave velocity, and 7) = 
period near the source. Values of 7 calculated from 
equation (8) agree well with the observed increase in 
period. For microseisms Gutenberg [14] found approxi- 
mately 
T? = T3 + 0.01D, (9) 
where D is expressed in kilometers. The increase in 
period with distance makes it difficult to use micro- 
seisms recorded at a distance D from the source to 
find the period 7) which prevails near the source. 
Effect of Differences in the Structure of the Earth’s Crust 
and of Faults on Microseisms. Microseisms in general 
have wave lengths of less than about 25 km; since their 
energy decreases exponentially with depth, they should 
be propagated mainly within the uppermost 25 km of 
the earth’s crust, and short-period microseisms should 
be propagated in even a thinner layer. (This is the main 
reason for the observed dispersion of surface waves.) 
In geologically disturbed areas the loss of energy due 
to absorption is greater. This was recognized first for 
Europe (Fig. 4), where the regions in which the micro- 
MICROSEISMS 
MAINLY AFFECTED 
BY STORMS NEAR 
@ BAY OF BISCAY 
Fic. 4.—Areas with relatively large microseisms in Europe 
for three different locations of storm centers. (After Gutenberg.) 
seisms are relatively large for a given location of the 
" source agree relatively closely with the geological units 
[15]. For example, microseisms originating m Scandi- 
navia are propagated far into Asia without much loss 
of energy, but their amplitudes decrease considerably 
in southerly directions where surface layers of different 
geological ages and different physical constants are 
passed by the waves. 
In the Pacific, microseisms with relatively large am- 
plitudes are recorded if the station and the source of 
the microseisms are on the same side of the andesite 
line. This line (sometimes called the Marshall line) is 
the surface trace of a discontinuity (considered to ex- 
tend to a depth of at least 30 km) which separates the 
less andesitic material of the upper layers in the bottom 
of the “‘Pacific Basin” from the more andesitic material 
with different physical properties on the continental 
side. In the western Pacific, the andesite line follows 
