Roll-Along Format for 2D Velocity (Vs) Cross Section
To generate a 2-D shear-velocity (Vs) cross section, MASW analysis requires preparation of a “roll-along”
type data set at the beginning of the analysis. A roll-along data set has multiple seismic records that have the
same source offset (X1) and receiver-array length (L), both of which should be optimally set for a planned
investigation depth (Zmax). A data set of non-roll-along format or a roll-along format with X1 and L not
optimally set can result in dispersion images adversely influenced by the near- and far-field effects of surface
waves. This, in consequence, can reduce the accuracy of the overall data analysis and the final result of 2D
velocity (Vs) cross section can become unreliable.
A roll-along data set can be prepared in two different ways; (1) data acquisition by using a land streamer and
(2) recompilation of a data set acquired by using the conventional fixed receiver array(s) with moving shot
points (often called "shoot through" approach). The former approach requires an additional field apparatus
(land streamer) to be procured and assembled before the field operation. Nonetheless, it can increase the
overall speed of field operation by an order of magnitude, especially for a large-scale project that requires a
collection of many field files (e.g., > 100). However, because of the indirect receiver coupling, the signal-to-
noise (SN) ratio of acquired data may decrease. On the other hand, the recompilation approach is a post-
acquisition process that selectively chooses for subsequent analysis those traces (channels) falling within an
optimum offset range determined by a proper combination of X1 and L. Both approaches are further
Data Acquisition with Land Streamer
The easiest way to collect a roll-along data set is to use a land streamer. A land streamer is a platform on
which geophones are attached without spikes so that all of them can be moved as one unit, facilitating the
mobility of receiver array (Figure 1).
Data Acquisition with "Shoot-Through" Approach
It is demonstrated here to use an ordinary receiver
array (of 24-channel with 5-ft receiver spacing; i.e., of
12-channel acquisition equivalent (Figure 2).
The receiver spacing (dx) is chosen in such a way that
the total length of the roll-along receiver array (i.e., 12-
channel array) is about twice as long as the maximum
investigation depth (Zmax); i.e., 11dx ≥ 2Zmax. Thus,
with dx=5ft, it is assumed that Zmax ~ 30 ft. The
receiver spacing will have to be accordingly changed
as Zmax changes. For example, to achieve Zmax ~ 90
ft, a receiver spacing three times longer (i.e., dx=15 ft)
is required. If the receiver spacing is not to be
changed, then the number of channels used for the
fixed array should be accordingly increased to extend
the array length of the roll-along format (e.g., from 12-
channel to 36-channel acquisition).
First, choose a source offset (X1), which is about 50%
of the receiver array length (L) for the roll-along data
to be prepared. Here, X1 of 6dx (i.e., 30-ft) is chosen.
X1 is usually chosen as a multiple of dx. Then, start
shooting (i.e., impacting) at X1 distance ahead of the
first channel (ch-1), and then move the shot point by
1dx each time for twelve times. The source point can
be slightly off the geophones when it reaches inside
the receiver array; for example, it can be ± 20-30% off
either inline or offline. This will generate a data set of
13 field records (files) of 24-channel acquisition that
will be used to generate a roll-along data set consisting
of 13 records of 12-channel acquisition, from which a
final 2-D Vs cross section of 60-ft (i.e., 12dx) lateral
distance can be obtained through the normal MASW
analysis. This approach is graphically illustrated in
Figure 2 with source/receiver (SR) configurations
specified in Table 1.
In Figure 2, a 24-channel receiver array is placed on
the ground with a receiver spacing (dx) of 5-ft, and
then the first shot point located at X1 (6dx) ahead of
the first channel (ch-1) generates a 24-channel field
record named “1001.dat”, and then the shot point
moves by 1dx to generate another (24-channel) file of
“1002.dat”, and so on (see Table 1). The
configuration shows total of 13 such files are collected
by moving shot points twelve times by 1dx each time
toward the receiver array. Actual impact points for
those intra-array shots can be slightly off the receiver
point as previously mentioned. But, they can be
assumed at the correct (geophone) locations during
the source/receiver (SR) setup without any reduction of
Preparation of Roll-Along Format
(X1= 6dx and 12-channel Acquisition)
The previous 24-channel fixed-array data set of total
13 files is now rearranged to generate the following roll-
along data set of X1=6dx (30-ft) and 12-channel
acquisition by selectively discarding front and back
channels’ data (Figure 3). This process is usually
automatically performed by the analysis software (e.g.,
ParkSEIS) during the recompilation process. The
horizontal distance of the final velocity (Vs) cross
section covered by this data set is indicated by the
The cross section in Figure 4 is obtained by following
the data-acquisition mode previously explained.
Originally, 13 field records of 24-channel acquisition
were collected and then reconfigured to generate 13
roll-along records of 12-channel acquisition. This roll-
along data set was used to generate this cross section
by following the normal MASW analysis procedure.
Extension of Survey Length with Shoot-Through
When a longer lateral dimension is needed for the
cross section than the distance enabled by one fixed
array previously illustrated, one can move the front 12
receivers to the other end of the original array after
collecting the 12-th field file. At that time, the channel
numbers will have to be reconfigured so that the 13-th
geophone now becomes the 1st channel for the
reconfigured receiver array. In this way, one can
continue to move the front 12 geophones as many
times as needed to extend the survey line length to
meet the whatever length required. This approach is
graphically illustrated in Figure 5 with source/receiver
(SR) configurations specified in Table 2.
Figure 5 illustrates how to extend the lateral distance
for the final 2D velocity (Vs) cross section by using
“fixed” receiver arrays at multiple locations. It aims at
generating a 12-channel roll-along data set to
generate the final 2D velocity (Vs) cross section by
using a 24-channel acquisition mode as previously
illustrated. After collecting twelve (12) field records by
continuously moving shot points for a given “fixed”
array (e.g., #1 array), the first twelve (12) geophones
will have to be moved to the other end of the previous
array. Then, all (24) geophones are reconnected to
make a new “fixed” array at a different location (e.g.,
#2 array). This procedure can be repeated as many
times as needed to extend the lateral survey
The chart in Figure 6 shows the source/receiver (SR)
configuration for the 12-channel roll-along data set that
would be obtained from the 24-channel data set by
following the above approach to extend the survey
distance. The extended lateral distance of the final 2D
velocity (Vs) cross section is indicated on the chart by
a yellow box.
Figure 1. A land streamer.
Figure 2. Chart of source/receiver (SR) setup for a "shoot-through" approach.
Table 1: File (record) numbers and surface coordinates corresponding to each source/receiver (SR)
configuration for a "shoot-through" approach displayed in Figure 2.
Figure 3. Chart of source/receiver (SR) setup for the roll-along data set compiled from the
shoot-through data set displayed in Figure 2.
Figure 4. A 2D velocity (Vs) cross section analyzed from the roll-along data set displayed in Figure 3.
Figure 5. Chart of source/receiver (SR) setup for a shoot-through approach to extend the lateral
distance of the final 2D velocity (Vs) cross section.
Table 2: File (record) numbers and surface coordinates corresponding to each source/receiver (SR)
configuration for the "shoot-through" approach displayed in Figure 5.
Figure 6. Chart of source/receiver (SR) setup for the roll-along data set compiled from the
shoot-through data set displayed in Figure 5.