doi: 10.1111/j.1365-2478.2009.00835.

Geophysical Prospecting

Low-frequency passive seismic experiments in Abu Dhabi, United Arab

Emirates: implications for hydrocarbon detection
Mohammed Y. Ali1 , Karl A. Berteussen1 , James Small2 and Braham Barkat1
1 The

Petroleum Institute, P.O. Box 2533, Abu Dhabi, United Arab Emirates, and 2 Managing Consultant, Norway

Received July 2009, revision accepted August 2009

ABSTRACT
Low-frequency passive seismic experiments utilizing arrays of 3-component broadband seismometers were conducted over two sites in the emirate of Abu Dhabi in the
United Arab Emirates. The experiments were conducted in the vicinity of a producing oilfield and around a dry exploration well to better understand the characteristics
and origins of microtremor signals (16 Hz), which had been reported as occurring
exclusively above several hydrocarbon reservoirs in the region.
The results of the experiments revealed that a strong correlation exists between
the recorded ambient noise and observed meteorological and anthropogenic noises.
In the frequency range of 0.150.4 Hz, the dominant feature is a double-frequency
microseism peak generated by the non-linear interactions of storm induced surface
waves in the Arabian Sea. We observed that the double-frequency microseism displays
a high variability in spectral amplitude, with the strongest amplitude occurring when
Cyclone Gonu was battering the eastern coast of Oman; this noise was present at
both sites and so is not a hydrocarbon indicator. Moreover, this study found that
very strong microtremor signals in the frequency range of 23 Hz were present in
all of the locations surveyed, both within and outside of the reservoir boundary and
surrounding the dry exploration well. This microtremor signal has no clear correlation
with the microseism signals but significant variations in the characteristics of the
signals were observed between daytime and nighttime recording periods that clearly
correlate with human activity.
High-resolution frequency-wavenumber (f -k) spectral analyses were performed on
the recorded data to determine apparent velocities and azimuths of the wavefronts
for the microseism and microtremor events. The f -k analyses confirmed that the
double-frequency microseism originates from wave activity in the Arabian Sea, while
the microtremor events have an azimuth pointing towards the nearest motorways,
indicating that they are probably being excited by traffic noise. Results drawn from
particle motion studies confirm these observations. The vertical-to-horizontal spectral
ratios of the data acquired in both experiments show peaks around 2.53 Hz with no
dependence on the presence or absence of subsurface hydrocarbons. Therefore, this
method should not be used as a direct hydrocarbon indicator in these environments.
Furthermore, the analyses provide no direct evidence to indicate that earthquakes
are capable of stimulating the hydrocarbon reservoir in a way that could modify the
spectral amplitude of the microtremor signal.

E-mail:

[email protected]
at The Petroleum Institute, Abu Dhabi, United Arab
Emirates.
Formerly


C

2009 European Association of Geoscientists & Engineers

2 M.Y. Ali et al.

INTRODUCTION
The ambient noise of the Earth is generated by many independent sources that affect the overall frequency band of the natural background wavefield. Frequencies below 1 Hz are largely
generated by oceanic and large-scale meteorological events
(Longuet-Higgins 1950; Peterson 1993; Webb 2007). At frequencies above 1 Hz, the noise wavefield in urban settings is
dominated by cultural sources, particularly traffic, whereas in
remote sites wind generated noise is the predominant source
(Peterson 1993; Withers et al. 1996; Young et al. 1996; Wilson et al. 2002; McNamara and Buland 2004; Marzorati and
Bindi 2006; Bonnefoy-Claudet, Cotton and Bart 2006b). Cultural noise typically exhibits daily and weekly cyclical variations linked to human activities (Yamanaka, Dravinski and
Kagami 1993; Bonnefoy-Claudet et al. 2006b).
The measurement of ambient noise in the frequency band
below 1 Hz can help in our understanding of the interaction
between the solid earth, oceans and the atmosphere (Grevemeyer, Herber and Essen 2000; Kedar and Webb 2005).
Ambient noise in the 110 Hz frequency range (commonly
known as microtremor) can provide a low-cost and noninvasive exploration solution in urban sites where geotechnical information is often difficult to obtain. As a result microtremor measurements have been used in many studies to
determine fundamental resonant frequencies, shear-wave velocities and thicknesses of unconsolidated shallow sediments
(Ohori, Nobata and Wakamatsu 2002; Hartzell et al. 2003;
Scherbaum, Hinzen and Ohrnberger 2003; Chavez-Garcia
and Luzon 2005; Kind, Faeh and Giardini 2005; Picozzi,
Parolai and Richwalski 2005; Maresca, Galluzzo and Del
Pezzo 2006; Cho, Tada and Shinozaki 2006; BonnefoyClaudet et al. 2006a; Tada, Cho and Shinozaki 2006, 2007;
Chavez-Garcia and Rodriguez 2007; Dutta et al. 2007;
Wathelet et al. 2008; Stephenson et al. 2009). These data
can be used to predict local amplification of ground motion during earthquakes and for the preparation of seismic
microzonation maps (Gaull, Kagami and Taniguchi 1995;
Ansal, Iyisan and Gullu 2001; Tuladhar et al. 2004; Bhattarai 2005). Such information is crucial for seismic hazard
assessment. Predictions based on microtremor data have been
verified using other seismic techniques. For example, patterns
of ground motion amplification observed during earthquakes
are comparable to amplification patterns modelled from microtremor data (Horike, Zhao and Kawase 2001; Rodriguez
and Midorikawa 2003; Cara et al. 2008; Haghshenas 2008).
Similarly shallow shear-wave velocities determined from the
analyses of microtremor data have been verified through ap-


C

plication of downhole measurements and seismic refraction

profiles (Liu et al. 2000; Louie 2001; Di Giulio et al. 2006).
In addition, microtremor data have been used to image subsurface sedimentary structures. Draganov et al. (2007, 2009)
used seismic interferometery to cross-correlate ten hours of
microtremor data acquired in desert areas of Saudi Arabia
and Northern Africa. The cross-correlation revealed several
P-wave reflection events that correlated well with reflections
deduced from an active seismic survey at the same locations.
In more recent times it has been suggested that microtremor
measurements can be used to determine the location, depth
and thickness of hydrocarbon reservoirs. Anomalously high
spectral amplitudes of microtremor signals in the 16 Hz
frequency range, with a peak around 3 Hz, have been observed and reported over a number of hydrocarbon reservoirs, primarily in a gas field in Austria and several sites in the
Middle East, including some oilfields in Abu Dhabi (Singer
et al. 2002; Dangel et al. 2003; Holzner et al. 2005a,b,
2006a,b,c, 2007a,b; Frehner et al. 2006, 2007; Rached 2006,
2009; Lambert et al. 2007, 2009a,b; Steiner, Saenger and
Schmalholz 2007; 2008a; Saenger et al. 2007b, 2009a; van
Mastrigt and Al-Dulaijan 2008; Nguyen et al. 2008, 2009;
Goertz et al. 2009; Saenger, Torres and Artman 2009c). These
studies have claimed that a strong correlation exists between
microtremor spectral anomalies and the presence of hydrocarbons over several established oilfields. Their findings suggest that microtremor anomalies diminish towards the rim
of hydrocarbon reservoirs and are totally absent above nonreservoir locations.
Furthermore, it has also been suggested that microtremor
analysis has several potential applications in hydrocarbon exploration and production. These include reconnaissance of
frontier exploration areas, optimization of well placement,
reservoir monitoring and as a complementary tool to structural imaging to reduce drilling risk (Holzner et al. 2005c;
Schmalholz et al. 2006; Bloch and Akrawi 2006; Graf et al.
2007; Saenger et al. 2007a, 2009b). Some studies have suggested that a linear relationship can be derived between
the observed microtremor signal and the total thickness of
hydrocarbon-bearing layers from various oilfields, mainly in
the Middle East (Dangel et al. 2003; Holzner et al. 2005c;
Bloch and Akrawi 2006; Saenger et al. 2007a). Other studies
have claimed that these anomalous microtremor signals are
produced from resonant scattering and amplification within
the porous multi-phase (or partially-saturated) reservoir rocks
(Holzner et al. 2005c, 2009; Schmalholz et al. 2006; Graf et al.
2007; Saenger et al. 2007a, 2009a,b; Walker 2008; Frehner,
Schmalholz and Podladchikov 2009). In these theories, it is

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 3

assumed that the low-frequency ambient noise (microseisms)

of the Earth is the stimulus for the anomalous microtremor
signals observed. These studies also claimed that non-linear interaction of the microseism signal with liquid hydrocarbons,
water and the pore-rock materials in the reservoirs distorts
the ambient noise, resulting in the anomalously high verticalcomponent microtremor ground velocities above hydrocarbon reservoirs. To date there has not been any published work
that successfully applies modelling or mathematical expressions to link microtremor signals with subsurface hydrocarbon reservoirs.
Steiner, Saenger and Schmalholz (2008b) applied time reverse modelling to argue that microtremor data can accurately
determine locations, depths and thicknesses of hydrocarbon
reservoirs. However, it has since been noted that the time reverse modelling presented in that study was largely controlled
by the a priori velocity model and was minimally affected by
the microtremor wavefield (Green and Greenhalgh 2009).
Furthermore, a growing number of case studies have cast
doubt upon the applicability of the microtremor technique
for hydrocarbon detection. Berteussen (2008a,b) and Ali et al.
(2007, 2009a,b,c,d) showed that the observed anomalous microtremor signals are not related to body waves originating
from hydrocarbon reservoirs. Rather, they are caused by low
apparent velocity surface waves travelling through shallow
sediments. In another study, Hanssen and Bussat (2008) analysed ambient noise recorded over an oilfield in the Sahara
desert of Libya and suggested that high spectral amplitude
microtremor noise does not originate from the underlying
reservoir but rather are surface waves caused mainly by anthropogenic noises.
In this paper we present the results of microtremor investigations carried out over an onshore carbonate oilfield
and around a dry exploration well in Abu Dhabi, United
Arab Emirates. The aims of the experiments were to determine the characteristics and origin of microtremor signals and to investigate whether microtremor signals can be
used as a hydrocarbon indicator. We performed frequency,
joint time-frequency, V/H (vertical-to-horizontal) spectral ratio, frequency-wavenumber and particle motion analyses to
characterize the recorded ambient noise.
SURVEY AREAS
Geological setting
Oilfield
The oilfield for the study is located onshore in the Abu Dhabi
Emirate, about 50 km southwest of Abu Dhabi city (Fig. 1a,b).


C

The field has a moderate relief anticlinal structure and is limited by three major faults, two NW-SE trending and a WNWESE trending fault system. The anticline is interpreted to have
resulted from vertical movements due to basement tectonics
and later compressional stress related to regional tectonics associated with the obduction of the Oman ophiolites in the
Late Cretaceous. In addition, the field is affected by the collision of the Arabian Plate with central Iran along the Zagros suture, together with the culmination of the Musandam
peninsula during Oligocene-Miocene times. This resulted in
deformations related to strike-slip movements. However, the
main phase of deformation occurred during the Late Cretaceous. The shallow subsurface structure of the field consists
mainly of soft Quaternary sediments overlying hard Tertiary
carbonates.
The producing zones are all within a series of stacked reservoirs of the Lower Cretaceous Shuaiba and Kharaib Formations, all from the Thamama Group (Barremien to Late
Aptian) (Fig. 2). These carbonate reservoir intervals are separated by very low porosity intervals that are commonly referred to as dense zones. The reservoir zones correspond to
the late transgressive and highstand system tracts characterized by parasequence sets that show shoaling upward trends
of predominantly aggradational and progradational stacking
patterns.
The vertical closure at the main reservoir level is approximately 40 m. This oilfield was selected as a suitable site for the
experiment because it has a clear and well-defined oil/water
contact (OWC) mapped from 3D seismic and well data according to geoscientists currently working on the field. However, there is a slight possibility that deeper reservoirs (e.g., in
the Upper Jurassic Arab and Permian Khuff Formations) may
exist in the field. Nevertheless, the oilfield provides a unique
environment in which to measure and study the microtremor
signal both above and outside the reservoir.
Dry exploration well
The deep exploration well is located about 100 km southeast of Abu Dhabi city (Fig. 1a,b). The well was drilled with
a near-surface geology characterized by tens of metres of
poorly-consolidated aeolian Quaternary sediments. The seismic interpretation indicated the presence of small closures at
Kharaib Formation level. The well targeted a closure with
an elongated structural feature striking in a NW-SE direction
encompassing two culminations, being bound from the west
by a normal fault dipping to the west. The main exploration
objectives of the well were to investigate the hydrocarbon
potential and reservoir development of the Mishrif, Shuaiba

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

4 M.Y. Ali et al.

Figure 1 a) Regional satellite map showing the study areas. Stars show the location and date/time of Cyclone Gonu. b) Local satellite map
showing the study areas. Red lines show major motorways. c) Location map of the survey on the oilfield showing the oil/water contact (OWC).
Red triangles represent the position of the seismometers above the oil reservoir, whereas blue triangles represent the location of the seismometers
above the water saturated zone. d) Location map of the survey on the abandon exploration well. Red triangles indicate the position of the
seismometers.


C

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 5

B225NW

(c)

A1800NW

B225NE

B
B15NW
B15SW

B15NE
B15SE

B225SW

A900NE
A450NW
A225 A450NE
NW
A225NE

Location A
A225SW
A450SW

OWC

Location B

A1800NE

A900NW

B225SE

A900SW

Well W1

A225SE
A450SE

A900SE

B1800NW

A1800SW

Location B
A1800SE

B1800NE

B900NW
B900NE

A112NW
A112NE
A56NW
A56NE

B900SW

A15NW
A5SW
A15SW

B900SE

B1800SW

A15NE

Well 1 (A)

A15SE

A56SW
A112SW

B1800SE
500

500

1000

Location A

(meters)

(d)

D1

D5

C1

C5
B1
B5
A5

BH Well

A1

A2

B2

C2

A4
B4

A3
B3

C4

C3

D4

50

50

(m e te rs )

Figure 1 Continued.


C

A56SE

1500

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

D3

A112SE

6 M.Y. Ali et al.

Figure 2 Summary of stratigraphic column of the UAE foreland basin showing petroleum systems of Mesozoic and Tertiary. Modified from Ali
and Watts (2009).

(Middle-Lower Cretaceous) and the deeper Upper Jurassic

Formations (Fig. 2). The well eventually penetrated more than
16 500 ft of Mesozoic and Cenozoic sediments but logging
data indicated that the characteristics of all prospective reser-


C

voirs were insignificant. Moreover, open hole drillstem tests

(DST) and production tests of the reservoirs were all found
to be water-bearing. While this area was specifically chosen
for the study primarily because of the lack of hydrocarbons

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 7

present, there is always the possibility that deeper reservoirs

could exist (e.g., in the Khuff Formation). However, the seismic data show no clear structures that could trap significant accumulation of hydrocarbons below the Upper/Middle
Jurassic Formations.

surements. The seismometers have a flat frequency response

from 0.33100 Hz, with a sensitivity of 2000 V/m/s. A sampling rate of 200 Hz was used for the first experiment and
100 Hz for the second experiment. The sensors were covered
and buried up to 0.5 m deep where possible for firm ground
contact and wind shielding.

Data acquisition

Data processing

The survey over the oilfield was carried out between 21 May
and 17 June of 2007. During the acquisition of the data, a
powerful tropical cyclone (Cyclone Gonu) hit the coast of
Oman (Fig. 1a). Cyclone Gonu developed in the eastern Arabian Sea on 1 June attaining peak wind speeds of 240 km/h
on 3 June. Cyclone Gonu made landfall on the eastern coast
of Oman on 5 June with sustained wind speeds of approximately 150 km/h, becoming the strongest tropical cyclone to
hit the Arabian Peninsula in recorded times (De Bhowmick
et al. 2007; le Comte 2008). It then turned northward into
the Gulf of Oman and dispersed after moving ashore along
southern Iran on 7 June.
The survey consisted of a single 2D profile running between locations A and B, with several detailed studies conducted around each of these locations using arrays of sensors (Fig. 1c). The array geometry was designed to optimize the detection of microseism and microtremor signals.
Location A is situated over the maximum oil column of the
reservoir, whereas location B was positioned over an area
that presumably contained no oil. The 2D seismic profile was
recorded simultaneously and intended to connect the two sites.
Two seismometers were placed at locations A and B to continuously record throughout the entire survey and to monitor
any long-term variations in the ambient noise. The deployment of the sensor arrays consisted of 7 individual configurations centred at location A and 4 configurations at location
B, each with varying aperture sizes (from 303600 m). Each
array consisted of 5 broadband stations and utilized a typical
recording period of 24 hours.
The survey over the dry exploration well was carried out
in January of 2009 and included the deployment of four arrays with varying aperture sizes (50500 m) centred at the
well (Fig. 1d). Recording times for the arrays varied from
124 hours.
The signals were recorded using 3-component broadband
seismometers, Guralp CMG-6TD for the first experiment over
the oilfield and CMG-3EX for the second experiment around
the dry exploration well. Each sensor was equipped with an internal 24-bit digitizer and external global positioning system
(GPS) receiver to allow precise synchronization of the mea-

Time, frequency and joint time-frequency analyses were performed to characterize the recorded ambient noise. Data were
first converted into units of metres per second by correcting
for the digitizer output and the seismometer gain factor. A
processing window of 60 seconds with a 5% cosine taper was
applied to the data to reduce spectral leakage. Fourier amplitude spectra were analysed, both without smoothing applied
and with the smoothing procedure of Konno and Ohmachi
(1998), using a b-value of 40. The mean was removed and
excessively noisy sections of the signal were excluded from
the analyses before stacking the data. Figure 3 shows that
strong local wind conditions and regional earthquakes can
contaminate the signal by generating high-frequency (>6Hz)
and low-frequency (<2 Hz) noises respectively. Such noisy
data were removed prior to analyses.
Time-frequency representations were used to analyse the
spectral content of the data. Time-frequency representations
have been successfully used for signal recovery at low signalto-noise ratio, for accurate estimation of the instantaneous
frequency, for signal detection in communication, for radar
and for the design of time-varying filters (Cohen 1995). The
time-frequency representation technique that was used in this
study is the short-time Fourier transform, defined as

x( )h( t)ej2 f d
(1)
X(t, f ) =


C

where x( ) is the signal under investigation and h(t) is a window function centred around t = 0. The Fourier transform
is generated by shifting the window so that it is centred on
a time of interest, multiplying x( ) by the shifted version of
h(t) and then taking the Fourier transform of the resultant
windowed signal. The window can be shifted such that the
resulting time intervals are contiguous or overlapping. In the
examples below (Figs 6 and 10), contiguous windows were
considered.
ANALYSIS OF AMBIENT NOISE
Continuous recording over a period of 27 days allowed the
time response of the ambient seismic noise to be correlated
with large-scale meteorological conditions and anthropogenic

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

8 M.Y. Ali et al.

Figure 3 Time series of vertical particle velocity showing a) a regional earthquake that occurred in southern Iran on early morning (local time)
of 8 July 2007. b) A noisy data set probably due to windy conditions. c) Quiet data set. d) Spectral amplitudes of the records displayed in
Fig. 3(ac).


C

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 9

Spectrum amplitude (m/s)

10

x 10

-6

Iran earthquake - A900 SW

8 June at 01:00-01:30
Wind - Location A
26 May at 17:00-17:30

1.0
Quiet record - A15 NE
27 May at 01:30-02:00
0.1

(d)
0.2

0.4

0.6

0.8

10

Frequency (Hz)
Figure 3 Continued.

noises. Figure 4 shows the ambient noise observed at the two

sites. Many of the features of the figure are well understood
and include microseism (0.11 Hz) and microtremor (26 Hz)
signals. In this section the noise spectra amplitude estimated
for the two survey areas in these bands are detailed.
Frequencies below 1 Hz
Figure 4(a-c) shows representative spectral amplitudes of the
ambient noise recorded at locations A and B of the oilfield
and close to the dry exploration well, respectively. In the frequency range 0.150.4 Hz, the noise spectrum is dominated
by a strong and easily recognizable peak at 0.25 Hz called
the double-frequency peak (Longuet-Higgins 1950). The peak
of the double-frequency microseism spectrum occurs at a frequency nearly twice that of the ocean surface waves (LonguetHiggins 1950; Bromirski and Duennebier 2002). It is believed
that this double-frequency microseism occurs as a result of
the non-linear interaction between two ocean swells with the
same frequency, propagating in opposite directions (LonguetHiggins 1950; Friedrich, Kruger
and Klinge 1998; Kedar and

Webb 2005; Tanimoto 2007; Webb 2007). The conditions

that generate a double-frequency microseism arise in shallow
water due to the interaction of incident ocean swells and reflected/scattered wave energy from coastal areas (Bromirski
and Duennebier 2002; Bromirski, Duennebier and Stephen
2005). This noise occurs at all sites and thus can not be correlated with the presence of hydrocarbons.
Frequencies above 1 Hz
Figure 5(a) shows the spectral amplitude of the 2D profile
running from location A to location B of the survey over
the oilfield. The figure shows a distinct spectral microtremor
anomaly that peaks at about 2.5 Hz. The microtremor signal is observed on all stations whether positioned over the

C

oil reservoir (location A) or over the water saturated zone

(location B), with no significant differences in the spectral
amplitudes of microtremors signal inside or outside the reservoir boundary. In addition, all three components (vertical,
north-south and east-west) for all stations recorded the signal
(Fig. 4). This observation is confirmed by the second experiment over the dry exploration well, which shows a distinct
microtremor signal at about 2.5 Hz on all three components
(Fig. 5b). The figure shows that the spectral amplitude of microtremors at the dry well are much higher than that of the
oilfield.
The spectral amplitude of microtremor signals recorded in
these surveys exhibited distinctive cyclical day-to-night and
weekdays-to-weekend variations. During the daytime a strong
microtremor signal consistently appears, whereas during the
night only weak microtremor signals were detected (Fig. 6a,b).
The 24-hour cycle has minimum amplitude at around 1 am
local time and maximum amplitude late in the morning. In
addition, low microtremor amplitudes were observed during
the weekend (Fridays in the United Arab Emirates) compared
with a typical weekday. In the dry well location, differences of
spectral amplitude of the microtremor signal were observed
between daytime and nighttime measurements. In contrast,
the microseism signal shows negligible day and night variation, suggesting that the source of the signal is not dependant
on the local anthropogenic activity.
These microtremor signals have been attributed to the nonlinear interaction of the microseism signal with the hydrocarbon reservoir, causing enhanced vertically polarized P-waves
in the microtremor band above the reservoir compared with
positions away from the reservoir (Holzner et al. 2005c; Graf
et al. 2007; Walker 2008; Saenger et al. 2009b). However,
the fact that all three components have recorded the signal
means that the signal cannot possibly be a direct P-wave travelling vertically up from below the sensors, unless it has been

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

10 M.Y. Ali et al.

-6

Spectrum amplitude (m/s)

x 10

10

(a)

1.0
Vertical
North-South
East-West

0.1

0.2

0.1

Spectrum amplitude (m/s)

x 10

10

0.4

0.6

0.8

8 10

Frequency (Hz)

-6

(b)
Double-frequency
microseism peak
(distant source)

Microtremor
peak

High-frequency
noise

1.0
Vertical
North-South
East-West

0.1

0.2

0.1
10

Spectrum amplitude (m/s)

Microtremor
peak
High-frequency
noise

Double-frequency
microseism peak
(distant source)

x 10

0.4

0.6

0.8

-6

(c)

8 10

Microtremor
peak
Double-frequency
microseism peak
(distant source)

1.0

Frequency (Hz)

High-frequency
noise

Vertical
North-South

0.1

East-West

0.1

0.2

0.4

0.6

0.8

8 10

Frequency (Hz)
Figure 4 Examples of typical spectral amplitudes of 60 minute periods of ambient noise recorded on vertical and horizontal components at
a) location A of the oilfield survey, 7 June 2007 at 00:0001:00. b) Location B of the oilfield survey, 7 June 2007 at 00:0001:00. c) Close
to the dry exploration well, 21 January 2009 at 00:0001:00. All locations show double-frequency microseism at a frequency of around
0.25 Hz, microtremor at around 2.5 Hz and high-frequency noise. Microseism and microtremor signals were observed on all three seismometer
components (vertical, north-south and east-west) at all recording stations. For sensor locations see Fig. 1.

scattered or converted. In addition, the strong diurnal variation in the microtremor signal suggests that the source responsible is possibly related to surface waves caused by the coupling of anthropogenic noise (e.g., traffic, production installations), which tend to have minimums at night and on weekends due to the cyclical nature of cultural noise. Such daily
and weekly variations in spectral amplitudes of microtremor
signals have been reported in many other studies that have
attributed cultural activities revealed by regular daily spectral


C

amplitude variations with minimums recorded at midnight

and maximums at midday (Yamanaka et al. 1993; BonnefoyClaudet et al. 2006b).

COMPARISON OF AMBIENT NOISE

WITH METEOROLOGICAL DATA
The correlation between the double-frequency microseism
peak and the presence of ocean storms is supported by the

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 11

x 10

-7

Spectrum amplitude (m/s)

10

(a)

Location A
A900SW
1800SW
OWC
B2700NE
B1800NE
Location B

8
6
4
2
0
2

Spectrum amplitude (m/s)

x 10

10

Frequency (Hz)

-6

(b)

BH well
B1
B2
B3
B4
B5

3
2
1
0
2

10

Frequency (Hz)
Figure 5 a) Spectral amplitudes of vertical components of the sensors located along a profile from location A to B on the oilfield on 7 June
2007 at 00:0001:00. b) Spectral amplitudes of sensors with array aperture of 100 m centred at the exploration dry well on 20 January 2009
at 00:0001:00. All locations (above and outside the oilfield) show distinct microtremor signal.

temporal variation of spectral levels observed in the first survey during the occurrence of Cyclone Gonu in 2007 (Figs 1a
and 7a). Spectral amplitudes of the double-frequency microseism increased as Cyclone Gonu developed then reached
their maximum on 6 June as Gonu approached the eastern
Omani coast. The spectral amplitudes immediately dropped
back again once the cyclone had passed the region. Therefore, the appearance of Cyclone Gonu over the Arabian Sea
during 46 June effectively explains the larger spectral amplitudes of double-frequency microseisms observed during the
survey.
It is not surprising that microseism waves were recorded
even at a distance of more than 500 km from the Omani
coast. The microseism energy propagates primarily as fundamental Rayleigh waves through the Earths crust and hence
does not attenuate rapidly and may be observed at continental sites far removed from the coastlines (Haubrich and
McCamy 1969; Barstow, Sutton and Carter 1989; Bromirski
and Duennebier 2002; Bonnefoy-Claudet et al. 2006b). For
example, Gerstoft, Fehler and Sabra (2006) reported observ-


C

ing microseisms in Southern California which were generated

by Hurricane Ktarina in New Orleans some 2700 km away.
Comparing Fig. 7(a,b) over the duration of the survey indicates there is no direct relationship between the strength of
the microtremor and the microseism signals. During the period in which Cyclone Gonu was battering the eastern Omani
coast the spectral amplitude level of the microseism signal increased by a factor of about 10. This increase is seen to be
essentially identical for both the vertical and horizontal components, whereas the microtremor signal remained seemingly
unchanged. Hanssen and Bussat (2008) also noted that there
is no correlation between the low-frequency (16 Hz) band
and microseism signal (<0.25 Hz). This is inconsistent with
the assertion made by Holzner et al. (2005c, 2009) that the
driving force in the generation of the anomalous microtremor
signals are microseism events.
Winds can be considered as broadband sources producing large amplitudes of noise at high frequencies (>1 Hz),
representing fluctuations in response to the variation of
the wind intensity (Withers et al. 1996; Young et al.

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

12 M.Y. Ali et al.

Figure 6 a) Time frequency display for data recorded from the vertical component of the seismometer at location A of the oilfield on 26 May
2007. b) Time frequency display for data recorded from the vertical component of the seismometer close to the dry exploration well on 20
January 2009. On both locations the microtremor signal exhibits strong diurnal variations with a strong signal during the day and a weaker
signal at night.

1996). Hence, we compared the microtremor measurements

with local wind speed and temperature obtained from a
weather station located approximately 30 km north-east of the
oilfield.
Figure 8(a-c) shows a comparison of microtremor signals
recorded on the oilfield with meteorological data (wind speed
and temperature). The figure shows that the maximum spectral amplitude of microtremor signal occurs at about 8:00 am
local time, which corresponds to the period with lowest av-


C

erage wind speed (23 km/s) and relatively low temperature

(33 C). The highest temperature (>45 C), which occurs at
around midday, correlates well with a period of relatively
low microtremor spectral amplitude, which is probably due
to there being less cultural noise at that time (lunch and prayer
breaks). The lack of a direct correlation between the fluctuations in spectral amplitudes of the microtremor and wind
speed is probably due to the isolating effects of burying the
seismometers to a depth of 0.5 m. However the separation

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 13

x 10

-6

(a)

50
45

Spectrum amplitude (m/s)

40
35
30
25
20

North-South
East-West
Vertical

15
10

Cyclone Gonu

5
0
2

10

14

13

12

11

16

15

17

Days (June 2007)

x 10

-7

10

Spectrum amplitude (m/s)

(b)
8

Friday

Friday

Friday

Thur

Fri

Sat

Sun

Mon

Tue

Thur
6 Wed 7
8

Fri

Sat

10

Sun

11

Mon

12

Tue

13

Wed

14

Thur

15

Fri

16

Sat

17

Days (June 2007)

Figure 7 a) Peak spectral amplitude of horizontal and vertical components of double-frequency microseism (about 0.25 Hz) obtained at
location B of the oilfield. b) Peak spectral amplitude of microtremor signal (2.5 Hz) horizontal and vertical components recorded at location B
of the oilfield. The spectral amplitudes of microseisms increased dramatically when Cyclone Gonu approached the coast of Oman, whereas the
microtremor signal remained unchanged. Spectral amplitudes of microtremor signals exhibit strong daily and weekly cyclical variations. Signals
were significantly stronger during normal working hours on weekdays compared with nighttimes and across weekends (Fridays in the United
Arab Emirates).

(30 km) between the weather station and the survey may also
affect this.

VERTICAL/HORIZONTAL SPECTRAL
RATIOS
The H/V spectral ratio technique (Nakamura 1989) consists of
estimating the ratio between the Fourier amplitude spectra of
the horizontal and the vertical components of the microtremor
recorded at the ground surface. The method takes advantage
of the fact that, in a soft soil layer, horizontal ground movements (mainly Rayleigh waves) are more strongly amplified
than the vertical movements. As a result, there is a correlation between the H/V peak frequency and the fundamental
resonance frequency of the site. The technique provides reliable estimates of fundamental resonance frequency and am-


C

plification factors of the uppermost soft layers (Nakamura

1989; Lermo and Chavez-Garcia 1994; Konno and Ohmachi
1998).
It has been suggested that the dominant amplitude peak
of the ratio of vertical to horizontal components (V/H) in
the microtremor range is higher over hydrocarbon reservoirs
(Lambert et al. 2007). The proponents of this assertion reason that hydrocarbon reservoirs emit distinctive P-waves that
cause an increased vertical polarization of the ambient noise
wavefield at the surface. This suggestion has resulted in the
premise that high V/H values (>1) can be used as a direct hydrocarbon indicator (Walker 2008; Lambert et al. 2009a,b;
Saenger et al. 2009b; Goertz et al. 2009; Nguyen et al.
2009).
For each station in this study, the spectral amplitudes for
the horizontal and vertical components were calculated over

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

14 M.Y. Ali et al.

x 10

Spectrum amplitude (m/s)

10

-7

(a)

North-South
East-West
Vertical

Friday

Friday

Friday

Thur

Fri

Sat

Sun

Mon

Tue

Thur
6 Wed 7
8

Fri

Sat

10

Sun

11

Mon

12

Tue

13

Wed

14

Thur

15

Fri

16

Sat

17

Days (June 2007)

Wind speed (km/hr)

30

(b)

25
20
15
10
5
1

10

11

12

13

14

15

16

17

10

11

12

13

14

15

16

17

Days (June 2007)

(c)

Temperature (C)

45

40

35

30

Days (June 2007)

Figure 8 Comparison of a) peak spectral amplitude of microtremor signal with b) wind speed and c) temperature. Maximum spectral amplitude
of microtremor occurs at around 08:00 am local time, which is close to when the wind speed is slowest. Highest temperature at midday correlates
well with relatively low spectral amplitude, which is probably due to lunch and prayer break.

a continuous 60-minute period of ambient noise. The Fourier

amplitude spectra were then smoothed, following the method
of Konno and Ohmachi (1998) and the quadratic mean of
the horizontal amplitude spectrum was divided by the corresponding vertical spectrum. The final result was obtained by
averaging the H/V ratios from all windows and then inverting
to obtain the V/H ratio.
The V/H spectral ratios measured over the oilfield (Fig. 9)
indicate that a peak (>1) exists at 23 Hz on most of the stations whether measured over oil-saturated or water-bearing
areas. The anomalous V/H curve obtained at location A, for
frequencies higher than 3 Hz, is probably related to localized


C

noise generated by the structures immediately surrounding

well A. The results from V/H analysis for the second experiment over the dry well show an identical V/H peak also at
about 2.5 Hz. Hence, these results would indicate that the
V/H method could not be used as a hydrocarbon indicator in
this case.

EARTHQUAKE TRIGGERED
MODIFICATION
It has been suggested that hydrocarbon reservoirs can be stimulated by earthquake activity to temporarily alter the local

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 15

1.8

(a)

V/H

1.4

1.0

Location A
A900SW
1800SW
OWC
B2700NE
B1800NE
Location B

0.6

0.2
1

10

10

Frequency (Hz)
1.8

(b)

V/H

1.4

1.0

BH well
B1
B2
B3
B4
B5

0.6

0.2
1

Frequency (Hz)

Figure 9 a) Mean V/H spectral ratio of microtremor data recorded at the oilfield for 60 minutes on 7 June 2007 at 00:0001:00. b) Mean V/H
spectral ratio of microtremor measurement at the dry well for 60 minutes on 20 January 2009 at 00:0001:00. V/H spectral ratios of 2.53 Hz
are observed on both sites. The V/H peaks of 2.53.0 Hz are observed on both sites. The peaks are not related to the location of the hydrocarbon
reservoir.

ambient noise wavefield (Nguyen et al. 2008). This stimulation is said to result in a significant spectral increase of the
microtremor signal above the reservoir both during the earthquake and for at least an hour after the event. To test this
observation, teleseismic and regional earthquakes that were
recorded during the survey over the oilfield were examined.
An earthquake that occurred in southern Iran on 8 June (local time) with magnitude of 4.2, for example, is shown in
Figs 3(a) and 10. The onset of this event is apparent with a
sharp increase in the amplitudes of all three components over
a wide frequency band. However, there is no clear indication as to whether the spectral amplitude of the microtremor
signal is amplified during or after the earthquake. Both the
vertical and north-south components actually indicate a slight
decrease of spectral amplitude for the microtremor frequencies immediately after the earthquake, which is inconsistent
with observations reported by Nguyen et al. (2008).


C

FREQUENCY-WAVENUMBER ANALYSIS
AND WAVE PROPAGATION ACROSS
THE ARRAYS
Frequency-wavenumber analyses of array data provide an
excellent means for discriminating the seismic wavefield by
phase velocity, propagation azimuth and frequency. These
in turn provide detailed information on the source generation and propagation modes of noise wavefield (Capon 1969;
Haubrich and McCamy 1969; Cessaro 1994; Satoh, Kawase
and Matsushima 2001; Okada 2003; Chevrot et al. 2007). In
this study, we applied high-resolution frequency-wavenumber
(f -k) spectral analysis to the array data, as proposed by Capon
(1969). In doing so, we were able to identify the directions of
approach and dominant phase velocities of the microseism
and microtremor waves as they moved over the array, thereby
distinguishing body waves from surface waves and possibly
identifying their origins. The method assumes plane waveform

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

16 M.Y. Ali et al.

Figure 10 Time frequency display for data recorded from a) vertical component and b) north-south component. c) East-west component showing
that a regional earthquake occurred in southern Iran on early morning (local time) of 8 June 2007 with magnitude of 4.2. The earthquake has
a well-defined signature characterized by a vertical streak of high energy over a wide frequency band.


C

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 17

propagation through the seismic array and aims to improve

the resolution in wavenumber space.
In this study the propagation azimuth and slowness of the
microseism and microtremor bands were measured in each
array by finding the peak power in slowness space of continuous 60-minute signal samples. The wavenumber domain
was computed on a uniformly sampled grid in f -k space using
vertical component signals from the array recordings.
Figure 11(a,b) shows the slowness maps for microseism
bands (centre frequency of 0.25 Hz) in the oilfield from crossshaped arrays with apertures of 3600 m at locations A and B
respectively. While such a cross-shaped array does not have
an isotropic mapping pattern, this does not dramatically affect the detection of the azimuth of the approaching wavefield. A straight arrow denotes the corresponding propagation back azimuth for the wavenumber vector at the peak
spectrum amplitude. The signal across both arrays exhibits
high coherency, showing a single well-defined peak at slowness corresponding to that of crustal Rayleigh waves with
apparent velocity of approximately 3600 m/s. This high apparent velocity indicates that the wavefield has interacted with
deeper more compacted carbonate rocks, consistent with these
low frequencies. The propagation back azimuth (i.e., direction from the array pointing back towards the source) of the
wavefront varies from 305327 , with the slight difference
on propagation azimuth most likely due to change in source
positions. The wavefront is interpreted as being microseismic
events generated by wave activity in the Arabian Sea, far to the
south-east.
Figure 11(c) illustrates the slowness map for the microtremor (centre frequency of 2.5 Hz) signal recorded at the
oilfield, with array aperture of 225 m. The propagation azimuth is relatively scattered although the maximum energy
response is about 180 pointing in the direction away from a
major motorway (Fig. 1b). The apparent velocity of the wavefronts is about 1150 m/s. Figure 11(d) illustrates the slowness
map for microtremor (centre frequency of 2.5 Hz) measurement recorded at the dry well with array aperture of 200 m.
Examination of the phase velocities of the wavefront indicates a tendency to cluster at approximately 800 m/s with a
propagation azimuth of 165 , again pointing to the nearest
motorway (Fig. 1b).
The study area is composed of shallow unconsolidated sediments, which lie directly over hard carbonate layers with
P-wave velocities far above 1150 m/s. Therefore, it is unlikely
that the observed wavefront is an ordinary P-wave originating
from the subsurface hydrocarbon reservoir. If the recorded
waves were actually coming directly from below the array,


C

then they should arrive simultaneously at all seismometers

(i.e., the apparent velocity would be infinite and the azimuth
undefined).
On the basis of these observations, the origin of the anomalous microtremor events is interpreted as comprising surfacecoupled waves excited mainly by traffic noise from motorways that cross the survey areas at a distance of about 10
15 km (Fig. 1b). This analysis correlates with the interpretation drawn from the spectral amplitude analyses but contrasts with other studies that have attributed the spectral peaks
of the microtremor events with the location of subsurface
hydrocarbon reservoirs (Dangel et al. 2003; Holzner et al.
2005c; Walker 2008; Saenger et al. 2009b; Lambert et al.
2009a).

PARTICLE MOTION
Assuming that the particle motion of double-frequency microseisms and microtremors (which are mainly composed of
Rayleigh waves) are elliptical and that the medium is azimuthally isotropic, then the wave should have a rotational
motion in a vertical plane oriented in the azimuthal direction
(i.e., in the direction of wave propagation) (Barstow et al.
1989; Bromirski and Duennebier 2002; Tanimoto, Ishimaru
and Alvizuri 2006; Bonnefoy-Claudet et al. 2006b). If, for
example, the wave is actually a Love wave then the motion
will be transverse to this. To test this proposition, processing
of the data in this study was initially band-pass filtered into
a selected frequency window. An azimuth of approach was
assumed and the horizontal signals were rotated into a radial component in the assumed direction of approach and the
transverse component at 90 to this azimuth. The root-meansquare (rms) of each of the radial and transverse components
were calculated. Finally the ratio of the radial energy and
transverse energy were computed. This process was repeated
for all azimuths from zero to 180 .
In Fig. 12 the ratio between maximum and minimum energy levels of horizontal components for a one hour interval (18:0019:00 local time for the days 3 and 5 to 8 June)
are presented. The time period reflects the days immediately
prior to Cyclone Gonu striking the coast of Oman up until
it dispersed two days later. For the frequency band 0.25 Hz
(double-frequency microseism) a clear correlation exists between the azimuth of maximum energy and the movement of
the cyclone along the Omani coast. Before the cyclone had
approached the coast, the particle motion tends to be more
focused with an azimuth of approach pointing around N98 E
(i.e., south-east towards the Arabian Sea). As the cyclone

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

18 M.Y. Ali et al.

Figure 11 Normalized energy response in slowness space for arrays of varying aperture sizes at a) location A, oilfield: centre frequency = 0.2 Hz,
array aperture = 3600 m, 4 June 2007 at 00:0001:00. b) Location B, oilfield: centre frequency = 0.2 Hz, array aperture = 3600 m, 16 June
2007 at 00:0001:00. c) Location A, oilfield: centre frequency = 2.5 Hz with array radius of 225 m, 26 May 2007 at 00:0001:00. d) Dry well:
centre frequency = 2.5 Hz with array radius of 100 m, 20 January 2009 at 00:0001:00. For locations of the sensors see Fig. 1. In each figure
the symbol X indicates the peak values. The distance between the centres and X gives the slowness of the waves at the frequency and the line
at the centre to X gives the direction of the wave propagation. The phase velocity and propagation azimuth (from the source) determined from
the maximum peak are written in the lower left-hand corner in each plot.


C

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 19

(a)

(b)
4

10

Energy

3 June 2007
2

0
0

20

40

60

80

100

120

140

160

180

5 June 2007

20

40

60

80

100

120

140

160

180

80

100

120

140

160

180

20

40

60

80

100

120

140

160

180

40

60

80

100

120

140

160

180

40

60

80

100

120

140

160

180

40

60

80

100

120

140

160

180

10

6 June 2007

Energy

60

6 June 2007

0
0

20

40

60

80

100

120

140

160

180

20

10

Energy

40

5 June 2007

7 June 2007

7 June 2007
2

0
0

20

40

60

80

100

120

140

160

180

20

10

8 June 2007

Energy

20

10

Energy

3 June 2007

8 June 2007
2

0
0

20

40

60

80

100

120

140

160

180

20

Azimuth (Degrees)

Azimuth (Degrees)

Figure 12 Ratio between maximum and minimum energy of the horizontal components plotted as a function of azimuth a) for microtremor,
centre frequency 2.5 Hz and b) for microseism, centre frequency 0.25 Hz. The arrow points the azimuth for maximum energy.

approaches the coast the azimuth shifts towards the east and
north-east, whilst becoming less coherent. This effect could be
expected as the storm gets nearer to the coast and produces a
wider distribution of sources. After the cyclone finally dissipates the azimuth of maximum energy returns to the previous
direction.
For the duration of the cyclone, the microtremor signals
exhibit minimal notable changes. A de-focusing of the amplitudes can be observed on 6 June but there is no visible
change to the peak azimuth. These results are consistent with
the results obtained from the frequency-wavenumber analysis and would indicate an independence of sources for the
microtremors and microseism signals.


C

DISCUSSION
The characteristics of ambient noise recorded over an oilfield
and an abandoned exploration well have been investigated
in relation to recent claims that signals in the microtremor
range can be applied to hydrocarbon detection. The analyses of the data provide a better understanding of the nature and origin of anomalously high microtremor signals that
have reportedly been observed over several hydrocarbon reservoirs in the region. The results for this study indicate that
high levels of microtremor signals are present above the hydrocarbon reservoir as previously claimed but that similar
levels of this signal are also observed over nearby assumed

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

20 M.Y. Ali et al.

(a)

Location A
A900SW
1800SW
OWC
B2700NE
B1800NE
Location B

1.4

H/V

1.2

1.0

0.8

0.6
1

Frequency (Hz)

(b)

BH well
B1
B2
B3
B4
B5

1.4

H/V

1.2

1.0

0.8

0.6
1

Frequency (Hz)

Figure 13 a) Mean H/V spectral ratio of 60 minutes microtremor data recorded at the oilfield on 7 June 2007 at 00:0001:00. b) H/V spectral
ratio of 60 minutes microtremor data recorded at the dry well on 20 January 2009 at 00:0001:00.

non-reservoir locations. Moreover, the results indicate that

the observed anomalous microtremor signals have originated
from surface waves with a propagation back azimuth pointing towards the nearest motorways in the area. Similar findings have been published from other studies. For example,
Hanssen and Bussat (2008) examined the relative traveltimes
of microtremor measurements over an oilfield in Libya and
concluded that the observed waveforms were actually surface
waves caused by anthropogenic sources such as production
facilities, traffic and the resonance frequencies of unconsolidated overburden in the area.
Whilst it could be suggested that had weak microtremor
signals from hydrocarbon reservoirs been present in the survey locations then they may have been masked by the strong
anthropogenic noise (e.g., from traffic) and thus not been detected at all without the application of an appropriate filter
such as a f-k filter (Nguyen et al. 2009). However, one of the
key design criteria in the planning of this survey was to utilize
3-component sensors in graduated arrays that would enable
full discrimination of laterally propagating waves from vertical, and to provide detailed information on the phase velocity
and direction of approach of any coherent wave energy. This


C

methodology is in direct contrast to previous surveys which

typically relied on simple lines of discrete sensors, thus making it impossible to differentiate individual wave trains based
on their propagation characteristics other than frequency content. As demonstrated in this study the best approach to discriminating waveforms originating from multiple sources is
to acquire data in arrays. Also, to minimize the effects of anthropogenic noise in this survey, whenever possible data for
analysis was specifically selected from the 24 hour records
when noise levels were at their lowest (i.e., during the night
and early morning).
Although our results clearly indicate that microtremor analysis cannot be used as a hydrocarbon indicator for this area,
the data do however have the potential to significantly contribute towards geotechnical characterization of shallow sediments for seismic hazard assessment. For example, it is well
documented in the literature (Nakamura 1989; Bodin and
Horton 1999; Cara, Di Giulio and Rovelli 2003; Guillier et al.
2005; Stephenson et al. 2009) that unconsolidated shallow
sedimentary layers cause amplification of ambient noise at
distinct frequencies in the range of 110 Hz, which are related to the geometry and the seismic properties of the soil

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 21

layer. As a result microtremor data can be used to estimate

fundamental resonant frequencies as well as shear-wave velocities and thickness of the uppermost soft layers.
Figure 13 shows the H/V (reciprocal of V/H) spectral ratio of microtremor measurements of the two survey areas.
The fundamental resonance frequency of the oilfield peaks at
about 1.7 Hz, with greater variability in terms of H/V envelope. The dry exploration well shows two well-pronounced
peaks at 1.3 Hz and 4.2 Hz. The peak at 1.3 Hz most likely
represents the fundamental resonance frequencies of the site,
whereas the peak at 4.2 Hz may be related to a contrast at
deeper depth or industrial origin.
Using the observed fundamental resonance frequencies we
calculated the thickness of the soft soil in both survey areas
and then compared that result with the well data, according
to the well-known equation (Parolai, Bormann and Milkereit
2002)
Vs,average
(2)
f H/V f0 =
4h
where V s,average and h are the average shear-wave velocity and
the thickness of sediments respectively. The shear-wave velocities were estimated to be 8151220 m/s from array analysis
using a high-resolution frequency-wavenumber (f -k) spectral
technique and from sonic log data from the well.
By use of this argument, the thickness of unconsolidated
sediments for the oilfield in this study was estimated to be
around 120180 m. This is consistent with the drilling data
close to location A, which penetrated 130 m of Quaternary
age aeolian sediments. The thickness of the unconsolidated
sediments at the dry exploration well was estimated to be
154 m but this could not be confirmed as the well did not
record any data at shallow levels.
This study does not conclusively disprove the assertion that
the interaction of microseism energy with porous multi-phase
hydrocarbon reservoirs can generate microtremor signals as
reported by Dangel (2003), Lambert (2009a) and Saenger
(2009b). However, for the sites investigated and described
in this paper any signals possibly originating from such interactions have been completely overshadowed by other effects,
primarily noise caused by cultural activity. These constraints
plus the complete lack of evidence in favour of the microtremor detection technique make such analyses unsuitable
for hydrocarbon detection in the environment for this study.
CONCLUSIONS
The following conclusions are drawn from this study:
r Double-frequency microseism signals are observable within
the frequency band of 0.150.4 Hz. The spectral amplitudes


C

r
r
r
r

of the microseism increased as Cyclone Gonu approached

the coast of Oman then declined once the cyclone passed
the region. There is no apparent correlation between the microtremor and microseism signals and therefore, the driving
force of the microtremor signal cannot be the microseism
events.
Microtremor (2.53 Hz) signals are observed over two sites
(over an oilfield and above a dry well). During the day
all sensors recorded strong microtremor signal, whereas all
sensors detected only weak signals over night. There is no
direct correlation between the maximum spectral amplitude
of microtremors and meteorological data. Cyclical daily and
weekly variations in the microtremor spectral amplitudes
clearly correlate with human activity.
V/H spectral ratios of 2.53 Hz are observed on both sites.
The V/H peak is not related to the location of the known
hydrocarbon reservoir.
Analyses of regional earthquake data show no evidence of
earthquake triggered spectral amplitude modification by the
hydrocarbon reservoir.
Particle motion studies clearly show that the microseism
signal was influenced by the action of Cyclone Gonu, while
the microtremor band displayed no such correlation.
The apparent velocity and propagation azimuth (from the
source) of microseism signals were 35003650 m/s and
305327 , respectively for the oilfield. These results suggest that the source of the microseism is the ocean swells of
the Arabian Sea.
The apparent velocity and propagation azimuth for the microtremor signal recorded at the oilfield and dry well are
8001150 m/s and 165180 , respectively. These results indicate that the observed microtremor signals originate from
surface waves propagating through shallow sediments having an azimuth directed from the nearest motorways.

ACKNOWLEDGEMENTS
We are grateful to the Oil Subcommittee of the Abu Dhabi
National Oil Co. (ADNOC) and its operating companies (OpCos) for sponsoring this project. We thank Mr Marwan Haggag for his logistical support of the fieldwork and in coordinating the project and Islam Md. Didarul for providing the
meteorological data.
REFERENCES
Ali M.Y., Berteussen K.A., Small J., Anjana B.T. and Barkat B. 2009a.
Recent passive experiments in Abu Dhabi. EAGE Passive Seismic
Workshop Exploration and Monitoring Applications, Limassol,
Cyprus, Expanded Abstracts, A36.

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

22 M.Y. Ali et al.

Ali M.Y., Berteussen K.A., Small J., Anjana B.T., Barkat B. and
Pahlevi O. 2009b. Recent low frequency passive seismic experiments in Abu Dhabi. 71st EAGE meeting, Amsterdam, The Netherlands, Expanded Abstracts, S037.
Ali M.Y., Berteussen K.A., Small J. and Barkat B. 2007. A low frequency passive seismic experiment over a carbonate reservoir in
Abu Dhabi. First Break 25, 7173.
Ali M.Y., Berteussen K.A., Small J., Barkat B. and Pahlevi O. 2009c.
Results from a low frequency passive seismic experiment over an
oilfield in Abu Dhabi. First Break 27, 9197.
Ali M.Y., Berteussen K.A., Small J. and Pahlevi O. 2009d. Microseism
and microtremor analyses over an oilfield in Abu Dhabi Implications for cyclone and hydrocarbon detection. 71st EAGE meeting,
Amsterdam, The Netherlands, Expanded Abstracts, SO34.
Ali M.Y. and Watts A.B. 2009. Subsidence history, gravity anomalies
and flexure of the United Arab Emirates (UAE) foreland basin.
Geoarabia 14, 1744.
Ansal A.M., Iyisan R. and Gullu H. 2001. Microtremor measurements
for the microzonation of Dinar. Pure and Applied Geophysics 158,
25252541.
Barstow N., Sutton G.H. and Carter J.A. 1989. Particle motion and
pressure relationships of ocean bottom noise at 3900 m depth:
0.003 to 5 Hz. Geophysical Research Letters 16, 11851188.
Berteussen K.A., Ali M.Y., Small J., Anjana B.T. and Barkat B. 2008a.
Analysis of low frequency passive seismic data from an experiment
over a carbonate reservoir in Abu Dhabi. Abu Dhabi International
Petroleum Exhibition and Conference, Abu Dhabi, United Arab
Emirates, SPE -117925-MS.
Berteussen K.A., Ali M.Y., Small J. and Barkat B. 2008b. A low
frequency, passive seismic experiment over a carbonate reservoir
in Abu Dhabi Wavefront and particle motion study. 70th EAGE
meeting, Rome, Italy, Expanded Abstracts, B046.
Bhattarai M. 2005. Seismic microzonation using H/V spectral ratios with single station microtremor survey. Individual Studies by
Participants at the International Institute of Seismology and Earthquake Engineering 41, 7386.
Bloch G. and Akrawi K. 2006. Application of passive seismic (IPDS)
surveys in Arabian Peninsula. EAGE passive seismic workshop
Exploration and monitoring applications, Dubai, United Arab Emirates, Expanded Abstracts, A28.
Bodin P. and Horton S. 1999. Broadband microtremor observation
of basis resonance in the Mississippi embayment, Central US. Geophysical Research Letters 26, 903906.
Bonnefoy-Claudet S., Cornou C., Bard P.Y., Cotton F., Moczo P.,
Kristek J. and Fah D. 2006a. H/V ratio: a tool for site effects evaluation. Results from 1-D noise simulations. Geophysical Journal
International 167, 827837.
Bonnefoy-Claudet S., Cotton F. and Bard P.Y. 2006b. The nature
of noise wavefield and its applications for site effects studies A
literature review. Earth-Science Reviews 79, 205227.
Bromirski P.D. and Duennebier F.K. 2002. The near-coastal microseism spectrum: Spatial and temporal wave climate relationships. Journal of Geophysical Research-Solid Earth 107, 2166.
doi:10.1029/2001JB000265
Bromirski P.D., Duennebier F.K. and Stephen R.A. 2005. Mid-ocean
microseisms. Geochemistry Geophysics Geosystems 6, Q04009.
doi:10.1029/2004GC000768.


C

Capon J. 1969. High-resolution frequency-wavenumber spectrum

analysis. Proceedings of the IEEE 57, 14081418.
Cara F., Cultrera G., Azzara R.M., De Rubeis V., Di Giulio G., Giammarinaro M.S. et al. 2008. Microtremor measurements in the
city of Palermo, Italy: Analysis of the correlation between local geology and damage. Bulletin of the Seismological Society of America
98, 13541372.
Cara F., Di Giulio G. and Rovelli A. 2003. A study on seismic
noise variations at Colfiorito, central Italy; Implications for the
use of H/V spectral ratios. Geophysical Research Letters 30, 1972.
doi:10.1029/2003GL017807.
Cessaro R.K. 1994. Sources of primary and secondary microseisms.
Bulletin of the Seismological Society of America 84, 142148.
Chavez-Garcia F.J. and Luzon F. 2005. On the correlation of seismic
microtremors. Journal of Geophysical Research-Solid Earth 110,
B08302.
Chavez-Garcia F.J. and Rodriguez M. 2007. The correlation of microtremors: empirical limits and relations between results in frequency and time domains. Geophysical Journal International 171,
657664.
Chevrot S., Sylvander M., Benahmed S., Ponsolles C., Lefevre
J.M. and Paradis D. 2007. Source locations of secondary microseisms in western Europe: Evidence for both coastal and pelagic
sources. Journal of Geophysical Research-Solid Earth 112, B11301.
doi:10.10292007JB005059
Cho I., Tada T. and Shinozaki Y. 2006. A generic formulation for
microtremor exploration methods using three-component records
from a circular array. Geophysical Journal International 165, 236
258.
Cohen L. 1995. Time-Frequency Analysis. Prentice-Hall.
le Comte D. 2008. Global Weather Highlights 2007: A mixed bag.
Weatherwise 61, 1618.
Dangel S., Schaepman M.E., Stoll E.P., Carniel R., Barzandji O., Rode
E.D. and Singer J.M. 2003. Phenomenology of tremor-like signals
observed over hydrocarbon reservoirs. Journal of Volcanology and
Geothermal Research 128, 135158.
De Bhowmick A.K., Thani M.A. and Al-Rawahi Y. 2007. Cyclone
GONU and reliability of main interconnected transmission system
of Oman. IEEE Conference, Perth, Australia, Expanded Abstracts,
4852.
Di Giulio G., Cornou C., Ohrnberger M., Wathelet M. and Rovelli A.
2006. Deriving wavefield characteristics and shear-velocity profiles
from two-dimensional small-aperture arrays analysis of ambient
vibrations in a small-size alluvial basin, Colfiorito, Italy. Bulletin
of the Seismological Society of America 96, 19151933.
Draganov D., Campman X., Thorbecke J., Verdel A. and Wapenaar
K. 2009. Subsurface structure from ambient seismic noise. 71st
EAGE meeting, Amsterdam, The Netherlands, Expanded Abstracts,
Z038.
Draganov D., Wapenaar K., Mulder W., Singer J. and Verdel A.
2007. Retreival of reflections from background-noise measurements. Geophysical Research Letters 34, L04305.
Dutta U., Satoh T., Kawase H., Sato T., Biswas N., Martirosyan A.
and Dravinski M. 2007. S-wave velocity structure of sediments
in Anchorage, Alaska, estimated with array measurements of microtremors. Bulletin of the Seismological Society of America 97,
234255.

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 23

Frehner M., Schmalholz S.M., Holzner R. and Podladchikov Y.Y.

2006. Interpretation of hydrocarbon microtremors as pore fluid
oscillations driven by ambient seismic noise. EAGE Passive Seismic
Workshop Exploration and Monitoring Applications, Dubai,
United Arab Emirates, Expanded Abstracts, A05.
Frehner M., Schmalholz S.M. and Podladchikov Y. 2009. Spectral
modification of seismic waves propagating through solids exhibiting a resonance frequency: A 1-D coupled wave propagationoscillation model. Geophysical Journal International 176, 589
600.
Frehner M., Schmalholz S.M., Podladchikov Y., Eth Z. and Oslo
U. 2007. Interaction of seismic background noise with oscillating
pore fluids causes spectral modifications of passive seismic measurements at low frequencies. 77th SEG meeting, San Antonio, Texas,
USA, 13071311.
Friedrich A., Kruger
F. and Klinge K. 1998. Ocean-generated micro
seismic noise located with the Grafenberg
array. Journal of Seis
mology 2, 4764.
Gaull B.A., Kagami H. and Taniguchi H. 1995. The microzonation of
Perth, Western Australia, using microtremor spectral ratios. Earthquake Spectra 11, 173191.
Gerstoft P., Fehler M.C. and Sabra K.G. 2006. When Katrina hit California. Geophysical Research Letters 33, L17308.
doi:10.1029/2006GL027270
Goertz A.V., Schechinger B., Koerbe M. and Krajewski P. 2009.
A low-frequency passive seismic survey in an urban setting in
Germany. 71st EAGE meeting, Amsterdam, The Netherlands, Expanded Abstracts, S039.
Graf R., Schmalholz S.M., Podladchikov Y.Y. and Saenger E.H. 2007.
Passive low frequency spectral analysis: Exploring a new field in
geophysics. World Oil 228, 4752.
Green A. and Greenhalgh S. 2009. Microtremor spectra: A proven
means for estimating resonant frequencies and S-wave velocities of
shallow soils/sediments, but a questionable tool for locating hydrocarbon reservoirs. First Break 27, 511.
Grevemeyer I., Herber R. and Essen H.H. 2000. Microseismological
evidence for a changing wave climate in the northeast Atlantic
Ocean. Nature 408, 349352.
Guillier B., Chatelain J.L., Hellel M., Machane D., Mezouer N., Ben
Salem R. and Oubaiche E.H. 2005. Smooth bumps in H/V curves
over a broad area from single-station ambient noise recordings are
meaningful and reveal the importance of Q in array processing:
The Boumerdes (Algeria) case. Geophysical Research Letters 32,
L24306. doi:10.1029/2005GL023726
Haghshenas E., Bard P.Y., Theodulidis N. and the SESAME WP04
Team. 2008. Empirical evaluation of microtremor H/V spectral
ratio. Bulletin of Earthquake Engineering 6, 75108.
Hanssen P. and Bussat S. 2008. Pitfalls in the analysis of low frequency
passive seismic data. First Break 26, 111119.
Hartzell S., Carver D., Williams R.A., Harmsen S. and Zerva A. 2003.
Site response, shallow shear-wave velocity and wave propagation
at the San Jose, California, dense seismic array. Bulletin of the
Seismological Society of America 93, 443464.
Haubrich R.A. and McCamy K. 1969. Microseisms; Coastal and
pelagic sources. Reviews of Geophysics 7, 539571.
Holzner R., Eschle P., Dangel S., Frehner M., Narayanan C. and Lakehal D. 2009. Hydrocarbon microtremors interpreted as nonlinear


C

oscillations driven by oceanic background waves. Communications

in Nonlinear Science & Numerical Simulation 14, 160173.
Holzner R., Eschle P., Dangel S. and Narayanan C. 2007a. Hydrocarbon related microtremors Verification of an analytical oscillator
model by the Navier-Stokes equations. 69th EAGE meeting, London, UK, Expanded Abstracts, P212.
Holzner R., Esehle P., Dewarrat R., Lambert M. and Graf R. 2006c.
Marine application of hydrocarbon microtremor analysis (HyMAS). 76th SEG meeting, New Orleans, Louisiana, USA, 2290
2293.
Holzner R., Eschle P., Frehner M., Schmalholz S. and Podlachikov
Y. 2006a. Hydrocarbon microtremors interpreted as oscillations
driven by oceanic background waves. 68th EAGE meeting, Vienna,
Austria, Expanded Abstracts, D036.
Holzner R., Eschle P., Frehner M., Schmalholz S. and Podladchikov
Y. 2006b. Interpretation of hydrocarbon microtremors as nonlinear oscillations driven by oceanic background waves. 76th SEG
meeting, New Orleans, Louisiana, USA, 22942298.
Holzner R., Eschle P., Meier P.F. and Dangel S. 2007b. Linear model
for low-frequency pore liquid oscillations observed in hydrocarbon
microtremor analysis (HyMAS). GeoArabia 12, 169.
Holzner R., Eschle P., Zurcher H., Graf R., Dangel S. and Meier
P.F. 2005a. Identification of hydrocarbon reservoirs by microtremor analysis (HyMAS) Successful and reproducible. 2nd
SEG/EGS/EPEX/EPA International Petroleum Conference, Cairo,
Egypt, 36.
Holzner R., Eschle P., Zurcher H., Graf R., Dangel S. and Meier
P.F. 2005b. Case study of successful identification of hydrocarbon
reservoirs by microtremor analysis (HyMAS). AAPG International
Conference, Paris, France.
Holzner R., Eschle P., Zurcher H., Lambert M., Graf R., Dangel S.
and Meier P.F. 2005c. Applying microtremor analysis to identify
hydrocarbon reservoirs. First Break 23, 4146.
Horike M., Zhao B. and Kawase H. 2001. Comparison of site response characteristics inferred from microtremors and earthquake
shear waves. Bulletin of the Seismological Society of America 91,
15261536.
Kedar S. and Webb F.H. 2005. The oceans seismic hum. Science 307,
682683.
Kind F., Faeh D. and Giardini D. 2005. Array measurements of Swave velocities from ambient vibrations. Geophysical Journal International 160, 114126.
Konno K. and Ohmachi T. 1998. Ground-motion characteristics estimated from spectral ratio between horizontal and vertical components of microtremor. Bulletin of the Seismological Society of
America 88, 228241.
Lambert M., Schmalholz S.M., Saenger E.H. and Podladchikov Y.Y.
2007. Low-frequency anomalies in spectral ratios of single-station
microtremor measurements: Observations across an oil and gas
field in Austria. 77th SEG meeting, San Antonio, Texas, USA, Expanded Abstracts, 13521356.
Lambert M.-A., Schmalholz S.M., Saenger E.H. and Steiner B.
2009a. Low-frequency microtremor anomalies at an oil and gas
field in Voitsdorf, Austria. Geophysical Prospecting 57, 393411.
doi:10.1111/j.13652478.2008.00734.x
Lambert M.A., Schmalholz S.M., Saenger E.H. and Steiner B. 2009b.
Passive seismic study at an oil and gas field in Voitsdorf, Austria.

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

24 M.Y. Ali et al.

EAGE Passive Seismic Workshop Exploration and Monitoring

Applications, Limassol, Cyprus, Expanded Abstracts, A34.
Lermo J. and Chavez-Garcia F.J. 1994. Are microtremors useful in
site response evaluation? Bulletin of the Seismological Society of
America 84, 13501364.
Liu H.-P., Boore D.M., Joyner W.B., Oppenheimer D.H., Warrick R.E., Zhang W. et al. 2000. Comparison of phase velocities
from array measurements of Rayleigh waves Associated with microtremor and results calculated from borehole shear-wave velocity
profiles. Bulletin of the Seismological Society of America 90, 666
678.
Longuet-Higgins M.S. 1950. A theory of the origin of microseisms.
Philosophical Transactions of the Royal Society of London A 243,
135.
Louie J.N. 2001. Faster, better: Shear-wave velocity to 100 meters
depth from refraction microtremor arrays. Bulletin of the Seismological Society of America 91, 347364.
Maresca R., Galluzzo D. and Del Pezzo E. 2006. H/V spectral ratios and array techniques applied to ambient noise recorded in the
Colfiorito Basin, central Italy. Bulletin of the Seismological Society
of America 96, 490505.
Marzorati S. and Bindi D. 2006. Ambient noise levels in north
central Italy. Geochemistry Geophysics Geosystems 7, Q09010,
doi:10.1029/2006GC001256.
van Mastrigt P. and Al-Dulaijan A. 2008. Seismic spectroscopy using amplified 3C geophones. 70th EAGE meeting, Rome, Italy,
Expanded Abstracts, B047.
McNamara D.E. and Buland R.P. 2004. Ambient noise levels in the
continental United States. Bulletin of the Seismological Society of
America 94, 15171527.
Nakamura Y. 1989. A method for dynamic characteristics estimation
of subsurface using microtremor on the ground surface. Quarterly
Report Railway Technical Research Institute 30, 2530.
Nguyen T.T., Lambert M., Saenger E.H., Artman B. and Schmalholz
S.M. 2009. Reduction of noise effects on low frequency passive
seismic data. 71st EAGE meeting, Amsterdam, the Netherlands,
Expanded Abstracts, S038.
Nguyen T.T., Saenger E.H., Schmalholz S.M. and Artman B. 2008.
Earthquake triggered modifications of microtremor signals above
and nearby a hydrocarbon reservoir in Voitsdorf, Austria. 70th
EAGE meeting, Rome, Italy, Expanded Abstracts, P025.
Ohori M., Nobata A. and Wakamatsu K. 2002. A Comparison of
ESAC and FK methods of estimating phase velocity using arbitrarily
shaped microtremor arrays. Bulletin of the Seismological Society of
America 92, 23232332.
Okada H. 2003. The Microtremor Survey Method. Geophysica. SEG.
ISBN 1560801204.
Parolai S., Bormann P. and Milkereit C. 2002. New relationships
between Vs, thickness of sediments, and resonance frequency calculated by the H/V ratio of seismic noise for the Cologne area
(Germany). Bulletin of the Seismological Society of America 92,
25212527.
Peterson J. 1993. Observations and modeling of seismic background noise. US Geological Survey, Open File Report, 93322,
195.
Picozzi M., Parolai S. and Richwalski S.M. 2005. Joint inversion of
H/V ratios and dispersion curves from seismic noise: Estimating


C

the S-wave velocity of bedrock. Geophysical Research Letters 32,

L11308, doi:10.1029/2005GL022878.
Rached G.R. 2006. Surface passive seismic in Kuwait. EAGE Passive
Seismic Workshop Exploration and Monitoring Applications,
Dubai, United Arab Emirates, Expanded Abstracts, A27.
Rached G.R. 2009. The challenge for surface passive seismic measurements in Kuwait. EAGE Passive Seismic Workshop Exploration and Monitoring Applications, Limassol, Cyprus, Expanded
Abstracts, A33.
Rodriguez V.H.S. and Midorikawa S. 2003. Comparison of spectral
ratio techniques for estimation of site effects using microtremor
data and earthquake motions recorded at the surface and in boreholes. Earthquake Engineering & Structural Dynamics 32, 1691
1714.
Saenger E.H., Lambert M.A., Nguyen T.T. and Schmalholz S.M.
2009a. Preliminary model of hydrocarbon reservoir related microtremors. 71st EAGE meeting, Amsterdam, the Netherlands, Expanded Abstracts, S035.
Saenger E.H., Schmalholz S.M., Lambert M.A., Nguyen T.T., Torres
A., Metzger S. et al. 2009b. A passive seismic survey over a gas field:
Analysis of low-frequency anomalies. Geophysics 74, O29O40.
Saenger E.H., Schmalholz S.M., Podladchikov Y.Y., Holzner R., Lambert M., Steiner B. et al. 2007a. Scientific strategy to explain observed spectral anomalies over hydrocarbon reservoirs generated
by microtremors. 69th EAGE meeting, London, UK, Expanded Abstracts, A033.
Saenger E.H., Torres A. and Artman B. 2009c. A low-frequency
passive seismic survey in Libya. EAGE Detective Stories Behind
Prospect Generation Workshop Challenges and The Way Forward, Muscat, Oman, Expanded Abstracts, 5028.
Saenger E.H., Torres A., Rentsch S., Lambert M., Schmalholz S.M.
and Mendez H.E. 2007b. A hydrocarbon microtremor survey over
a gas field: Identification of seismic attributes. 77th SEG meeting,
San Antonio, Texas, USA, Expanded Abstracts, 12771281.
Satoh T., Kawase H. and Matsushima S. 2001. Estimation of S-wave
velocity structures in and around the Sendai basin, Japan, using
array records of microtremors. Bulletin of the Seismological Society
of America 91, 206218.
Scherbaum F., Hinzen K.G. and Ohrnberger M. 2003. Determination
of shallow shear wave velocity profiles in the Cologne, Germany
area using ambient vibrations. Geophysical Journal International
152, 597612.
Schmalholz S.M., Podladchikov Y.Y., Holzner R. and Saenger E.H.
2006. Scientific strategy to explain observed spectral anomalies
over hydrocarbon reservoirs generated by microtremors. EAGE
Passive Seismic Workshop Exploration and Monitoring Applications, Dubai, United Arab Emirates, Expanded Abstracts, A06.
Singer J.M., Barzandji O., Leu W., Rode E.D., Akrawi K., Linthorst
S. and Dangel S. 2002. Spectroscopic identification of tremor phenomena over hydrocarbon reservoirs. 64th EAGE meeting, Florence, Italy, Expanded Abstracts, H-46.
Steiner B., Saenger E.H. and Schmalholz S.M. 2007. Time-reverse
modeling of microtremors: A potential method for hydrocarbon
reservoir localization. 77th SEG meeting, San Antonio, Texas, USA,
Expanded Abstracts, 21152119.
Steiner B., Saenger E.H. and Schmalholz S.M. 2008a. Case studies on
2D- and 3D-time reverse modeling of low-frequency microtremors

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125

Low-frequency passive seismic experiments 25

Application to reservoir localization. 70th EAGE meeting, Rome,

Italy, Expanded Abstracts, B045.
Steiner B., Saenger E.H. and Schmalholz S.M. 2008b. Time reverse
modeling of low-frequency microtremors: Application to hydrocarbon reservoir localization. Geophysical Research Letters 35,
L03307. doi:10.1029/2007GL032097
Stephenson W.J., Hartzell S., Frankel A.D., Asten M., Carver
D.L. and Kim W.Y. 2009. Site characterization for urban
seismic hazards in lower Manhattan, New York City, from
microtremor array analysis. Geophysical Research Letters 36,
L03301, doi:10.1029/2008GL036444
Tada T., Cho I. and Shinozaki Y. 2006. A two-radius circular array method; inferring phase velocities of Love waves using microtremor records. Geophysical Research Letters 33, L10303,
doi:10.1029/2006GL025722
Tada T., Cho I. and Shinozaki Y. 2007. Beyond the SPAC method:
Exploiting the wealth of circular-array methods for microtremor
exploration. Bulletin of the Seismological Society of America 97,
20802095.
Tanimoto T. 2007. Excitation of normal modes by non-linear interaction of ocean waves. Geophysical Journal International 168,
571582.
Tanimoto T., Ishimaru S. and Alvizuri C. 2006. Seasonality in particle motion of microseisms. Geophysical Journal International 166,
253266.
Tuladhar R., Yamazaki F., Warnitchai P. and Saita J. 2004. Seismic microzonation of the greater Bangkok area using microtremor


C

observations. Earthquake Engineering & Structural Dynamics 33,

211225.
Walker D. 2008. Recent developments in low frequency spectral analysis of passive seismic data. First Break 26, 6977.
Wathelet M., Jongmans D., Ohrnberger M. and Bonnefoy-Claudet
S. 2008. Array performances for ambient vibrations on a shallow
structure and consequences over Vs inversion. Journal of Seismology 12, 119.
Webb S.C. 2007. The Earths hum is driven by ocean waves over the
continental shelves. Nature 445, 754756.
Wilson D., Leon J., Aster R., Ni J., Schlue J., Grand S. et al. 2002.
Broadband seismic background noise at temporary seismic stations observed on a regional scale in the southwestern United
States. Bulletin of the Seismological Society of America 92, 3335
3342.
Withers M.M., Aster R.C., Young C.J. and Chael E.P. 1996. Highfrequency analysis of seismic background noise as a function of
wind speed and shallow depth. Bulletin of the Seismological Society
of America 86, 15071515.
Yamanaka H., Dravinski M. and Kagami H. 1993. Continuous measurements of microtremors on sediments and basement in Los Angeles, California. Bulletin of the Seismological Society of America
83, 15951609.
Young C.J., Chael E.P., Withers M.M. and Aster R.C. 1996. A comparison of the high-frequency (>1Hz) surface and subsurface noise
environment at three sites in the United States. Bulletin of the Seismological Society of America 86, 15161528.

2009 European Association of Geoscientists & Engineers, Geophysical Prospecting, 125