Analysis of Acoustic-Seismic Coupling

In recent experiments performed at the TU Dortmund seismic and acoustic signals were measured. The soil velocity was measured with geophones, which were buried at different depths in the ground, and the sound pressure with microphones, placed at the surface above the geophones. The ratio of seismic to acoustic spectral amplitude gives the coupling coefficient between these signals, which is depending on the frequency and the elevation angle as well as on the burying depth of the seismic sensors.

A plot of the measured coupling coefficient is given below:

In the graph the coupling coefficients between a microphone at the surface and on the one hand a geophone at 0.19 m depth (red) and on the other hand a geophone at 0.51 m depth (blue) are shown. As the source a speaker was used (solid line), which produced pink noise (equal amount of sound power in each octave), at grazing incidence. The circles and triangles indicate measured signals from a helicopter, which mostly produces signals at harmonics of the rotation frequencies of its rotors.

Surface waves:During the analysis of the measurements it was found that for specific frequencies and elevation angles of the acoustic waves the seismic signal was increased significantly. This is believed to be caused by constructive interference of surface waves (i.e. Rayleigh waves) which only occurs when the effective wavelength of the incident acoustic wave (projection of the incident wave to the ground) matches the Rayleigh wavelength as shown in the sketch below:

Displaying the acoustic-seismic coupling coefficient for different frequencies and elevation angles of the incident acoustic helicopter signal we find the coupling strength is increased for certain frequencies and angles (see graph below, measured during a CTBT campaign). While the strongest acoustic signal is measured when the helicopter is directly above the sensor the seismic signal reaches its maximum a little while before the helicopter reaches the sensor and after passing it.

The investigation of the influence of acoustically induced surface waves in detail is topic of the current research project. To distinguish soil vibration caused by surface waves excited at some distance and local acoustic excitation some sensors (microphones and geophones) are shielded from the incident acoustic signal. For that we used a wooden box cased with acoustic damping material. So the acoustic waves couple to the ground outside the box and can - for certain frequencies and elevation angles - excite surface waves which can travel into and be measured inside the box as shown in the sketch below:

Microphone Geophone

While an incident acoustic wave excites soil vibration locally wherever it hits the ground (red arrows), superposition of surface waves occurs only for matches of the effective wavelength of incident wave and surface wave. The surface waves propagate and interfere and can be measured in the box, where local excitation is reduced due to acoustic damping.

The damping of the used box is frequency dependent and increases with the frequency. First evaluations show reduction of acoustic power by two orders of magnitude, while seismic power is reduced by a factor 2 to 5.

On the left a power spectrum of the sound pressure of pink noise, emitted by the speaker, is shown with applied acoustic damping (red curve) and without (black curve). Averaged values are given likewise in green and in blue. On the right a power spectrum of the acoustically induced soil velocity is shown, measured with two geophones, buried at 0.3 m each. The sensors of the red curve were placed under the acoustic damping box and the sensors of the black curve were placed outside.

Measurements

In May 2013 we performed a measurement campaign to verify the surface-wave hypothesis. The measurement took place close to the Airport Münster-Osnabrück (FMO), Germany. We installed the setup in line with the runway in 4 km distance to cover the range from grazing incidence up to incidence angles of 90° for the signals of taking off and landing jet aircraft. The turbines of jet aircraft produce broadband noise. Additionally a speaker was used to emit pink noise, pure CW sine signals and sweeps in a frequency range between 10 Hz and 5000 Hz. The speaker was mounted on a crane so the elevation angle could be varied.

Microphones and geophones were used to measure sound pressure and soil velocity. One third of the sensors were shielded by a wooden box, cased with acoustic damping material: two microphones were placed inside the box and nine geophones were buried under it in different depths between 0.15 m and 0.60 m. The box and the crane with speaker can be seen in the picture to the right.

For different measured signals the time development of the spectral power is displayed in the graphs to the left:The upper shows the sound pressure of a flyover of two jet aircraft (left and middle) which produce broadband signals by the turbines.

In the second graph the acoustically induced soil velocity is displayed for pure sine waves emitted by the speaker: The frequency of the sine was increased from 10 Hz to 100 Hz with increment of 2 Hz, from 100 Hz on the increment was 10 Hz. Every frequency was played for a duration of 10 seconds.Multiples of the 50 Hz mains hum are also visible.

Outlook

The recent measurements from May 2013 will be evaluated in the next ten months. First conclusions can be given:

The acoustic-seismic coupling coefficient is dependent on the frequency and the elevation angle of the incident signal as well as the depth in the ground.

The influence of superposing surface waves on weak seismic measurements cannot be neglected, the analysis is performed currently.

Burying seismic sensors might be an efficient way to reduce disturbing airborne signals in sensitive SAMS measurements.

Abstract:

For the verification of the Comprehensive Nuclear Test-Ban Treaty a worldwide network of sensors, the International Monitoring System (IMS), is used to detect every underground explosion with a yield of 1 kt TNT equivalent. However, the localisation of the hypocentre of a suspected nuclear explosion with the IMS has an uncertainty of approximately 10 km. So according to the treaty text an on-site inspection (OSI) can be performed to locate the hypocentre with a precision of about 0.1 km. One method for an OSI is the setup of a local seismic network in the inspection area, the seismic aftershock monitoring system (SAMS), to detect aftershock events caused as a consequence of the explosion. These events show a very weak magnitude so man-made noise can disturb the SAMS measurements.In a project for the Research Award for Young Scientists and Engineers we analyse airborne acoustic signals which excite soil vibrations when coupling into the ground that can disturb SAMS measurements. Such signals can be caused by vehicles or helicopters, used by inspectors during an OSI. The analysed signals are broadband (measurements of jet aircraft) and periodic continuous wave (CW) signals, produced by a speaker. The research is focused on surface waves, excited by acoustic-seismic coupling in the surroundings of a sensor. The measured seismic signal is increased for constructive interference of the surface waves at the position of the sensor, which can occur for specific acoustic frequencies and elevation angles of the signal. The analysis shall give recom-mendations for an OSI to prevent or reduce such disturbing signals.

References

1 J. Altmann, F. Gorschlüter, Removing Periodic Noise to Detect Weak Impulse Events, poster presented at International Scientific Studies (ISS 09) Conference, Vienna, 10-12 June 2009.2 F. Gorschlüter, J. Altmann, Suppression of Periodic Disturbances in Seismic Aftershock Recordings for Better Localisation of Underground Explosions, Pure and Applied Geophysics, SP Birkhäuser Verlag Basel, p. 1-13,

http://dx.doi.org/10.1007/s00024-012-0617-y .3 Altmann, Jürgen; Gorschlüter, Felix; Liebsch, Mattes: Investigations of Periodic Disturbances on Seimsic Aftershock Recordings, poster presented at EGU General Assembly 2012, Vienna 22-27 April 2012.

In this graph the acoustically induced soil velocity is shown for the speaker emitting pink noise three times for 30 seconds each (broadband signals). The low frequency disturbances between the pink noise events are most likely caused by adjustment work at the crane. In the middle of the graph white lines of decreasing frequency are visible: They correspond to the overflight of a propeller aircraft

Questions for pending evaluation

The analysis of the measured data has just started, several points will be in the focus of the ongoing research:

Positioning of seismic sensors: In which depth are signals coupled to the soil on the surface detectable? Is there an optimal depth to bury seismic sensors to reduce such disturbances? Is shielding of seismic sensors from airborne sound helpful to reduce disturbances?

Frequency dependency: Which frequencies are transmitted through acoustic-seismic coupling? Which frequencies can possibly excite coherent surface waves? What are the characteristics and differences between broadband and periodic incident signals?

Surface waves: What is the influence of surface waves on measurements of acoustically induced weak seismic events? Can they be distinguished from locally excited soil vibrations?

Elevation angle: If surface waves are excited, for which angles of incidence of acoustic waves?

1cos(α)

λairλ soil ~α

λ soil

λair

air

soil

Incident acousticSignal

Propagating surface wave

Cou

plin

g C

oeff

/ (m

/s)/P

a

Acoustic−Seismic Coupling Coefficient :

C ( f ,α)=Aseismic( f ,α)Aacoustic( f ,α)

Acoustic-Seismic Coupling in Porous Ground - Measurements and Analysis for OSI SupportMattes Liebsch, Jürgen Altmann

Experimentelle Physik III, Technische Universität Dortmund, D-44221 Dortmund, Germany

T3 - P7

Two taking off jet aircraft, sound pressure

Pure CW sine waves, soil velocity