Subsurface mapping by ambient noise tomography

Subsurface Mapping by Ambient Noise Tomography

Importance of Subsurface mapping

Most of the geological maps allows for understanding about the local lithology and geological structures which is important for geological interpretations. To further support geological research and engineering work, understanding the subsurface lithology and structures underneath is also necessary. Borehole drilling is one of the common and traditional methods of exploring the subsurface, but it has several limitations. Other than being invasive to the ground surface, only small-scale structures can be found via borehole drilling. Countering these limitations, geophysical survey becomes a non-invasive practical alternative.

Ambient Noise as Alternative Geophysical Survey Method

Major geophysical survey techniques include electrical resistivity, gravity anomaly and seismic. Among all techniques, seismic survey is often used on the detection of subsurface structures, which is possible by correlating the velocity anomaly with the geological structures. Both active and passive seismic source could be used, while earthquake is one of the major passive seismic sources. For regions with frequent earthquake (seismic active regions), the analysis of seismic source will be easier. For some seismic inactive regions like Korea Peninsula, however, the seismic noise analysis is more difficult. Instead of the correlation from strong seismic sources like earthquake, the potential of weak ambient noise on subsurface structure mapping is discovered in recent years.

Source of Ambient Noise

Data collection of ambient noise is the prior stage of subsurface mapping, which is important for further analysis and correlation. The seismic noise can be transmitted by either body wave (P-wave, S-wave) or surface wave (Rayleigh wave, Love wave). The seismic source can be

For conducting seismic research and exploration, active seismic sources would be created intentionally to record the velocity change of seismic waves. Some examples of tools creating active seismic source include hammer[note1], airgun, and even artificial explosion, which may create seismic waves with similar magnitude as large earthquakes. Other than the artificial seismic sources, passive ambient noise can also be recorded and analysed. Ambient noise refers to the background noise originated either from natural events or anthropogenic activities. The use of ambient noise on velocity structure modelling has received more attention, especially in seismically inactive regions.

Nature of Ambient Noise

From an extensive range of frequency, ambient noise can be further classified into several categories, which are based on their origins.

Anthropogenic

The anthropogenic ambient noise, excluded those artificial seismic sources produced intentionally for research, are originated from human activities. Considering the ocean ambient noise source as an example, there are noises that are created unintentionally by human activities, such as shipping and offshore engineering work[ref2]. The importance of shipping reflects on the well-developed trading and commercial industries. International immigration and emigration of products and goods can be done via shipping. The shipping activities are thus becoming frequent. During the shipping process, the mechanical waves can be driven up along water surface and propagate through the ocean. Other than shipping, offshore engineering work can also produce surface waves. Engineering works, which are usually done on the continent to usually fulfil the demand of urban development, include but not limited to borehole drilling, foundation construction and geophysical surveys. Extended from the continent, reclamation has been actively carried out by many countries to create more land for urban development. Those engineering works can thus also be carried out offshore. The processes of offshore drilling and exploration create continuous mechanical waves that can also propagate through ocean.

Regarding the continental urban areas, there are more examples of human activities creating the background noise. Other than the engineering works, the urban traffic is the major component of urban ambient noise. Although the mechanical waves of the continent are not as visible than those from the ocean, they can still be transmitted via the soil and rock layers. Cars travelling on the road can produce repeatable vibration on the road which can then be transmitted through the soil layers.

Natural Noise

Natural ambient noise refers to the background noise produced from the natural events. Our natural environment is not stationary but constantly changing every moment because the nature itself is continuously modified by the weather, tectonic movements and biogenic activities. They can also produce low frequency background noise that can be further analysed. Some of the most significant events are listed below.

Earthquake, as mentioned before, is one of the most remarkable events that can produce the most significant background noise. Earthquakes can be caused by the movement along two faults or plates. Seismic waves are released and penetrate through the soil and rock layers, causing the continents to shake. There are numerous earthquakes happened every day with different scales around the world. By recording the seismic waves produced from the earthquake, we can further understand the interior structure of the Earth.

The ocean wave is another possible natural ambient noise source. As mentioned before, the ocean wave can be produced by anthropogenic activities, such as commercial shipping and offshore engineering work. Besides, natural wind and marine animals can also induce weak ocean wave propagating through the ocean, which can be recorded by seismometers.

Other than the natural ambient noise propagated through solid and liquid, ambient noise propagated via air can also be recorded if it is strong enough. One of the examples is typhoon. Typhoon is one of the natural hazards whose frequency and strength are magnified by global warming. The strong thermal current creates a low-pressure zone where the air can circulate around the centre. In such case, the vibration of the air is strong enough to be recorded by seismometers.

Variation of ambient noise

To evaluate whether the collected ambient noise source can be further analysed, we must carefully consider if there are any regular variations or patterns of certain ambient noise source. Referring to the urban noise source, it may experience a daily variation, where the human activities are conducted mostly in daytime and reduced in nighttime. The ambient noise should thus increase in daytime while reduce in nighttime. Apart from the temporal variation, the spatial variation can also matter. For example, the commercial shipping is usually concentrated on certain routes. The corresponding amplitude of ambient noise should also decrease when moving away from the shipping routes. Nevertheless, it is still difficult to distinguish the ambient noise sources.

Seismic Velocity Structure Modelling

The collected ambient noise can be further analysed to produce velocity structure maps which are used to correlated with the possible subsurface structures. The processes involved are complex and require multiple mathematical calculations.

Data collection method

As mentioned before, the data collection of ambient noise is the primary step of any seismic research. The most common tool for seismic data collection is the seismometer. Other example could be the geophones. There are also seismic stations or observatories authorized by different official bodies. For example, Hong Kong Observatory has set up several seismic stations in different locations in Hong Kong. The seismic waves are recorded by the seismometers and shown as seismographs. Semimoist and geophysicists can then identify the arrival time of different body waves and surface waves. The seismic waves usually arrive in the order of P-wave, S-wave, Rayleigh wave, and Love wave. Nevertheless, the analysis of ambient noise is more difficult than simply identify the waves above.

Ambient Noise Data Processing

Compared the seismographs of ambient noise with those recording active seismic sources, or simply reviewed any seismographs, you can discover the 'thickness' of the seismic waves. Despite the occasional increase in amplitude due to the active seismic source, the entire seismic waves are also in certain amplitudes, showing that there are certain activities that also produce weak seismic waves. The ambient noise seismographs are required to correlate together such that the velocity map can be produced.

Cross-correlation of ambient noise

Cross-correlation is a widely used tool in several aspects. It is used to discover the time difference between two events and the factors behind. In seismology, cross correlation is also used in analysing the ambient noise. The basic idea of ambient noise cross correlation is to find the green's function of the two seismic ambient noise. Imagine if there are two seismic stations with one seismic source. The seismic wave will arrive both seismic stations at different time if their distances from the seismic sources are different. Seismologists try to find the travelling time lag between two stations. Expanding the sample into an area with multiple ambient noise sources, the seismic waves from the two stations can be cross-correlated if the source distribution is even across the whole sample area. Otherwise, the green's function will be uneven and difficult to correlate any relationship between two factors.

Inversion

Inversion is one of the techniques used in ambient noise tomography. Simply speaking, inversion of a function refers to finding the original parameters that output the function itself. In ambient noise tomography, inversion of the cross-correlation function is an important step to obtain the subsurface velocity structure. The cross-correlated seismic waves can be inverted either linearly or non-linearly.

Subsurface structure correlation with velocity imaging

From the collection of ambient noise data to cross-correlation and inversion, a velocity image would be the finally produced. Figure x from Sager et. al (2018) shows the velocity model. As shown by the figure, the red parts refer to relatively low velocity zones while the blue parts show the relatively high velocity zones. This series of diagrams shows how the complexity of the waveforms could affect the quality of the image. For image created with higher complexity, the number of velocity zones increases, showing the higher quality of the image. This can show more small-scale features.

Linkage of Geological Structure with Velocity zone

Before interpreting the velocity zone, it is necessary to understand how the seismic velocity varies. In general, P and S wave travel faster in high density medium. Only P wave can travel through any medium while S wave can only travel through solid. Therefore, low velocity zone can refer to some vacuum space in the subsurface layer, such as void space and faults. Conversely, the high velocity zone may refer to the lithology with closely packed rocks, such as igneous rock. To correlate the velocity zone with geological structure, it is necessary to consider the size and shape of the velocity zones, and more importantly, the resolution of the subsurface velocity image. The resolution of the image can affect the scale of the subsurface we can interpret. Sometimes, fieldwork is also needed in order to better correlate the velocity map.

Example of Subsurface Structure

Here are some figures showing the related subsurface structures with the corresponding velocity map.