Dr Taka’aki Taira – Taking the Earth’s Pulse – Listening to Seismic Noise

Apr 29, 2017Earth & Environmental Sciences

Research Seismologist Dr Taka’aki Taira at the University of California at Berkeley and his colleagues investigate changes in Earth’s structures related to earthquake rupture and volcanic eruption by exploring ways to listen to and interpret ambient seismic noise.


The Earth Not Only Shouts at Us –  It Also Whispers

Earthquakes are scary things. Just the shaking and rolling of the normally stable ground in a mild temblor is enough to frighten people and animals. And the damage large earthquakes cause – including the human death toll – seems to make the news on a regular basis. For example, recently a violent earthquake struck Gorkha, Nepal, killing 9,000 people and injuring 22,000. A century ago it was the historic 1906 San Francisco earthquake – one of the largest natural disasters in the history of the United States resulting in the destruction of 80% of the city of San Francisco and the loss of over 3,000 lives – that resulted in a veritable eruption of scientific interest in earthquakes, seismic motions and methods to observe and forecast these kinds of disasters. For the last century, scientists the world over have studied the creaking and cracking of the Earth’s crust in an effort to warn us of impending catastrophes. But the big sounds that accompany such phenomena as earthquakes and volcanic eruptions are not all that you can detect with a seismograph. You can also hear background whispering of the Earth’s crust – the so-called ‘seismic noise.’

Seismic noise is the rather nonspecific term for the fairly persistent, low frequency vibration of the Earth’s crust from any number of causes. Also known as ambient vibrations, it is referred to as ‘noise’ because it is ordinarily an unwanted part of the signals recorded by seismometers. Other disciplines besides seismology also term it ‘noise’ because it is a nuisance for things that are sensitive to vibrations, like precision telescopes or the commercial growing of crystals. On the other hand, measuring ambient vibrations can be helpful in engineering, where projects such as the building of bridges and high-rise buildings require calculations of the elastic properties of the soil to determine whether the structures will be susceptible to shifting due to earthquakes and other seismic events. In other words, seismic noise may indicate whether the ground is firm enough or not. This is where Dr Taka’aki Taira and his colleagues concentrate much of their recent professional attention – studying seismic noise in relation to actual earthquakes for determining whether noise is not noise at all, but a possible indicator of stress in the tectonic plates near earthquake prone areas.

The Science of Seismic Fault Dynamics is Itself a Dynamic Science

Since the 1950s and 1960s, scientific understanding has been that the Earth’s rigid outer layer – the lithosphere – is broken up into seven or eight major tectonic plates, along with a number of smaller plates, that essentially float on the fluid inner layers of the Earth. These tectonic plates are always slowly moving, separating in some places and colliding in others. But the important thing is when they collide their edges grind together causing massive friction. Where plates collide – called a fault – the friction of that collision can cause energy to accumulate over time, which can suddenly release when the amount of built up energy overcomes the force of friction between the two edges. According to this model, an earthquake results from a sudden slip on a fault. This slip causes the edges move against each other, releasing energy in waves that travel through the Earth’s crust and causing vibrations and movement that can lead to damage and destruction. Simply speaking, the tension builds up over time until the edges of the fault can’t hold it anymore and an earthquake results. Thus scientists have been focusing on observing the deformation at the Earth’s surface to measure the total strain accumulation by this tectonic plate model. But this is not enough. We need to know when the Earth can’t endure the accumulated strain anymore, i.e., the strength of the crust.  Moreover, scientists recognise the strength might vary over time. When the crust is weakened, earthquakes would occur more frequently even though the accumulated stain is smaller. This would lead to a more complex pattern of earthquake cycles. Without knowing the crustal strength, it is extremely difficult to forecast earthquakes. Therefore, observing the temporal variations of the fault strength has been a long-sought goal of the earthquake science community over the last few decades.

Together with Paul Silver of Carnegie Institution for Science, Dr Taira and his team developed a new means to monitor fault strength by analysing seismic waveforms and microearthquake activities, and published their findings in Nature. The team found the first field evidence showing the fault strength was temporally weakened, and this temporal weakening was responsible for clusters of earthquakes in Central California. More importantly, the reduction of the fault strength they found in California was induced by dynamic stress changes from a distant 2004 magnitude 9.1 Sumatra earthquake that had occurred on the other side of the world.

The implication of their finding is that distant large earthquakes may increase the risk of subsequent earthquakes around the globe. More recently, Dr Taira has concentrated on recording ambient vibrations – generally considered nonspecific background noise – and correlating them with specific volcanic and earthquake activity.

‘Noise’ Can Be Considered a Heartbeat of Seismic Activity

In the American Pacific Northwest, Pacific Oceanic tectonic plates have slid below the North American Plate for millions of years. Heat from this tectonic subduction has given rise to numerous volcanoes from California all the way up to British Columbia over the recent geologic past – say, the last 30 million years. It is also responsible for seismic activity in the Lassen volcanic area, located at the southern edge of the Cascade Mountain Range. Here, at the Lassen Volcanic Center, is where Dr Taira and his colleagues have their seismology listening post. The Center sits above a hydrothermal system that is feared might be the site of hydrothermal explosions at some time in the future. Dr Taira monitors this area, aiming to develop a new way of forecasting volcanic phenomena by using seismic noise correlated with actual seismic and geologic activity.

Dr Taira and his colleague Florent Brenguier, of the Université Grenoble Alpes, analysed ambient vibrations at six stations in the Northern California Seismic Network around Lassen Peak, a mountain in the Lassen volcanic area. The data from these stations was electronically processed to show changes in the speed of seismic waves traveling through this area. Variations in seismic wave speed are indicative of changes in tectonic stress in the area. Essentially, they established a quasi-real-time velocity monitoring system through the use seismic interferometry with ambient vibrations. Their monitoring system showed the variability of seismic velocity over time in response to stress changes from earthquakes and from seasonal environmental changes.

Interestingly, dynamic stress changes from an actual magnitude 5.7 local earthquake produced a measurable velocity reduction 1 km below the surface. Calculations from the changes surrounding this earthquake indicated that the Lassen hydrothermal system contained highly-pressurised hydrothermal fluid deep beneath the surface. Dr Taira’s measurements also show that the long-term seismic velocity changes closely follow snow-induced vertical pressure almost immediately. That is, winter snow accumulation on the surface actually pushes down with enough force to cause changes in the seismic velocities in the subsurface hydrothermal fluid. Dr Taira feels that this is most consistent with a hydrological load model, where surface loading presses the hydrothermal fluid out, leading to an increase in the opening of cracks in the crust. The weight of snow forces fluid already below the surface to cause cracks in the various subsurface rocks and sediment. At any rate, this effect of the hydrothermal fluid movement is correlated with reductions of seismic velocity. This allowed Dr Taira and his team to deduce that heated hydrothermal fluid is responsible for the long-term changes in seismic velocity, and those changes in velocity can be used to understand – basically in real time – what is going on with the subsurface fluids. This is exciting news, giving hope that monitoring of the ambient vibrations can tell us what the fluids and rocks kilometres deep are doing. This in turn might give an early warning of volcanic or earthquake activity.

Is This an Isolated Finding?

Dr Taira’s results from the Lassen monitoring are very exciting, but they aren’t the first time he’s used ambient vibration monitoring to listen to earthquakes. Recently, in the journal Geophysical Research Letters, Dr Taira and his group published their studies of ambient vibration-based monitoring they used to look at the temporal variations of crustal seismic velocities before, during, and after the 2014 magnitude 6.0 earthquake in the South Napa area. This South Napa earthquake is the largest earthquake in the San Francisco Bay Area, since the 1989 magnitude 6.9 Loma Prieta earthquake. They saw a velocity drop immediately after the South Napa earthquake. The spatial variability of the velocity reduction correlated best with the pattern of the peak ground velocity of the South Napa mainshock. This told them that fracture damage in rocks induced by the dynamic strain is likely responsible for this velocity change. About half of the velocity reduction was recovered at the first 50 days after the South Napa mainshock.
This velocity recovery is a fascinating observation. Dr Taira believes that these findings after the earthquake may actually indicate a healing process of damaged rocks. This implies that fault lines can ‘heal’ themselves very rapidly to some extent after they have released energy in an earthquake. Again, what some folks consider noise – these ambient vibrations – is giving Dr Taira and his colleagues an important new technique to monitor seismic activity.

Future Needs and Directions?

What else can we find out about earthquakes and volcanoes by eavesdropping on them? With the advancements of computer resources and instrumentation, Dr Taira and his colleagues are pushing the limits to uncover more of these secrets. Their efforts spent watching and listening to ‘noise’ will hopefully yield more information on the genesis of earthquakes and allow us more precision in our ability to understand the underlying mechanisms of earthquakes and volcanic eruptions. However, their monitoring system is not complete. Their present system puts out a daily velocity change of the ambient vibration pattern. If they had more computing power, however, they could perform massive cross-correlation computations and perhaps give an hourly update, even streaming it in real-time online. This would enable researchers around the world to detect changes in the velocity of ‘noise’ that accompany and perhaps precede a volcanic eruption or an earthquake.

In addition, Dr Taira tells Scientia that he recently obtained a grant from National Science Foundation with Rice University and Lawrence Berkeley National Laboratory to perform an active source experiment at Parkfield, central California. Recent advancement of the instrumentation allows the team to generate and detect the seismic waves in high precision. Previous experiments carried out by another researcher team found velocity changes that preceded small earthquakes. This measurement could lead to a sort of ‘stress meter’ at greater depth to better understand how fault-zone stress is related to earthquakes. Dr Taira has recently joined this active source experiment project as a co-Principal Investigator and is planning to go into the field soon to begin monitoring. This is important research on an important subject – not noise at all. 

Meet the researcher

Dr Taka’aki Taira
Berkeley Seismological Laboratory
University of California, Berkeley


Dr Taira’s research interests include earthquake and observational seismology, transient stress changes at seismogenic depth, subsurface hydrothermal fluid migration, source mechanism of fluid-induced earthquakes, developing seismic array methodologies, seismic imaging of crustal structure, seismic wave propagation, and modeling of conduit flow dynamics. He has authored or co-authored nearly 30 articles published in peer-reviewed journals and other professional proceedings, as well as multiple oral presentations, media releases and interviews. Dr Taira is heavily involved in the operation of the Northern California Earthquake Data Center. Dr Taira’s research has been recognised by the Young Scientist Award from the Seismological Society of Japan in 2011 and the Best Young Scientist Poster Award at the 2013 International Continental Scientific Drilling Program science conference at Potsdam, Germany. He also received the nation’s top scientific honour – the Young Scientists’ Prize from the government of Japan in 2016.




T: (+1) 510 642 8504
E: taira@berkeley.edu
W: http://earthquakes.berkeley.edu/~taira



Paul G. Silver (deceased), Carnegie Institution for Science, USA

Florent Brenguier, Université Grenoble Alpes, France

Fenglin Niu, Rice University, USA

Thomas M. Daley, Lawrence Berkeley National Laboratory, USA

Robert M. Nadeau, University of California, Berkeley, USA

Douglas S. Dreger, University of California, Berkeley, USA

Roland Bürgmann, University of California, Berkeley, USA

Barbara Romanowicz, University of California, Berkeley USA

David R. Shelly, United States Geological Survey, USA

Kiyoshi Yomogida, Hokkaido University, Japan

Robert B. Smith, University of Utah, USA



National Science Foundation

United States Geological Survey

Southern California Earthquake Centre

France Berkeley Fund

University of California MEXUS (Institute for Mexico and the United States)



T Taira and F Brenguier, Response of hydrothermal system to stress transients at Lassen Volcanic Center, California, inferred from seismic interferometry with ambient noise, Earth, Planets and Space, 2016, 68, 162.

T Taira, F Brenguier and Q Kong, Ambient noise-based monitoring of seismic velocity changes associated with the 2014 Mw 6.0 South Napa earthquake, Geophysical Research Letters, 2015, 42, 6997–7004.

T Taira, DS Dreger and RM Nadeau, Rupture process for microearthquakes inferred from borehole seismic recordings, International Journal of Earth Sciences, 2015, 104, 1499–1510.

T Taira, R Bürgmann, RM Nadeau and DS Dreger, Variability of fault slip behavior along the San Andreas Fault in the San Juan Bautista Region, Journal of Geophysical Research: Solid Earth, 2014, 119, 8827–8844.

T Taira, PG Silver, F Niu and RM Nadeau, Remote triggering of faultstrength changes on the San Andreas fault at Parkfield, Nature, 2009, 461, 636–639.