The Earth Story

Peak Ground Acceleration When you look at a seismogram, a classic one, what are the units on the vertical axis? That’s the earthquake, right? But what are we actually measuring?

explainxkcd.com/wiki/index.php/2344:_26-Second_Pulse

Earthquake in Japan

On November 22 local time, a large earthquake struck off the coast of Japan. The reported Moment Magnitude of this quake was 6.9, indicating a substantial energy release, but because the quake occurred tens of kilometers offshore no portion of land received more than light shaking. This video clip was shared automatically online after the quake – the feed links to the seismic network that monitors Japan and automatically records and shares clips and charts showing peak ground acceleration during the quake.

Peak ground acceleration is another measure of the intensity of the shaking that is distinct from “Moment Magnitude” – the magnitude scale is an attempt to measure the total energy released of the quake, the Peak Ground Acceleration is a property that more directly affects structures and people. You don’t feel “magnitude” of a quake, you feel Peak Ground Acceleration. The colors of this clip show how intense that peak ground acceleration was – light shaking at most, gradually moving across the country. The background motion before the quake reflects the slight movement any seismic instrument feels – it could be automobiles, small earthquakes, wind, waves, etc.

This quake happened off the coastline of Fukushima province – the same province devastated by the 2011 earthquake and tsunami. This quake produced a small tsunami as well; it occurred on a similar type of fault and likely produced uplift of the seafloor. Damage, even to the crippled Fukushima nuclear reactor site, from this quake is limited.

Perhaps the most interesting note regarding this quake is that it likely counts as an aftershock of the 2011 event. The technical definition of an aftershock I was taught is a quake occurring within 2 fault rupture distances before seismic activity returns to background levels. Most faults have a normal amount of “background seismicity” – small to medium sized quakes, most too small to feel, that occur during a given time period. When a major earthquake occurs, it shifts stresses throughout a region and local faults can respond by producing quakes up to several years afterwards. This surge of earthquakes is defined as the aftershock sequence. With a large quake like this one it is difficult to say that it’s a true aftershock, meaning that it couldn’t have happened without the 2011 monster, and debating whether it fits the definition is something of an academic exercise, but I’d say this quake does show how long aftershock sequences can last after major quakes. It will be several more years before seismicity at this fault returns to levels seen before the 2011 quake, and this rupture was a reminder of that.

-JBB

Video Source: https://twitter.com/EqAlarm/status/800811671757340672 https://twitter.com/nippon_en/status/800814682827128832

Reference: http://bit.ly/2gHJQgi http://on.doi.gov/2gbT7Me

What is the difference between the modern Moment Magnitude scale for measuring earthquake energy and the classic idea for the scale defined by Caltech Seismologist Charles Richter? This video explains why the scales were updated and how they compare.

Moulin Seismics

Moulins are deep vertical shafts on a glacier/ice-sheets that helps drain meltwater from the surface. Meltwater streams and supraglacial lakes (lakes on the surface of the glacier/ice-sheet) drain their water into moulins and make their way right to the bottom where they help lubricate the path of the glacier, thereby accelerating its flow (watch this clip for an example - http://bit.ly/2coiqHL). Since these drainage systems directly influence the basal motion of glaciers and ice-sheets an understanding of these systems are essential for calculating glacier mass balances and thereby global sea level rise.

The geometry of moulins and their changes owing to continuous meltwater flow are barely known since they are difficult to investigate due to their accessibility. However, recently for the first time, scientists working on the Russell glacier in Greenland have used seismics to understand the geometry of these structures. The vibrations emitted as large amounts of water draining through a moulin can be measured as ground displacement by an array seismometers installed at the surface (along with a few seismometers lowered into boreholes). The analysis of the seismic wavefield can help constrain the geometry of these structures.

The walls of the moulins act as resonators to the acoustic waves produced by the inflowing water as they reach the collected water at the base. The tremor waveform has two different sources: one from the bottom of the resonator, and the other from the solid icewalls. The waveform patterns were similar to the resonance from an organ pipe that is plugged at one end. Just as the resonance from an organ pipe depends on the length of the tube, the resonance in the moulin depends on the level of water collected at the base.

Image Credit: Alex Buisse - http://bit.ly/2cow1Ox Ian Joughin - http://bit.ly/2ct2MM4 Source: http://bit.ly/2cc2ngo Original Paper: http://bit.ly/2cWvHtd

Geology professor at Yale telling the story of how Connecticut’s geology tracks back hundreds of millions of years through multiple cycles of continent collisions, and also showing deployment of seismic instruments that can study that (including students in the process).

Landslides leave a seismic signature

Scientists have long used seismometers to gain information about earthquakes, deciphering what happened and even estimating possible damage. But, that’s not all they can do. If you’ve ever walked past one (or, well, jumped around in front of one to watch your own seismic signal…and yeah, go do that if you haven’t, check the links at the bottom for more), you know that seismometers pick up many signals, even from small movements like people or trucks moving by.

Modern seismometers are programmed to recognize these small shakings and filter them out in the hunt for the signals from actual earthquakes. But what if they were filtering out information on a threat to people which could be detected, like a landslide?

Large landslides shake the ground enough to give substantial seismic signals, but most landslide signals aren’t recognized by seismometers. New research in the journal Science seeks to change that.

The researchers, led by Lamont-Doherty Earth Observatory, took seismic data on the largest landslides and found trends within the data. Just like seismologists can classify earthquakes and tell where they happen based on seismic data, these researchers found they could interpret landslides in much the same way.

The different portions of a landslide: breaking away, traveling down a steep slope, shallowing and traveling across a flat area, landing in a ravine, all give distinct seismic signatures. The recorded seismic data even could give the size of the landslide, all in real time.

The group even found that what was once thought to be a single slide in India based on the deposit was actually a series of seven distinct landslides over four days.

Some obvious future applications could be determining areas where a slide has blocked a river, threatening commerce or creating a flash flood risk or in determining where to send resources after a major storm.

Landslides kill thousands of people per year. Taking this data and turning it into an accurate warning system or helping further understanding of the conditions that give rise to a slide will be challenging, but if the seismic stations are already in an area, or the area is very prone to landslides, this research could help develop systems that will save lives worldwide.

-JBB

ScienceNOWarticle: http://news.sciencemag.org/sciencenow/2013/03/the-shaky-side-of-landslides.html

Original paper: http://www.sciencemag.org/content/339/6126/1416

Perspectives article: http://www.sciencemag.org/content/339/6126/1395.summary

Landslide in the Philippines, image source: http://thewatchers.adorraeli.com/2011/07/28/philippines-on-alert-for-landslides/

Landslide death toll, from Nature: http://www.nature.com/news/death-toll-from-landslides-vastly-underestimated-1.11140

Turn your computer into a seismometer: http://qcn.stanford.edu/sensor/

Earthscope: Epic Science

The Earthscope is one of the biggest scientific projects on the planet, what Popular Science called the #1 most epic experiment in the universe. There are multiple arms to the project, but the most expansive is called the Transportable Array, a suite of 400 seismometers that has rolled east across the US over the last 8 years. TA stations were deployed for 2 years in a north-south stripe across the country, then picked up and moved east to the next footprint and redeployed. The result is unprecedented seismic coverage of the North American continent. The mission of the project is to use this massive seismic array and other data sets such as GPS and satellite data as a telescope pointed into the Earth, exploring the structure of the continent and underlying mantle.

The big idea behind the project is to use North America as a "natural laboratory" to study the entire cycle of continental growth, modification, and breakup. Several hundred papers have been published based on Earthscope data so far, including one that used TA seismic data to image a previously unknown, massive magma chamber under the Yellowstone hotspot. The Earthscope team also produces a wide range of educational resources and data products; I particularly like the "Birthquake" app, which gives you a snapshot of the earthquakes that happened on your birthdate. Also enthralling are the ground motion visualizations produced from TA data which show earthquake waves sweeping across the country.

The TA reached the East Coast in 2013, and is now moving to Alaska - an even more daunting task. The array will continue to record until 2018; it will leave behind a number of legacy stations which have been adopted by local institutions, permanently improving the state of seismic coverage in the US. -CEL

Images: "This/these images were created by the EarthScope National Office, a National Science Foundation funded project. EarthScope scientists study the structure and evolution of the North American continent using three primary observatories, the Plate Boundary Observatory (PBO), USArray, and the San Andreas Observatory at Depth (SAFOD). For more information, visit www.earthscope.org"

Sources: earthscope.org usarray.org http://bit.ly/1QwJPoz Ground motion visualizations: http://bit.ly/1ORDh3Z Birthquake: earthscope.org/birthquake

Multiquake

A geologist working in the field can, with a bit of training, spot a number of faults in a single outcrop. This annotated image shows multiple faults in green lines with red and purple lines marking the boundary of distinct layers that have moved along those faults.

In the field, we can see the deformation from each fault. However, multiple faults in an area poses an interesting thought experiment, what would happen if two faults moved at once or almost at once? How would you recognize the earthquake generated from each fault?

This question is not an abstract one; earlier this week, deep beneath South America, a pair of earthquakes occurred within only a few seconds of each other. Each quake produced the standard sequence of waves that occurs during an earthquake and in this case the quakes were far enough apart in time that nearby seismometers could recognize the distinct pulses from each.

However, this will not always be the case. Large earthquakes on megathrust faults in the ocean release huge amounts of energy and they are also surrounded by smaller faults. If a megathrust fault ruptures, it can trigger earthquakes on these smaller faults and unless there is a seismometer very near the site of the second quake, its waves will be overwhelmed by the waves of the larger quake.

This scenario is somewhat analogous to trying to pick out a single voice in a crowded room – known to neural scientists as the “Cocktail party” problem. Our brains have figured out how to recognize the pattern of waves associated with specific voices even if they are swamped out by louder noises. Can we do something similar with earthquakes?

A pair of scientists from the University of Liverpool recently argued this was exactly the case for a 2011 earthquake off the coast of Chile.

In 2011, a magnitude 7.1 earthquake took place on a section of the Chilean Subduction zone. Usually after an earthquake there are a series of aftershocks in the area of the fault that ruptured, but the aftershock pattern for this quake was unusual. The computer-derived location of the fault rupture had fewer aftershocks than two nearby areas, one about 10 km from the calculated source and another cluster about 30-40 km away.

The researchers also looked at the exact shape of the waves measured by seismometers and found that there were a number of patterns that couldn’t be well explained by a single earthquake.

Following the geological evidence, they told their computers to simulate the type of waves produced if the first quake rapidly triggered a second quake at the site of the extra aftershocks. They found that if they considered a second quake at the second location occurring just as the waves from the first quake arrived, they could explain the exact shape of the seismometer signal. In other words, the first quake triggered a second quake at almost the same time, and the second quake was buried in the first.

The scientists were able to pull apart the signals from these two quakes because they had the aftershock series and because this portion of South America has a number of seismometers that record earthquakes. However, there are important implications for hazard assessment worldwide.

When an earthquake occurs, computers rapidly process the data to estimate the strength and location of the quake. That information is then used in the emergency response, which is typically most necessary near the earthquake epicenter. However, if a second earthquake is buried in the first and the computer can’t recognize it, the computer could misidentify the location of the quake by tens of kilometers.

Figuring out routines to allow computers to identify the second quake in the “crowded room” scenario is beyond my expertise, but thanks to the work of these scientists its now a problem we realize needs solved. 10-20 kilometers of error might not seem like a lot, but if a local tsunami wave is generated, 10 kilometers can be the difference between putting a warning in the right spot or completely missing the location of greatest tsunami risk.

We can figure out with our eyes how to identify multiple faults in this picture, our ears can figure out how to identify a voice in a crowded room, now we need to teach computers to do that problem rapidly with earthquakes because it can help save lives.

-JBB

Image credit: https://flic.kr/p/dfhDtC (CC licensed)

Original paper: http://bit.ly/1Spo1tW

Crowded room problem http://n.pr/1NUQPXm

Multiple Peru quakes this week: http://bit.ly/1NBfBS6

This video was taken in a USGS server room during the Gorkha quake in Nepal. It’s in the same site as an active seismometer that recorded the earthquake, meaning you can watch ground motion as measured by seismometer at the same time you see what is happening in the room.  Data comes from the PEER center at UC Berkeley.

Dragons, Frogs and Earthquakes

In the aftermath of the Nepal earthquake it wasn’t long before the news of the disaster had travelled globally, and aid began being sent to help those in need. In this day and age Japan even has a text alert for Tsunami warnings, which was invaluable and saved many lives in 2011.

Obviously, this ability to send information around the world at lightning speed hasn't always been the case. Previously, in large countries where the population density is sparse, it could take days to weeks for news of such a disaster to reach those capable of providing large scale aid.

In ancient China a brilliant scientist called Zhang Heng came up with a solution, he created a machine that could tell officials when and in which direction an earthquake had struck. In 132 AD Zhang produced a machine called the “Dragon Jar”.

It was a jar roughly 1 metre wide (3 feet) and 1.7 metres tall (5.5feet) that had eight dragon heads evenly spaced around the edge facing outwards. The heads represented the points of a compass, and in each of their mouths they held a ball bearing. Beneath the dragons, at the base of the jar, were 8 frogs each facing towards one of the dragons with their heads tilted upwards and their mouths open. The jars contents are unknown although many believe there to be some form of pendulum at its centre.

When an earthquake struck, even if too weak for a human to feel, the pendulum would swing with the grounds movement. This would result in one of the dragons dropping the ball into a frog’s mouth, giving an indication of the direction that the tremor had come from. It would also alert any attendants to the earthquake as the ball would make a loud noise as it landed in the frog’s mouth. With this knowledge search parties and aid could be sent out in that direction, leading to a much swifter response and the saving of many people’s lives. It is proposed that the device alerted officials to an earthquake 400miles away that had been too weak for them to feel but had been 7.0 on the Richter scale.

While Zhang was praised for his invention, his staunch belief that the tremors were a natural phenomenon and not an act of an angry god prevented him from rising through Chinese society. Unfortunately, none of his devices have been found and notes written by Chang are brief and incomplete. However, while we may not know how it worked, the invention was still a marvel of science that is impressive even today!

Reference: http://on.doi.gov/1EcsAC7

Further Reading: http://bit.ly/1EwfJfj

Image Credit (of a replica): Houfeng Didong

Reading a seismogram This is the actual seismogram recorded by a seismometer during a magnitude 5.0 earthquake on Jan Mayen island in 1995. There’s a lot of information in a graph like this if you know how to read it. A classic seismogram is made by a pen held by a weight in a fixed position on top of a spinning roll of paper. During a quake, the inertia of the weight keeps the pen from moving, while the paper drum moves due to the waves. The pen then marks the drum as it moves back and forth. Modern instruments detect these waves electronically, but the principle is the same. In this plot, time moves from left to right; the left side starts and the right side finishes. 3 different features stand out. The first hint of the quake is a small pulse of energy that decays away with some fine structure to it. Then, there’s a second pulse of energy, slightly less intense than the first. Just after this second pulse is when the biggest, most intense shaking starts. This is the classic sequence of waves created in an earthquake. The fastest moving waves are called p-waves, a pressure wave moving through the earth. The next to arrive is an s-wave, a shear wave also moving through the earth. Finally, the largest pulse of energy arrives in the form of surface waves known as Rayleigh and Love waves. The surface waves are where the largest ground motions take place and when most of the damage is done. They also have a lower frequency than the p and s waves, meaning that the wave arrivals become more spread out. The time between waves arriving at a seismic station depends on how far the station is from the quake. The farther away the station is, the bigger the difference in arrival times. The difference between the p wave arrival and surface wave arrivals can be as little as a fraction of a second for earthquakes very nearby and as long as tens of minutes for earthquakes far away.  Each wave bounces and refracts off slightly different regions within the earth, leading to the complex structure in each portion. That complexity can be used by computers as a way of measuring actual differences inside the Earth.  -JBB Image credit:  https://www.flickr.com/photos/rockbandit/309794495 Read more: http://www.geo.mtu.edu/UPSeis/reading.html http://tinyurl.com/o74xor4