19.6: Preparing for the Worst

Is it Possible to Predict Earthquakes?

To quote the USGS, “No.”

Crazed Cats, Earthquake Weather, and Seismic Statistics

When most people think of earthquake prediction, they imagine something like the daily weather forecast they might listen to as they dress for work. Within a reasonable margin of error, meteorologists are able to predict if one is going to need an umbrella that particular day.

Earthquake prediction is not like that. In fact, it’s never going to be like that, though not for lack of trying. Over decades, seismologists have thoroughly researched ideas about ways that the earth might signal a major snap is about to happen. One idea is the measurement of radon gas; if there is a sudden release of radon gas, does this mean it is being liberated by small subterranean fractures forming just before an earthquake? Possibly, but not proven. If groundwater levels abruptly shift, does this mean a quake is imminent? Possibly, but not proven. If local electromagnetic fields fluctuate, could this be a sign? I think you can predict where this is going.

No single or combined set of these kinds of possible indicators has been found to be a reliable barometer for the imminent arrival of a quake. Sometimes radon increases in well water and then a quake happens; other times, no change in radon occurs and then a quake happens. It’s inconsistent.

One common misconception is that animal behavior can indicate a quake’s approach. The idea here is that common household pets possess sensory apparatus unknown to science capable of detecting hitherto undiscovered indicators of the impending arrival of an earthquake. In this way of thinking, when cats run around the house chasing invisible things, this indicates something seismic is about to happen.

The problem with this kind of thinking is that cats are always running around the house chasing invisible things. It’s only after a major quake that a pet owner might remember the feline behavior and think, “Yesterday Mr. Whiskers was acting strange. He must have known the quake was coming!” But there is no connection between a cat’s behavior and a quake; even if one particular cat played with unusual friskiness just before a quake, plenty of other cats slept in the sun, not to mention the legions of tired dogs snoring on the couches of owners who do not allow them on the furniture. None of these events are related to each other. This pareidolia can be considered an error of patternicity, which Michael Shermer defines as imagining a meaningful pattern in meaningless noise.

Another mistaken idea about earthquake prediction is something termed “earthquake weather.” This is usually defined as unseasonably warm and calm days. But there is no pattern of quakes happening more often on warm days, or cold days, or rainy days, or dry days. Understand that in California, most quakes originate many kilometers deep. Surface conditions affect only a few centimeters of the surface rocks and soils, so there is no mechanism for surface weather to have any effect on kilometers-deep rocks.

What we can do instead is talk about earthquake probabilities. For example, we can say that given studies of past quake events, we can make reasonable estimates of the likelihood of repeat (see Video \(\PageIndex{1}\) and Video \(\PageIndex{2}\) below).

Video \(\PageIndex{1}\): Earthquake Prediction 1/2

Watch the following video about earthquake prediction.

Video \(\PageIndex{2}\): Earthquake Prediction 2/2

Watch the following video to learn more about the dangers of rip currents and how to spot them.

In terms of earthquake probability, geologists have developed techniques in the specialty of paleoseismology that allow detailed dating of when past quakes have happened. When several identified quakes can be dated over a span of time, we can estimate a recurrence interval, or the statistical time between major quakes in a particular location.

One way this works is to dig a deep trench across an active fault, wait for the soil to dry, then scrape a smooth surface on the trench sides. Seismic disruptions in the soil may reveal themselves. Often these disruptions have organic material to be collected and carbon dated, providing detailed timing of the earthquakes. Such trench studies are very valuable for their high resolution of timing and solid evidence for recurrence intervals. One such paleoseismic investigation was led by the USGS’s Jim Lienkaemper, who authored a series of papers using trench data collected at Tyson’s Lagoon, in Fremont, about the Hayward fault’s activity. These studies found evidence for 12 major quakes over the previous 1900 years, with a recurrence interval of 165 ± 65 years. That timing helped quantify our understanding that the Hayward fault is due for a major shake.

Lessons from the Parkfield Experiment

Imagine if one location in Central California had substantial earthquakes on a regular basis–say, every 22 years on average. What if scientists descended upon this place to install arrays of instruments to measure, poke, and prod the ground, its water, its gases, its electrical characteristics–in short, everything you could think of that might be an indicator of a coming quake?

What if you installed all this… and then the quake did not come?

This situation describes the Parkfield experiment, a project begun in the mid 1980s. Magnitude 6 quakes had occurred in this small town in 1857, 1881, 1922, 1934, and 1966. This averages to 22 years between quakes, with a range of 12-32 years apart. This regularity led researchers to conclude that if they installed a diverse array of sensors, they should be able to detect something changing before, during, and after the next quake.

It was calculated with 90-95% confidence that the next Parkfield quake should occur before 1993. But this first year of the Clinton presidency came and went–no quake. Nineteen ninety-four came and went, as did the next year, and the next. The expense of maintaining the equipment and labor to keep this project going became problematic. The dot.com boom and the inevitable dot.com bust happened–still no Parkfield quake. The terrorist attacks of 9/11 occurred and Parkfield remained frustratingly quiet. How long do you keep this going–and how can the project be abandoned, after all the expense of setting it up?

Finally, on 28 September 2004, Parkfield shook with a M_w6.0, as shown below in Video \(\PageIndex{3}\):

Video \(\PageIndex{3}\): Parkfield quake

A video capturing the long-anticipated 2004 Parkfield quake [no audio].

Scientists eagerly retrieved and analyzed the data from the dense sensor arrays. Now we could see what was going on in the weeks, days, and hours before the quake happened. Was there any clear signal of the impending shake? Something that could then help us recognize similar changes in other faults and advance us closer to the goal of predicting quakes as we predict the weather? Did Parkfield reveal a method to predict earthquakes?

Parkfield did not. Nothing significantly unusual or anomalous was detected. The quake just happened, apparently without a detectable precursor.

The real lesson of the Parkfield experiment, then, is that we should not expect to ever be able to predict a quake the way we predict weather. We probably will never predict the date and time, location, and magnitude of an impending quake. While we cannot, of course, dismiss the possibility that future research may uncover some previously unimagined way to predict earthquakes, the main lesson of the Parkfield experiment is one of limitations.

But this doesn’t mean we’re helpless. We can determine meaningful earthquake probabilities. We can definitively say, for example, that tomorrow there will be several minor quakes in California, even if we don’t know the exact place or magnitude. We can detail where earthquakes have happened in the past, and from that, know that these areas have a higher probability of being future locations. We can perform trench studies to determine recurrence intervals and from that understand future risks.

We can’t say the time and date, place, and magnitude of a future quake. But we can give ranges of time in years, approximate areas, and based on past quakes, expected magnitudes. We can say, for example, that there is a high risk along the Hayward Fault, because we know the history of where and how frequently it has snapped. That’s different from a weather forecast that tells you if you’re going to need an umbrella on a particular day; however, using probabilities we can say that umbrellas are more likely to be needed in December than in July, and that information can be useful too.

Can We Have an Early Warning System?

Though California has long been an international hub of earthquake research, the state did not implement an earthquake early warning system until 2019, nearly three decades after the first early warning system was introduced in Mexico City. Early warning systems in countries like Mexico and Japan are designed to detect large, subduction-created quakes that occur far offshore. Quake waves travel on the order of the speed of sound, so detection of a distant quake could in these situations provide significant time between detection and alert and the arrival of the first seismic waves. This situation is similar to that of the Cascadia Subduction Zone, and indeed ShakeAlert, an early warning system, is likely to be very effective in Del Norte and Humboldt Counties, but most of California’s population and critical infrastructure are found along the transform boundary, not the subduction zone.

The idea behind an early warning system is that once a quake is detected, the system sends an alert to cell phones in affected regions. Earthquake waves move about the speed of sound, while electronic signals move at the speed of light. In theory, an early warning system could send signals ahead of an approaching quake, pop up warnings on cell phones, and allow people to move to safer locations prior to the arrival of the quake waves.

To this end, the United States Geological Survey has created a network of seismic stations linked to a phone warning system called ShakeAlert. This is relatively new and is expected to expand into other states. One practical application of ShakeAlert is the connection to the Bay Area Rapid Transit (BART) system allowing automated braking of commuter trains upon detection of a significant quake, replacing a previous system that operated manually.

A few seconds of warning is a good thing. However, a critical examination reveals a few issues. In a very large quake, as the seismic waves spread along a fault, the size may develop rather than be instantaneously recognizable, meaning that early detections may underestimate the size; in other words, because huge quakes displace large areas along faults, the signal for “big quake” may evolve rather than announce itself immediately. By then areas close to the rupture will already be feeling the shaking.

The proximity of faults and people presents another problem with useful earthquake warning systems. California is most likely to have large quakes on the San Andreas, Hayward, or one of the numerous thrust faults branching off the San Andreas. These faults are often right in the midst of urban areas. There will be little time for instruments to detect the quake, for algorithms to process the information, for cell numbers in the affected region to be triggered, and then for people to actually pick up their phone and see the alert. For example, when a major quake happens on the Hayward fault, people in Berkeley and Oakland will experience the full force in mere seconds. In Los Angeles, when the fault that nearly cuts the famous Capitol Records Tower activates, people examining celebrity handprints in the concrete of the nearby Grauman’s Chinese Theatre will feel the quake long before a warning reaches their phones.

Drop & Cover or Triangle of Life?

If you could receive a quake alert on your phone, what would you be able to do?

One long-standing idea about how to survive a quake is to “drop and cover,” meaning to get underneath something such as a table and to cover one’s head. This is what the USGS and the Red Cross recommend, however this protection has limitations. It is possible that if a building is collapsing, and heavy beams are crashing down, these might flatten a table and squish anyone hiding underneath. On the other hand, if one was not under a table that could happen as well. As Conan O'Brien once humorously pointed out when interviewing the USGS's Lucy Jones, even an IKEA table offers more protection than nothing.

A recent idea that needs to be addressed here is the so-called “triangle of life,” the concept that rather than ducking underneath a table, one might have a better chance being on the side of a table. The idea is that a falling beam hitting a table could create a triangular-shaped opening in which one could survive. There is, however, no real scientific support for this concept, which the USGS describes as a “misguided idea” that “doesn’t apply to buildings constructed within the United States." Your best bet is still to drop and cover, to get underneath something sturdy as a shield against falling objects.

The temptation to exit the building one is in might be strong–after all, collapsing buildings are what kill people. Leaving the building seems to make sense, but there is great danger of objects from the sides of the building raining down upon those leaving. For example, if one happens to be near the exit of a skyscraper when a quake happens, and one runs outside to flee the building, one could be heading straight into a curtain of falling glass panes that have detached from the upper levels of the building, shattering on the sidewalk like so many transparent guillotines. Deadly glass panes plummeting from high rises already occur with alarming frequency during large wind storms, and in an earthquake we can expect streets to be hit with heavy glass falling with deadly force.

So staying inside is dangerous, going outside is dangerous–what choice does one have? Even this question rests on a flawed assumption. In reality, when a major earthquake happens, you probably will not be able to move at all. One of the authors (Newton) personally experienced this phenomenon during the 1989 Loma Prieta quake, in which the shaking was so severe that he found himself unable to stay upright when attempting to walk across a living room, and then upon falling to the ground, discovered that the debris of broken glass and brick fragments meant that even crawling was not an option. You may make a plan about what you should do, but in practice you may not be able to move at all and you will ride out the quake in whatever place you happen to find yourself.

Preparation: the Best and Only Option

While there is no way to predict when specific quakes will happen, you can take a number of actions today that can make your life easier afterwards.

Many people may start at a place such as the Red Cross’s earthquake preparation page. However, a critical examination shows that much of this information is problematic.

The Red Cross earthquake preparation list advises that “What you should do before an earthquake” includes “Get a fire extinguisher and learn how to use it safely.” This is about as useful as the demonstration of how a seatbelt works shown to every airline passenger before every flight. The Red Cross notes the importance of “eat[ing] healthy food and get[ting] enough sleep.” Cheers, thanks for that. The Red Cross also advises, “It’s normal to have a lot of bad feelings, stress, or anxiety”; it’s good to know that the default psychological condition for most people won’t change following a quake. The Red Cross then warns, “Be ready to live without power, gas, or water,” without providing any specifics that would actually help.

With the paucity of good preparation information, people may consider purchasing an “earthquake preparation kit.” Such kits are sold by a number of companies and typically contain similar items, such as emergency food rations, a water bag for collecting water, bandaids, flashlights, a poncho, a survival blanket, and a whistle.

These items are all quite useless. In an actual post-quake environment, emergency food is a low priority; you probably have plenty of non-perishable items around your house, and if not, then humans can go days without food without any ill effect. Water will be the major requirement, but an empty bag for containing water does very little if broken water lines mean that no water comes out from your household tap (and, of course, stores will sell out of bottled water immediately following a quake). You likely already have household bandaids; in any event, they will do nothing for the serious medical issues following a major shake. These kits often emphasize flashlights and glowsticks, but people forced into urban camping may find that bedtimes are far earlier than our current long, light-sustained nights. Ponchos are unnecessary if one has access to one’s normal clothes–which, if you’re able to find your home’s emergency kit, that means you probably still have access to your house. Likewise, survival blankets are unnecessary; even in winter, the majority of California does not experience life-threatening low temperatures. And most useless of all is a whistle. The idea of the whistle’s inclusion in the emergency kit involves you being trapped under a collapsed building; if you can signal by blowing a whistle, then this may guide help to your position (the standard is to make a noise once every minute, so rescuers can work and then stop to listen on a schedule). However, there’s a fatal logic error here. If you have access to the emergency kit with the whistle, then you are not trapped; if you are trapped, you won’t have access to the emergency kit. Outside of the absurdly unlikely circumstance that your home collapses on you in such a way that you are within arm’s reach of this kit, then the whistle will be just as useless as the rest of the items found in such earthquake emergency kits.

Here’s what you really need instead, in order of importance:

• Water. Water will be the first thing you need following a devastating quake, but our fragile infrastructure guarantees that many will face household taps that gurgle air rather than provide drinking water. Store bottled water at home, store water in your car, store water at work; you never know where you’re going to be when the quake happens. You really cannot store enough water in preparation for what is about to happen. As a related item, consider a camping water filter or a LifeStraw; these can remove bacteria (and depending on the grade, some viruses) from natural waters, allowing you greater potential water sources. Do not imagine that the government is going to send water trucks into cities to slake the thirst of those affected by a big quake; the freeways and roads will be disrupted and there won’t necessarily be a way for such trucks to reach neighborhoods, nor for you to travel to where the water is located. This is going to be the biggest health concern; people can last about three days without water.
• Garbage bags. Lots of garbage bags–think Costco-level. These will not be for throwing out trash, but rather dealing with an issue related to the collapse of the water and sewage system. Within a few hours of a major quake, everyone is going to need to use the toilet. However, not everyone has a sufficiently large or private backyard. A simple solution is this: line the inside of your now non-flushable toilet with a garbage bag. Then seal it up and dispose of it (perhaps remembering how your neighbors let their dog onto your lawn but failed to pick up the waste).
• Baby wipes. Every geologist learns, often during the undergraduate capstone course known as Field Camp, what it is like to go many weeks without showering. Geologists also learn that with a sufficient quantity of baby wipes, one can mimic a shower in such a way as to keep oneself somewhat clean. A baby wipe bath is not only hygienic, but good for one’s mental outlook. Dry shampoo is also a pro-tip.
• Generator. In many parts of the world unconnected to a power grid, and in seemingly every overcrowded public campground these days, generators are the default way of making electricity. Post-quake, generators may be a good way to make electricity until the reconstruction of our power grid. Many generators include a dual-use option that allows you to use fuel from barbeque propane, which is safer than having loose containers of gasoline around your house. Consider that some areas, post-quake, will be cut off from the electrical system for a very long time. While we are accustomed to thinking that when the power goes off in a storm, it will be restored in a few hours, imagine that instead it’s not just the power lines that have fallen, but the poles themselves that have been uprooted; moreover, the roads to get to the poles may be disrupted and impassable to repair vehicles. Before the electrical lines are reconnected to your house, technicians will have to replace the poles; before they can do that, engineers will have to stabilize the roads and construction workers will have to repave the streets. All this is going to take a very long time. While a generator solution may not be for everyone in every situation, it is worth considering before the quake.
• Cash. A group of esteemed philosophers once posited, “Cash rules everything around me, dollar dollar bill y’all.” We should take this admonition to heart. Withdraw and store some money, with the idea that following a quake, when all electricity is out and ATM machines and Venmo and credit cards no longer work, it would be useful to be able to buy some needed items.
• USB power batteries. There are a number of relatively inexpensive USB batteries which can power electronics, such as lights and cell phones. One good thing is that the cell system may have the robustness to survive a quake, unlike the situation with land lines that are very vulnerable. Ad-hoc cell towers are sometimes added at places such as concert venues to increase signal, and it is plausible that such cell towers could restore cell service much quicker than other utilities.
• Natural gas shutoff wrench. In the United States, there is no requirement that residential homes have an automatic natural gas shutoff valve, one which when triggered by substantial acceleration turns off gas coming into homes. If the quake breaks the fragile natural gas lines throughout your house, and you start to smell gas, you have a very limited time in which to shut off the gas main. Know where your gas main is, and attach to the main with a string a specialized wrench to allow you to immediately turn off the gas if you smell a leak. Of course, in a big quake, there could be gas main breaks throughout city streets; this happened in the Marina District in San Francisco following the 1989 Loma Prieta quake and sparked the major fires there. You can’t do anything about the possibility of city gas mains, but you can try to protect your home with a shutoff wrench.
• Home modifications. Everyone can take a few simple measures to improve the safety within a home, including removing glass-framed pictures from areas where you might walk in the night after a quake. Bookshelves can be easily secured to walls. Heavy objects can be placed lower, to reduce the risk of injury if they fall. If you own a home, there are a number of options for long-term strengthening your property, such as the installation of cripple walls and bolting the house to the foundation. In a similar vein, buying earthquake insurance guarantees that you will not receive the unpleasant surprise awaiting virtually every other homeowner--that your home insurance won't cover earthquake damages, and the federal government under FEMA is unlikely to offer much money to rebuild.

You sometimes hear that people should plan to be independent for 72 hours following a major quake. This is optimistic. For some people, the damage following a big quake won’t be measured in hours but in weeks and months and years--that’s the timeframe to consider for earthquake preparation.

Do not imagine that FEMA or the California Office of Emergency Services or local governments have stockpiled warehouses full of necessary items and stand ready to spring into action. Assume a situation closer to the bungled response to the natural disaster of the 2005 Hurricane Katrina in New Orleans, where tens of thousands of Americans were left ignored and abandoned by their state and federal governments. If history is any guide, you should plan to be on your own. Following a giant quake, there will be no cavalry coming to your rescue–there are only the preparations you make for yourself and your family right now.

References

1. Wald, L., & Scharer, K. (n.d.). Introduction to Paleoseismology. https://www.usgs.gov/programs/earthquake-hazards/introduction-paleoseismology
2. Lienkaemper, J., Williams, P.L., Guilderson, T.P., 2010. “Evidence for a Twelfth Large Earthquake on the Southern Hayward Fault in the Past 1900 Years.” Bulletin of the Seismological Society of America, Vol. 100, No. 5A, pp. 2024–2034, October 2010, doi: 10.1785/0120090129
3. Ludwin, R. (n.d.). Earthquake Prediction. https://pnsn.org/outreach/faq/earthquake-prediction
4. Earthquake Hazards Program (2022, March 9). Earthquake Early Warning Around the World. Usgs.gov. Retrieved June 20, 2024, from https://www.usgs.gov/programs/earthquake-hazards/science/earthquake-early-warning-around-world
5. USGS. (n.d.). What is the "Triangle of Life" and is it legitimate? Usgs.gov. https://www.usgs.gov/faqs/what-triangle-life-and-it-legitimate