Earthquake alerts do not predict the next big quake. They detect an earthquake after it has begun and attempt to warn people before stronger shaking reaches their location. That distinction matters after a jolt, when people often want to know whether the event was the main earthquake and whether another is coming.
Scientists cannot reliably say exactly when, where or how large a future earthquake will be. Earthquake early warning can still help, however. It is a race between seismic waves, sensors, data processing, communications networks and human reaction time. In favorable circumstances, an alert can provide seconds to take protective action. In other cases, shaking may begin before an alert arrives, or no public alert may be issued.
Those limits are a consequence of physics and system design, not necessarily a failure of the warning system.
Earthquake prediction and earthquake early warning are different
Earthquake prediction would mean identifying a specific future earthquake before it starts, including its time, location and likely magnitude. Scientists can identify active faults, assess long-term seismic hazard and estimate earthquake probabilities over years or decades. But there is no reliable scientific method for predicting an individual major earthquake on a particular day and in a particular place.
Earthquake early warning begins only after a fault has started to rupture. Seismic sensors detect the first energy released, automated systems rapidly estimate the event, and alerts may be sent to areas expected to receive damaging shaking.
An alert that says shaking is expected is therefore not a forecast. The earthquake is already underway, even if the person receiving the alert has not yet felt it.
How earthquake warnings work
Earthquakes generate several types of seismic waves. P-waves, or primary waves, generally arrive first and travel faster than other major waves. They often produce less intense shaking than the later-arriving S-waves, or secondary waves. Surface waves can also contribute substantially to shaking and damage, especially during large earthquakes.
The difference in arrival times creates the opportunity for warning. A network of seismic sensors detects initial ground motion. Software compares signals from multiple stations, estimates an earthquake’s location and magnitude, and calculates where potentially damaging shaking may occur. If alert thresholds are met, the system can distribute a warning through available public and institutional channels.
All of this must happen within seconds. The system must distinguish an earthquake from local noise, confirm that multiple sensors are detecting the same event, estimate likely impacts and send a message that can be acted on immediately.
In California, Oregon and Washington, the U.S. Geological Survey’s ShakeAlert system provides earthquake early warning information. Public notifications may be delivered through channels such as Wireless Emergency Alerts, supported smartphone alert services and participating state or local emergency-management partners. Availability depends on location, device settings, connectivity and local alerting policies.
Why warning time varies
There is no standard amount of warning before an earthquake. An alert may arrive seconds before stronger shaking, after shaking has started, or not at all. People close to the earthquake source may have little or no useful warning because damaging waves can arrive before detection, analysis and alert delivery are complete.
Several factors affect warning time and accuracy:
- Distance from the rupture: People farther from the source may have more time between initial detection and stronger shaking.
- Sensor coverage: A dense sensor network can generally detect and characterize an event more quickly than a sparse one.
- Earthquake depth and rupture geometry: Deep and shallow earthquakes, as well as ruptures that spread along a fault, can produce different shaking patterns and warning challenges.
- Magnitude estimation: The first signals reveal only part of an event. A rupture can continue growing, requiring updated estimates.
- Communication delays: Data transmission, alert processing and delivery to a device all take time.
- Local conditions: Soil, basin structure, building characteristics and a person’s location can affect how shaking is experienced.
A warning-time estimate should be treated as an estimate, not a guarantee. Even a short alert can be valuable if it prompts immediate protective action.
Why an alert may arrive after shaking starts
Receiving an alert after feeling an earthquake is a common source of confusion. Often, the person was simply too close to the source for the system to get ahead of the waves. In other cases, a person may first feel relatively light initial motion and receive an alert before stronger shaking arrives.
The system may also need additional sensor data before it can determine that an event meets the threshold for a public alert. Delays in communications or device delivery can further reduce warning time. A phone may also have limited connectivity or settings that prevent a particular alert service from reaching it.
Not every detected earthquake generates a public warning. Warning systems are designed to alert areas where they expect shaking to exceed defined thresholds. Small, remote events, or events expected to produce only weak shaking in populated areas, may be recorded without triggering a notification.
This is the key difference between earthquake detection and public warning: scientists may detect many more earthquakes than the public is alerted about.
What USGS earthquake alerts can and cannot do
Early-warning information can support automated protective actions where organizations have built systems to use it. Depending on local arrangements, this may include slowing trains, opening fire-station doors or pausing sensitive industrial processes.
For the public, the most useful response is simpler: act immediately. An earthquake alert is not a prompt to search for more information before taking cover. By the time someone reads the details, strong shaking may be close.
Initial estimates of location, magnitude and expected shaking can change as more data become available. That is normal. Early warning systems must make decisions while an earthquake is still unfolding.
Why changing estimates and missed expectations happen
Early warning systems operate with incomplete information. Waiting for more certainty could reduce unnecessary alerts, but it would also use up the seconds that make a warning useful. Alerting too readily can cause disruption; alerting too cautiously can leave people without a warning they might have used.
An alert may be revised, may arrive for shaking that feels milder than expected, or may not arrive for an earthquake that a person found alarming. Individual experience is not a complete measure of the system’s estimate. Local ground conditions, nearby structures and upper-floor locations can make shaking feel stronger in one place than another.
The purpose of an earthquake alert is not to tell every person exactly what they will feel. It is to provide a brief opportunity to reduce harm where the physics allows it.
What to do when an earthquake warning arrives
The standard immediate response is Drop, Cover and Hold On. Drop to your hands and knees, take cover under a sturdy desk or table if one is nearby, and hold on until the shaking stops. If no shelter is nearby, get low and protect your head and neck with your arms.
- Stay indoors if you are already indoors. Do not run outside during shaking.
- Avoid windows, glass, tall furniture and objects that could fall.
- Do not use elevators.
- If you are in bed, remain there and protect your head and neck with a pillow unless there is an immediate hazard above you.
- If outdoors, move to an open area away from buildings, utility lines and other falling hazards when possible.
- If driving, pull over safely when practical and avoid stopping beneath bridges, overpasses, power lines or other hazards.
After shaking stops, check for injuries and immediate hazards such as fire or gas leaks. Follow directions from local officials. The USGS can provide earthquake information, while local emergency authorities are generally the primary source for evacuation instructions, road conditions and community-specific guidance.
Aftershocks are possible, but not precisely predictable
After a noticeable earthquake, concern about aftershocks is reasonable. Larger earthquakes are commonly followed by smaller events as the affected fault system adjusts. Scientists can estimate aftershock likelihood statistically based on patterns from previous earthquakes, and agencies may communicate elevated short-term risk after a significant event.
They cannot identify the exact time, location or size of the next aftershock. Reports that aftershocks are possible or likely should encourage practical preparedness, not the belief that another earthquake can be scheduled or predicted.
Keep shoes, a flashlight, medication and essential supplies accessible. Know how to receive local emergency information. If a building appears damaged, follow official guidance and do not re-enter if authorities advise against it.
Earthquake technology is not a crystal ball
Earthquake early warning is not an attempt to see into the future. It is infrastructure designed to act during the narrow interval between the first detected seismic waves and potentially stronger shaking.
People who expect alerts to predict the next earthquake may see their limits as failure. People who understand that an alert detects a rupture already in progress can use even a short warning for its intended purpose: protecting themselves immediately.
Earthquake alerts will always be shaped by geography, sensor coverage, communications and uncertainty. Their value lies not in certainty, but in turning a few hard-won seconds into safer decisions.