ShakeAlert

ShakeAlert is an experimental earthquake early warning system (EEW) for the West Coast of the United States and the Pacific Northwest sponsored by the United States Geological Survey (USGS).

The system uses sensors to detect P waves, compressional primary waves created by earthquakes and traveling faster than the damaging S waves. ShakeAlert is currently in a demonstration phase and has been sending real-time alerts to selected beta users since January 2012. The system will issue automated alerts to give people time to take protective actions like "drop, cover and hold on" in the event of a quake, preventing injuries caused by falling debris, automatically stopping public transport systems, preventing cars from entering bridges or tunnels, automatically shutting down industrial systems and gas lines, and triggering specific protocols in hospitals and other sensitive work environments. Initially the system will cover the west coast of in North America which is exposed to significant seismic risk along the San Andreas fault zone or the Cascadia subduction zone.

The second algorithm crucial to ShakeAlert is the tc-Pd Onsite algorithm. By using amplitude pd and period tc of the first signs of shaking, this onsite algorithm more accurately predicts the intensity and size of the earthquake than the ElarmS do. The tradeoff for using these algorithms for the earliest detection possible means having a less reliable approach than regional warning algorithms, however some argue that the added seconds to prepare are more important than reliability. Lastly, the Virtual Seismologist, as known as the VS method, imitates the analysis of a human scientist in terms of capacity, but does so at a faster rate. A Bayesian framework is used with inputs of acceleration, velocity, and displacement. The last step requires of all these algorithms to come together in a decision module. This decision module broadcasts the probability, size, and other characteristics of the earthquake.

The ShakeAlert system alerts the surrounding populations of an earthquake that has already happened. When the system detects the earthquake, the first step is to release several horn noises at a high volume level. Next, an automated voice yells out the words earthquake. The shaking that follows can start after varying intervals of time- anywhere from a couple seconds to several minutes. The purpose of then ShakeAlert system is not to predict an earthquake before it happens, but instead to determine and locate the earthquake as soon as it begins to happen. Within a single second, ShakeAlert can detect the location and severity of the earthquake to warn people of its presence. The people that these type of signals reach can vary. In the state of California, where earthquakes happen quite often, some of those most important listeners of this system include the Bay area’s rapid transit system (BART). Other important organizations that need to know the early onset of earthquakes are Fire Departments all along the west coast, as well as places like Disneyland, where a large population of people is limited to a relatively small geographical space. BART is already directly connected to the ShakeAlert system, so if the system were to send out early warnings of an earthquake, the metro system will automatically pause and adjust its trains to prevent serious accidents or derailments. Doug Given, one of the main leaders of the project, also serves as the National Earthquake Early Warning Coordinator for the U.S. Geological Survey. He explains in an interview with National Public Radio, the way in which the system is designed to alert the public. He explains that the system plans to reach populations in a similar way that the public receives Amber alerts.

...
Wikipedia