1. The observation is an arrival record, not a photograph of the interior

A seismometer records ground motion over time. The earlier P wave is compressional; the S wave is a shear wave. Comparing arrival times and wave types across stations helps infer the paths and structures the waves encountered. A weak signal missing at one station alone does not establish that the material beneath it is liquid.

An IRIS educational example combines worldwide records of the 1994 Northridge earthquake with wave paths. Waves from the same earthquake follow different routes to different stations, and different wave types have different speeds. Read a record by first asking which wave arrived, at which station distance, and at what time.

2. A shadow zone does not mean every seismic wave disappears

Liquids cannot transmit seismic shear motion. The absence of S waves traveling through the outer core is central evidence for its liquid state. P waves, however, can cross it: the core does not block every seismic wave.

In the introductory IRIS diagram, the first P arrival disappears beyond roughly 100° from the earthquake, while PKP waves that cross the outer core reappear beyond roughly 140°. These are angular distances between the earthquake and station measured at Earth's center. The approximate 100–140° range belongs to this educational account; it is not an identical detection boundary for every earthquake.

First identify the missing arrival, then compare records at other distances and a model of the paths. A shadow zone concerns arrivals with particular paths, not an absence of every wave type or all ground motion at every time.

3. Arrival separation is not automatically usable warning time

For a hypothetical calculation, let both waves follow the same straight 60 km path at constant speeds. Choose 6 km/s for P and 3.5 km/s for S. Using t=distance/speed gives arrivals at 10 s and about 17.1 s, a separation of about 7.1 s. These speeds and the path are assumptions for the example, not universal Earth constants.

Even if S arrives at 17.1 s in this example, an alert recipient does not necessarily have the full 7.1 s. Network detection, location and magnitude estimation, and delivery take time. Near the origin, there may be no time to receive an alert before shaking.

USGS describes ShakeAlert as detection of an earthquake that has already begun, not prediction of a future earthquake date. To assess actual warning, distinguish calculated wave-arrival separation, the time a system issues its message, and the time the user receives it.

4. Sources

IRIS: Exploring the Earth Using Seismology explains records and the outer-core inference. USGS earthquake early-warning overview explains detection, estimation, and delivery. The 60 km calculation is not a measurement of warning performance during an actual earthquake.