EARTH PULSE / METHODOLOGY
How to interpret an earthquake record
An earthquake record can report magnitude, depth, location, shaking, and other products, but those values answer different scientific questions and carry different limits. Earth Pulse turns selected public earthquake records into short explanations of one well-supported feature of an event.
The stories are educational interpretations produced by an automated system, not official assessments, forecasts, hazard estimates, or safety guidance.
01 / PURPOSE
How Earth Pulse interprets an earthquake record
Earth Pulse helps a curious reader understand what a reported magnitude, depth, location, shaking product, or other earthquake record may mean without turning a limited observation into a broader claim. For a selected event, it looks for one useful public question that the available material can answer within clear boundaries.
A story explains selected evidence in the record; it does not independently investigate the earthquake. Earth Pulse is not a substitute for a seismologist, an emergency service, or the official source behind the data.
What can an Earth Pulse story explain?
Depending on the evidence available for a particular event, a story may answer one bounded question such as:
- How do depth, magnitude, and surface shaking fit together?
- How large might this earthquake have been underground?
- Was this earthquake part of an officially identified sequence?
- Have earthquakes this large been recorded nearby before?
- How can an earthquake begin hundreds of kilometres underground?
- What does the available evidence show about how the rock moved?
These are examples, not promises for every event. Each story answers one bounded question only when the available evidence can support it; otherwise no story is published.
02 / PROCESS
From observation to publication
The system separates finding suitable material from writing about it. Each stage narrows what the final explanation is allowed to say.
- Observation The system reads a named public source and preserves its reported values, product type, and update time.
- Context Only relevant scientific context or bounded comparisons are attached. Context can explain meaning, but it cannot prove an event-specific cause.
- Selection The available material must support one useful public question and an approved type of story. Selecting a topic is not itself a scientific finding.
- Validation Structured rules record what each source supports, what it cannot establish, and whether two products depend on the same underlying evidence.
- Explanation An AI language model drafts plain-language prose from that prepared material. Separate sentence-level checks reject claims that exceed it.
- Publication A story becomes public only when automated evidence checks pass and its source record is current. Material source changes can trigger re-evaluation or removal.
03 / EVIDENCE
What counts as earthquake evidence
A source-backed record is a value or statement that can be traced to a named source and version. This means SOLAST can show where a claim came from; it does not mean that every value is exact, final, or independent of other products.
- Published event record
- Reported event time, location, magnitude, depth, status, and other fields tied to an identified source record.
- Observation or analytical product
- Recorded ground motion, public felt reports, mapped shaking, or wave-based products, each kept with its own method and limits.
- Official interpretation
- An interpretation explicitly published by the source, such as a sequence relationship or regional event context. It remains attributed to that source.
- Scientific context or comparison
- Established background knowledge or a bounded calculation that helps explain scale and setting without becoming proof about the event's cause.
04 / DATA BOUNDARIES
What an earthquake record can — and cannot — show
Different parts of an earthquake record answer different questions. A clean number, point, or line on a map can look more certain — and more complete — than the underlying measurement or estimate really is.
Magnitude and intensity describe different things
Magnitude describes the reported overall size of an earthquake. Intensity describes the severity of shaking at a particular place, whether reported by people or represented by an appropriate shaking product. One magnitude value therefore cannot, by itself, establish how strongly the earthquake was felt, what damage occurred, or what danger existed at a specific location.
Depth, hypocenter, and epicenter are not interchangeable
Reported depth describes the estimated depth of the earthquake’s starting location, or hypocenter. The epicenter is the estimated point on Earth’s surface directly above the hypocenter. These describe the estimated start of the event, not the complete underground rupture; on their own they do not identify the exact fault or the cause.
A focal mechanism does not name the fault
A focal mechanism describes a style of rock movement compatible with recorded seismic waves. It can constrain the type of faulting, but it does not by itself establish which of the two possible nodal planes ruptured or identify the exact named fault.
| Data | Shows | Does not show on its own |
|---|---|---|
| Magnitude | The reported overall size of the earthquake. | Shaking intensity, damage, or danger at a particular place. |
| Reported depth / hypocenter | The estimated depth of the earthquake’s starting location (hypocenter). | The full rupture, exact fault, cause, or local effects. |
| Epicenter | The estimated surface point directly above the hypocenter. | The shape or complete area of the underground rupture. |
| Did You Feel It? | Where respondents reported felt effects and their reported intensity. | An instrument-only measurement, a complete population sample, or no shaking where reports are absent. |
| ShakeMap | The mapped distribution and severity of shaking using observations and estimates. | A direct measurement at every location or a damage forecast. |
| Focal mechanism | A style of rock movement compatible with recorded seismic waves. | Which of the two possible planes ruptured or the exact named fault. |
| Catalog neighborhood | Recorded earthquakes close in time and space within stated search limits. | A causal connection or a formal foreshock-aftershock relationship. |
| Plate boundaries | The event's broad regional tectonic setting. | The exact fault or cause of that earthquake. |
The explanation may connect a record with established science, but scientific context constrains what can be said; it does not independently confirm an event-specific cause.
05 / UNCERTAINTY
How uncertainty stays visible
Earth Pulse does not try to make an uncertain record sound definitive. The same limits attached to the prepared evidence must remain visible in the published explanation.
- Values and conclusions remain attributed to their source or product.
- Words such as reported, estimated, calculated, may, and approximately are preserved when they matter.
- Precision is not added beyond what the source supports.
- Products that share underlying inputs are not described as independent confirmation.
- Missing evidence and the important question that remains unresolved are stated rather than filled with a guess.
Automated evidence checks test whether every sentence remains within the prepared record and its limits. They do not independently verify the earthquake, constitute scientific peer review, or certify an explanation of its cause.
06 / SELECTION
Why some events have no story
Most earthquakes will not receive an Earth Pulse story. Publishing nothing is the correct result when the available material cannot support one useful explanation without guessing or overstating certainty.
- No approved public question is sufficiently supported by the available products.
- Required information is missing, conflicting, or has changed materially.
- The uncertainty cannot be communicated clearly in a short explanation.
- The draft does not pass the automated evidence checks.
Missing prose does not mean that an event is unimportant. The Story Atlas is a curated educational collection, not a complete seismic catalog, and must not be used to compare earthquake frequency, danger, or regional activity.
07 / SOURCES & LIMITS
Sources, revisions, and limits
Story families use named public sources appropriate to the question they answer. The labels below describe the evidentiary role behind each family, not a claim that every earthquake can support every question. A calculated product, a public report, and general scientific context are not treated as interchangeable forms of proof.
- Official event contextUSGS event pages, detail records, and official event products.
- Underground scaleUSGS event geometry and bounded source-specific scale comparisons.
- Earthquake sequenceExplicit relationships published in USGS event products; no sequence is inferred from timing alone.
- Historical neighborhoodBounded searches of the USGS Comprehensive Earthquake Catalog.
- Reported and calculated shakingDid You Feel It? reports and ShakeMap estimates, kept distinct.
- Depth and surface shakingUSGS event depth together with separately identified public-report or ShakeMap evidence.
- Shallow start and weak shakingUSGS event geometry and bounded surface-shaking observations with their uncertainty retained.
- Deep-focus earthquakeUSGS event geometry and source-backed regional context; the reported depth is not treated as exact.
- Triple-junction settingUSGS location data and published scientific plate-boundary datasets used only for broad setting.
- Ocean-ridge settingUSGS location data and published scientific plate-boundary datasets, without claiming an exact fault.
- Plate-boundary settingUSGS event records and the broad global boundary geometry described by Bird (2003).
- Local earthquake concentrationBounded USGS catalog results showing events close in time and place, without assigning a sequence type.
- Rock movementPublished USGS moment-tensor or focal-mechanism products, including their ambiguity and method limits.
Core event material is attributed to the USGS Earthquake Hazards Program, and each story links to its original USGS event record. An AI language model drafts the prose; SOLAST controls the evidence selection, boundaries, automated checks, and publication rules. The model itself is not treated as a source.
Earthquake records can be revised after publication. If a required source value changes materially, the related story can be re-evaluated or hidden rather than silently presented as current.
Unknown remains unknown. Earth Pulse does not infer an exact fault, complete rupture shape, reason the earthquake happened at that moment, future activity, local damage, or danger unless an appropriate source explicitly establishes the narrower claim. A short explanation also cannot represent every scientific interpretation of an event.
Earth Pulse is not a warning, forecast, damage estimate, hazard assessment, or safety service. For current official information and emergency guidance, use the relevant geological survey, warning center, and local authorities. See the full legal notice and disclaimer.