Every earthquake generates a torrent of data — magnitude, depth, waveform, energy release, aftershock probability. Seismic dashboards compress this into visual layers you can read in seconds — if you know what to look for. This guide teaches you how.
📊 OPEN LIVE DASHBOARDThe USGS processes data from over 2,000 seismometers worldwide and catalogues approximately 20,000 earthquakes every single day. The vast majority are too small to feel. A handful are significant. An earthquake dashboard takes this firehose of data and compresses it into a visual interface designed to answer one question immediately: is something unusual happening, and where?
But dashboards are only useful if you can read them. A scatter of coloured dots on a dark globe looks dramatic, but without understanding the encoding — what colour means, what size means, what the axes of the sidebar charts represent — you are looking at decoration, not data. This guide breaks down every layer of a modern seismic dashboard so you can interpret what you see in real time.
A well-designed earthquake dashboard encodes at least five independent variables simultaneously: location (latitude/longitude), depth (kilometres below the surface), magnitude (energy released), time (when the event occurred), and uncertainty (how reliable the measurement is). The most advanced dashboards — including the Pandita Data live map — add real-time feed status, tectonic plate overlays, and historical context layers.
Depth is the single most important factor in determining how destructive an earthquake feels at the surface. A shallow M5.5 at 10 km depth can cause more damage than a deep M6.5 at 500 km. Every serious dashboard colour-codes depth, and reading this encoding correctly is the most critical skill for interpreting seismic maps.
| DEPTH RANGE | COLOUR | CLASSIFICATION | SURFACE IMPACT | COMMON CONTEXT |
|---|---|---|---|---|
| 0–20 km | ■ RED | Very shallow | Maximum — localised intense shaking | Crustal faults, volcanic, induced |
| 20–70 km | ■ ORANGE / YELLOW | Shallow | High — felt over wide area | Most damaging tectonic earthquakes |
| 70–150 km | ■ GREEN | Intermediate | Moderate — attenuated with depth | Subducting slab earthquakes |
| 150–300 km | ■ CYAN / TEAL | Deep intermediate | Low — rarely felt without M6+ | Deep slab, Wadati-Benioff zone |
| 300–700 km | ■ BLUE / PURPLE | Deep focus | Minimal — almost never damaging | Deep subduction (Tonga, Indonesia) |
The 2023 Morocco earthquake (M6.8, 26 km depth) killed nearly 3,000 people. The same year, a M7.1 struck Tonga at 210 km depth — and barely generated a felt report. The Morocco event released less than one-tenth the energy but occurred in the shallow crust directly beneath populated areas. When reading a dashboard, colour always comes before size.
The magnitude number displayed on a dashboard is a shorthand for a measurement that has gone through multiple processing stages. Understanding how it is derived, and what its limitations are, will prevent you from misreading what the dashboard is telling you.
When an earthquake first appears on a dashboard, the reported magnitude is an automated estimate based on the first few seismometers to record the event. For large earthquakes, this initial magnitude can be off by 0.3–0.5 units. The 2011 Tōhoku earthquake was initially reported as M7.9, then revised to M8.9, then to the final M9.1. If you see a magnitude labelled "preliminary" or flagged with a timestamp, treat the number as approximate. Refresh 15–30 minutes later for a more reliable figure.
One of the most powerful tools in seismic analytics is a chart you will sometimes see in the sidebar of advanced dashboards: the frequency-magnitude plot, also known as the Gutenberg-Richter (GR) relation. It states that for every one-step increase in magnitude, the number of earthquakes drops by approximately a factor of ten.
The b-value is the slope of this line. When b ≈ 1.0, the seismicity follows normal tectonic behaviour. When the b-value drops significantly below 1.0, it can indicate that stress is accumulating on a fault — fewer small earthquakes relative to larger ones suggests the system is locked and building toward a larger rupture. Conversely, elevated b-values (above 1.0) are common in volcanic regions and areas with high geothermal activity, where many small events dominate.
Temporal changes in b-value have been observed before some major earthquakes. A progressive decrease in b-value in the region surrounding a locked fault may indicate increasing stress concentration. This is not a prediction — it is a statistical shift in the character of seismicity. Advanced dashboards that display b-value evolution over time windows (e.g. rolling 90-day b-value) provide one of the few empirically grounded forecasting signals available in seismology.
Some dashboards — particularly those linked to specific seismometer stations — display raw or processed seismogram traces. These wiggly lines encode the actual ground motion recorded by a sensor, and learning to glance-read them is one of the most useful analytical skills you can develop.
The time gap between the P-wave and S-wave arrivals (the S-P interval) is the key diagnostic. An S-P interval of about 8 seconds corresponds to an earthquake roughly 64 km away. At 30 seconds, the source is approximately 240 km distant. Seismologists use three or more stations to triangulate the epicentre — and this triangulation is exactly what automated dashboards perform hundreds of times per day.
The USGS publishes multiple GeoJSON feeds at earthquake.usgs.gov, each filtered to a different magnitude threshold and time window. When a dashboard says "all earthquakes in the past 24 hours," it is pulling from a specific feed endpoint. Knowing which feed you are looking at determines the completeness and latency of what you see.
| FEED | MAGNITUDE THRESHOLD | TIME WINDOW | TYPICAL EVENT COUNT | USE CASE |
|---|---|---|---|---|
| all_hour | All magnitudes | Past 1 hour | 50–200 | Real-time monitoring, swarm tracking |
| all_day | All magnitudes | Past 24 hours | 500–2,000 | Daily overview, dashboard default |
| 2.5_day | M2.5+ | Past 24 hours | 50–120 | Felt earthquakes, alert systems |
| 4.5_day | M4.5+ | Past 24 hours | 5–20 | Significant events, global monitoring |
| significant_month | Significant only | Past 30 days | 5–15 | Major event review, reports |
| all_month | All magnitudes | Past 30 days | 10,000–30,000 | Research, statistical analysis |
Pandita Data's live earthquake map defaults to the 2.5_day feed — showing felt earthquakes worldwide in the past 24 hours. This is the optimal balance between information density and clarity. Switching to the all_day feed shows thousands of micro-earthquakes that, while scientifically valuable, create visual noise for general users.
The difference between casually glancing at a dashboard and actually reading it is pattern recognition. With practice, certain visual signatures become immediately meaningful. Here are the five patterns that seismologists train themselves to spot, and that you can learn to recognise on any dashboard.
Even experienced dashboard users make these errors. Awareness of them immediately improves the quality of your interpretation.
1. Confusing detection increase with activity increase. When a new seismometer is installed in a region, previously undetected micro-earthquakes suddenly appear. The dashboard looks like a swarm appeared overnight — but the earthquakes were always there. Check whether the network was recently expanded before concluding that seismicity has increased.
2. Ignoring depth when comparing magnitudes. "There was an M5.5 near us and nothing happened, so this M5.5 should be fine." This is only true if the depth is comparable. A shallow M5.5 at 5 km is a completely different experience from an M5.5 at 150 km. Always compare depth alongside magnitude.
3. Expecting aftershock sequences to be complete. Immediately after a large earthquake, the seismic noise is so high that many smaller aftershocks cannot be detected. The dashboard will show fewer events in the first hours than actually occurred. This is called catalogue incompleteness, and it temporarily biases the visible aftershock count downward.
4. Reading preliminary magnitudes as final. As noted above, early automated magnitudes can shift significantly. Social media amplification of preliminary numbers — before the seismological community has confirmed the magnitude — is a persistent source of misinformation.
If you want to go beyond consuming dashboards and start building your own analysis, here are the primary data sources that feed every serious seismic dashboard worldwide. All of them are free and publicly accessible.
| SOURCE | COVERAGE | FORMAT | LATENCY | BEST FOR |
|---|---|---|---|---|
| USGS Earthquake Hazards | Global (M2.5+ outside US) | GeoJSON, CSV, KML | ~2–5 minutes | General-purpose, dashboards, alerts |
| EMSC (Europe-Med) | Euro-Mediterranean | QuakeML, JSON API | ~5–15 minutes | European focus, felt reports, Aegean |
| IRIS DMC | Global (waveforms) | miniSEED, SAC, RESP | Near real-time | Seismogram analysis, research |
| ISC Bulletin | Global (reviewed) | ISF, CSV | Months (reviewed) | Research, historical catalogue |
| NOA Greece | Hellenic region | Bulletin, HTML | ~5–10 minutes | Aegean, Santorini, Hellenic Arc |
| GeoNet NZ | New Zealand | GeoJSON, QuakeML | ~2 minutes | Tonga-Kermadec, Alpine Fault |
Pandita Data aggregates primarily from the USGS GeoJSON feeds, supplemented with EMSC and NOA data for enhanced Aegean coverage. The live earthquake map you see on this page pulls from the 2.5_day endpoint and refreshes every 60 seconds. Every dot on the map is backed by a full GeoJSON feature containing 40+ properties — location, depth, magnitude, uncertainty, felt reports, tsunami flag, alert level, and more.
Beyond the basic map view, advanced seismic dashboards may display additional derived metrics. These are the numbers that operational seismologists monitor during crisis situations, and understanding them gives you a significant advantage over a casual viewer.
A seismic dashboard is not just a map with dots. It is a multi-layered analytical instrument encoding location, depth, magnitude, time, uncertainty, and statistical context simultaneously. To read it effectively, follow this sequence every time you open a dashboard:
1. Scan for colour first. Red/orange clusters demand attention — these are shallow events with the most surface hazard potential.
2. Check dot size second. Large dots are significant earthquakes. If a large dot is also red, it is the highest-priority event on the map.
3. Look for clustering. Is activity concentrated in one region? Does it align with a known fault or volcanic system?
4. Check recency. Are the events happening right now (bright, solid dots) or are they decaying remnants of an earlier sequence (faded)?
5. Context check. Open the event details. What is the exact depth? Is the magnitude preliminary or reviewed? Is there a PAGER alert? Has a tsunami warning been issued?
6. Compare to normal. Does this region always look like this, or is today different? The Gutenberg-Richter relation and historical event rates help calibrate expectations.
With these six steps — colour, size, clustering, recency, context, and comparison — you can extract more information from a 30-second dashboard glance than most people get from reading an entire news article about an earthquake. The live map below is your practice ground. Open it, zoom in, and start reading.