The Network Felt Another Earthquake

On June 24, multiple onocoy stations in Japan picked up motion signatures from a significant earthquake event. Different stations, different arrival times, matching ground-motion patterns: exactly the kind of signal you would expect when a tectonic wave moves through a dense reference network.

This is not the first time we have seen it.

In December 2025, onocoy stations also picked up vibrations from earthquakes near Izmir and Bangladesh. At the time, we described it as an unexpected use case: not what the network was built for, but a useful reminder that high-precision GNSS infrastructure can observe more than positioning alone.

The same is true here. onocoy is not becoming an earthquake-monitoring company. But when thousands of reference stations continuously measure position at high precision, ground motion sometimes shows up in the data.

What the Stations See

A GNSS reference station is built to provide stable correction data. For that to work, the antenna position has to be known and stable. If the antenna moves, even slightly, the data reflects that movement.

Most of the time, movement is not dramatic. It can be a bad installation, a flexible pole, roof vibration, thermal effects, or an antenna mounted somewhere it should not be. From the network's perspective, those signals are usually quality issues.

But sometimes the ground itself moves.

During an earthquake, seismic waves travel through the earth and cause small but measurable displacement. A high-quality GNSS station can record that displacement as a change in position or velocity. When several nearby stations show related motion signatures around the same event window, the pattern becomes much more interesting.

That is what we saw in Japan on June 24.

Why This Is Interesting Nonetheless

onocoy has not suddenly become an early-warning system.

Dedicated earthquake early warning systems use purpose-built seismometers, accelerometers, communication infrastructure, and alerting systems. Countries such as Japan operate highly advanced monitoring networks for exactly this purpose. GNSS reference stations are not a replacement for that infrastructure.

The interesting point is different: a decentralized GNSS correction network can produce useful secondary signals. It was built for centimeter-level positioning, but its continuous measurements can also help identify motion events, installation problems, and unusual behavior across the network.

That makes earthquake observations a useful stress test for the data.

If the same event appears across several stations with plausible timing and motion patterns, it tells us something about how sensitive the network is to real ground movement. If a single station shows motion while neighboring stations do not, that may tell us something else: the antenna, mast, roof, or receiver setup might be unstable.

Both outcomes are useful.

The Practical Value Today: Better Quality Assurance

The immediate value is not public earthquake alerts. It is station quality assurance.

Every correction network depends on station integrity. A reference station that slowly shifts, vibrates, or moves under load can degrade the quality of the corrections it produces. The faster we can detect that, the faster we can label the station correctly, notify the operator, and protect the quality of the data clients consume.

Motion detection helps separate three very different cases:

→ Real ground movement: multiple stations show a consistent event pattern.
→ Local installation problems: one station moves in a way nearby stations do not.
→ Normal noise: small fluctuations that do not indicate either an event or a station issue.

That distinction matters. A station should not be penalized for recording a real geophysical event. But a station mounted on a moving structure should be flagged, because its correction data may not be reliable enough for production use.

Better motion labeling means better network quality.

Why Density Matters

A single station can show motion. A dense network can show a pattern.

That is why coverage matters beyond the usual correction use case. As more stations join the network, onocoy gains more independent observation points. The value is not just more coverage for RTK users. It is also a better ability to compare stations against each other.

If an event appears across a region, the network can treat it differently than a one-off anomaly. If only one antenna moves, the system can investigate that station more closely.

This is where decentralized infrastructure becomes interesting. Thousands of independently operated stations can act as a distributed sensing layer, even when that was not the original reason they were deployed.

Again, this does not turn the network into a dedicated seismic system. But it does create a useful additional signal on top of the positioning data.

What Comes Next

The next step is better labeling and better tooling.

For onocoy, earthquake observations are less about publishing event screenshots and more about improving how the network understands motion. The goal is to distinguish real ground movement from station instability, and to do it automatically over time.

That work supports the core business: reliable GNSS corrections for agriculture, drones, surveying, machine control, logistics, and other applications that depend on accurate positioning.

If motion detection helps us identify weak installations faster, clients get better data. If it helps us avoid false assumptions during real geophysical events, miners get fairer treatment. If it occasionally reveals an earthquake moving through the network, that is a powerful demonstration of what high-density GNSS infrastructure can observe.

Live network: console.onocoy.com/explorer