Read on Twitter
Let's talk about seismic networks and #fakequakes. A modern network comprises 100s of individual stations that transmit data to a central hub, where sophisticated automated software detects earthquakes within seconds and generates real-time reports. https://twitter.com/celestelabedz/status/1327268376440279041
Uncorrelated noise at individual stations isn't a problem. But network software can be fooled. In the old days, a burst of microwave noise from one repeater could be misinterpreted as a sudden flurry of tiny quakes near the stations using that repeater (fun times).
A #fakequake can happen other ways, including when network software gets fooled by signals from an earthquake outside the network. Network software has gotten both faster and better in recent decades, but hiccups do still happen.
Human error can be involved as well, e.g. when an old earthquake location is updated and an alert goes out. Because, yeah, #shifthappens. But if you're old enough to remember ancient times, i.e., the 1980s & 1990s, you gotta be impressed by just how good modern networks are.
PS. I gather the software glitch this time may have been on the NV side, not the SCSN side. Another way software can go wrong is to generate a "split", i.e. 1 event turns into 2. I think this results from a confusion between P & S waves. @Kate6HTN?
