4 ms·
"The researchers found that, in the final two hours before a quake, the ground's motion often began to align with the expected motion. That alignment seemed to
by coin 3y ago
"The researchers found that, in the final two hours before a quake, the ground's motion often began to align with the expected motion. That alignment seemed to intensify as the time of the quake approached."
The GPS satellites aren't detecting anything. Researchers on the ground use positions from GPS to detect slight movements leading up to the earthquake.
- defrost 3y agoSounds like you'd prefer: Using GPS as a possible earthquake predictor (phys.org) https://phys.org/news/2023-07-gps-earthquake-predictor.html https://phys.org/news/2023-07-gps-earthquake-predictor.html The precursory phase of large earthquakes https://www.science.org/doi/10.1126/science.adg2565 https://www.science.org/doi/10.1126/science.adg2565 (14 hours ago) https://news.ycombinator.com/item?id=36814593 https://news.ycombinator.com/item?id=36814593
- ithkuil 3y agoYeah it is indeed frustrating when people talk like if GPS satellites "know where you are", yet that's a common misconception. It's not really that hard to explain how GPS works at a high level and it can be taught in schools. You first have to start with triangulation on 2d maps by measuring the direction (on a compass) you see 2 or more reference points (e.g. lighthouses). Then you hand-wavely describe that the same thing can be done in 3d and with moving "lighthouses in orbit" and that the trick has to do with knowing exactly where the satellite is right now (and you can compute that by knowing precise timing and each satellite orbit parameters) Then you emphasize the fact that the satellites don't know anything about you, not unlike lighthouses don't know anything about who's looking at them from far away
- namibj 3y agoJust use a ruler, set some offset to start at from your position, keep the offset the same for the >=dimensions+1 measurements, and show how you can recover both time (your ruler offset is your clock's offset from GPS time) and position just from the pseudoranges. The equations are fairly easy: (x - satellite[i].x)**2 + (y - satellite[i].y)**2 == (time_offset - satellite[i].pseudorange)**2 for your number of satellites, and it naturally extends to including a `z` for 3D, too.