Speaker 0:  nature usually gives us a heads up, but we finally found a way to listen to the earth shutter before it blows. forget massive sensor arrays, a single instrument can now predict volcanic eruptions within 92% accuracy, giving us a life saving eight hour head start. this is called the jerk signal, and it predicts interrupt eruptions. sorry. not interruptions. researchers have spent ten years perfecting a system that detects jerk signals, a rapid rate of change in ground acceleration causing a magma fracturing rock. it's a specialized early warning system that catches the mechanical preparation of a volcano before the lava ever breaks the surface. this turns volcanic volcano monitoring from an expensive multi sensor luxury into an accessible single instrument reality for under monitored regions worldwide. what do you think about this, john?

Speaker 1:  i think this is cool. and just so you guys know, this has been validated. they validated this at, volcano, and it was tested over ten years at piton de la fommez, which is one of the most active volcanoes. and the result is ninety two percent of eruptions were successfully forecast. warning times range from minutes up to eight hours, and they only had about fourteen percent false positives where there was magma movement without eruption. and so this is a major improvement because volcano prediction, as we know, it's been highly uncertain. even detecting if interruption will happen has been difficult. and so this moves forecasting closer to operational reliability. and my question is, is this gonna work with, what's the largest volcano in the it's it's in the us, and it hasn't erupted in years.

Speaker 0:  well, and they're just really cold. reliable, what they call it.

Speaker 1:  yeah. something. i wonder if this will work with that. hopefully, they can find a way for that because if that one erupts, we're we're screwed. but, i think this is cool because, the way that it works is it uses essentially a single brand a single broadband seismometer. right? so traditional volcano monitoring requires networks of sensors, gas measurements, satellite deformation tracking. but the jerk method, it extracts far more information from existing seismic data. so it detects signals that were previously hidden in noise. and it's so the cool thing about this is it's less about new hardware and more about new signal processing and interpretation, which i think is cool. and i didn't read anything about using ai to predict, but i'm curious to i'm guessing if they to integrate ai into this, they might even get more accuracy. what are your thoughts?

Speaker 0:  well, there's just to go back, the biggest volcano in the us is mauna loa lau, located, on the big island in hawaii. it's the largest active volcano in the us. so that was the active one. so

Speaker 1:  sorry. the supervolcano in yellowstone is called the yellowstone caldera. that's the one i'm thinking of.

Speaker 0:  that's yeah. that's the next one on my list still. yep. or, we're screwed if those happen. i like to think of this update like a high-tech home security system. most sensors only go off, when the front door is kicked open when that's the eruption. but the jerk method is like a vibration sensor on the frame that hears the intruder's drill bits grinding into the lock. it catches the work being done to get inside long before the door actually swings wide open. i think this is really, great and could save a lot of lives. so i'm really excited, you know. and this is low cost compared to traditional rays, requires only one broadband, seismometer. it's currently active at the piton de la fournes expanding to mount aetna, this year.

Speaker 1:  yeah. oh, i remember mount aetna. i studied mount aetna in school. i totally forgot about mount aetna.

Speaker 0:  i'm not sorry. you'll your dollars in mount aetna is probably superior to mine at this point.

Speaker 1:  i just know it's in italy, and that's all i remember. but, anyways, it's a straddle volcano.

Speaker 0:  yeah. right now, it's primarily for government observations and disaster management agencies. you know, this is great immediate impact for those living near, active peaks. this means receiving evacuations, text hours earlier than previously possible. yeah. with 92% across 24 historical and live eruptions, with lead time two minutes to eight point five hours, events warning. this is amazing. i'm excited about this. you were gonna say, john? sorry.

Speaker 1:  yeah. i'm excited. so the big challenge here is that the signals are still extremely weak, and they're easily buried in environmental noise. so wind, ocean waves, human activity, things like that. so with this innovation, hopefully, they're gonna be able to identify more consistent patterns in these tiny fluctuations and distinguish real magma movement from random noise. and, hopefully, that the more data that they gather, the more accurate they're gonna be and if they introduce ai. so this is really cool. i'm excited for this.

Speaker 0:  yeah. the difference is standard seismology looks at velocity or acceleration. this system isolates the rate of change of acceleration, allowing it to filter out a lot of the noise and identify specific transient pulses created by magma intrusions fracturing the lithosphere. so this moves us away from the multiparameter era of the late twentieth century toward a smarter algorithmic monitoring where signal processing does the heavy lifting previously required by, dozens of gps and gas sensors. so the system has to account for things like earth tides, gravitational pull of the moon, and sun actually flexes the solid ground enough to mask these tiny jerk signals. so researchers use complex algorithms to subtract the moon's influence. we talked about the twenty twenty six mount expansion, project pos four dyke, for pos four d i d y k e is integrating this with fiber optic sensing, effectively turning miles of internet cables into one giant continuing, ear to the ground. and the system is so sensitive, it can detect aborted eruptions. on december 5, 2025, the system screamed incoming, but no lava appeared. skeptics called it a fail, but lava data proved magma did move and stall just beneath the surface, and the system literally heard the volcano clear its throat and decided not to cough.

Speaker 1:  crazy.

Speaker 0:  yeah. so a single sensors setup lacks spatial data. it can tell you an eruption is coming, but it struggles to pinpoint exactly where the volcano on the volcano. the flank might unzip compared to, full sensory, grid. so you know what's happening, but you don't quite know where it's gonna break through. so if you're living in a volcanic hazard zone, look for this kinda stuff and advocate for them to be putting this in your area. yeah. so we're moving from a reacting to eruptions to hearing them think, providing a critical window for human survival. if you lived at the base of volcano, would you trust your life to a single sensor, or do you think we still need old school multimillion dollar sensor grids to feel safe? let us know in the chat.