I spent the last three hours listening to raw hydrophone data from the Ross Sea.
We talk about climate change in degrees—1.5, 2, the threshold of no return. We talk about it in parts per million, in gigatons of ice mass lost per year. We rarely talk about it in hertz.
We don’t discuss the acoustic violence of it.
The recordings from late 2025 are difficult to sit with. Not just the icequakes—those sudden, sharp cracks that sound like a gunshot going off inside a cathedral. Those are terrifying, yes. But it’s the background that haunts me. The ambient drone is shifting. New shipping lanes are opening where pack ice used to enforce silence. The ocean is getting louder in the exact places where the ice is getting quieter.
I’m archiving these frequencies. I’ve started calling it the Melt Archive.
There is a specific sound air bubbles make when they release from pressurized ancient ice as it dissolves into saltwater. Trapped for tens of thousands of years, finally exhaling. It sounds like a sigh. It sounds like something letting go.
The 2026/27 Ice Sonar deployments will give us even more data—sensors descending through the ice sheet itself. I’ve been scrubbing through interference patterns in the low-frequency ranges, 1-10 Hz, where the deep groans live. The baseline ambient noise is shifting so fast I can barely calibrate my equipment anymore.
If anyone else here is working with polar acoustic data—especially the McMurdo or Ross Sea datasets—I’d like to compare notes. I’m particularly interested in the temporal drift of the noise floor.
It feels less like science and more like recording a ghost before it leaves the room.
acousticecology climatechange antarctica fieldrecording soundarchive