Surface echoes can put whales hundreds of meters off in passive acoustic tracking

An oceanographer chasing what he assumed was a bug in his own code has instead described a propagation effect that biases how researchers locate whales by sound. John Spiesberger and Eugene Terray report in Physical Review E that interference between a call's direct path and its surface reflection can shift the apparent arrival time enough to misplace a nearby animal by hundreds of meters ^[1][5].

The operational stakes are set by how far low-frequency calls travel. A fin whale can be heard 100 kilometers (62 miles) away underwater on a single hydrophone, according to Spiesberger, and localization works by comparing when that sound reaches receivers spread across the seafloor ^[3][4]. Arrival-time differences are converted to position using an assumed sound speed, roughly 1,500 meters per second in seawater ^[9]. If the assumed speed is wrong, the geometry is wrong.

Spiesberger found the discrepancy while refining a program meant to compute the correct sound speed for his tracking equations ^[8]. The first result came back near 1,000 meters per second; later runs sometimes produced 3,000 ^[8][10]. That is a spread of roughly a factor of three, from about two-thirds of the nominal speed to twice it ^[19]. He assumed a coding error, and only after a few hours of checking concluded the behavior was physical: the receiver was picking up both the direct signal and its echo off the surface when the whale was shallow ^[10][11].

The mechanism is what physicists call temporal interference, the same effect that makes an over-the-air TV picture fade when two paths arrive out of phase ^[12]. Because the two arrivals interfere, the peak of the received energy can move earlier than the direct path alone would allow ^[13][7]. Spiesberger is careful about what that does and does not mean: what appears to speed up is the location of the strongest peak, not the signal carrying information ^[14]. Physicists have known for more than a century that a wave whose shape is changing can appear to outrun light while the information in it cannot, and Spiesberger says causality is not overturned ^[15]. The Einstein framing is the paper's route to publication in a physics journal; the consequence for anyone running hydrophone arrays is duller and more useful, which is that a standard assumption is systematically off in a specific, identifiable geometry.

Scale check on the error: a 300-meter range mistake corresponds to about 0.2 seconds of travel time at 1,500 meters per second ^[20]. That is large compared with the timing precision these arrays are built to deliver, and it is concentrated where whales spend time, near the surface. The authors say the correction could improve tracking used by conservationists ^[21].

The honest caveat is that this is theory and simulation so far ^[16]. Spiesberger plans a bench test with microphones, bouncing sound off a hard floor as a stand-in for the sea surface, and says an optical version using a beam splitter, a reflector and a detector could follow ^[17]. He says which medium proves easier to measure first is an open question ^[18].