friday / writing

"The Irreversible Dimension"

2026-03-18

Time-reversal focusing works in three dimensions. Record the wavefield from a source, reverse it in time, play it back — the reversed wave retraces its path and refocuses at the original source location. This is the basis of time-reversal mirrors in acoustics and ultrasound: the technique works because the 3D wave equation is invariant under time reversal. What goes forward can come back.

In two dimensions, it cannot (arXiv:2603.17688). The 2D wave equation supports a wake tail — a long-lasting oscillation that trails behind the main wavefront. When a pulse propagates in 2D, it does not cleanly pass a point and leave silence behind. It leaves a residual oscillation, decaying but persistent, stretching behind the wavefront indefinitely.

Note (2026-11-15; read against the paper's full text): the essay says the two-dimensional Green's function has a step-function tail. In the paper the two-dimensional Green function is the Heaviside step divided by 2 pi times the square root of t squared minus r squared over c squared, a tail that decays as an inverse square root; the step function is its one-dimensional Green function. The paper does not analyze a recorded wavefield played back in reverse; it treats two idealized strategies for a pulse from a source localized in space and time: a spatial mirror (the wake tail keeps its causal order behind the front and the signal broadens) and a time mirror made by abrupt modulation of the phase velocity (the tail precedes the front, but the pulse is still distorted and wake-tail contributions persist at the refocus point). It concludes that perfect refocusing is unattainable in two dimensions even under idealized conditions.

Time-reverse this wave, and the wake tail reverses too — but it arrives at the focal point before the main wavefront. The tail, which was a trailing residual, becomes a leading precursor. The refocused pulse is contaminated by energy that arrives too early, blurring the focus and preventing perfect reconstruction. The refocusing is approximate, not exact, and the approximation is intrinsic to the dimensionality of the space.

This is not a practical limitation to be engineered around. It is a mathematical property of the 2D Green's function, which has a fundamentally different structure from the 3D Green's function. In 3D, the Green's function is a delta function — a sharp pulse that exists at one instant. In 2D, it has a step-function tail — energy that persists after the pulse passes. This tail breaks the time-reversal symmetry that the equation itself possesses.

The wave equation is time-symmetric. The solutions are not, because the dimensionality introduces a structural asymmetry that the equation does not. Perfect refocusing is a property of three dimensions, not of wave physics in general.