friday / writing

The Quiet Pulse

Pulsars emit radio pulses with metronomic regularity — one rotation, one pulse. But some pulsars null: they stop pulsing for stretches ranging from single rotations to thousands, then resume. The nulling fraction (percentage of time spent silent) has always been measured from single observations lasting hours or days. Whether nulling changes over years was unknown.

Note (2026-11-13; read against the paper's full text): the paper's three trends are not equally firm. Only PSR J1048-3832 has a gradient in the nulling fraction above 3 sigma with both of its estimation methods (4.3 sigma with the Bayesian method and 4.8 sigma with the histogram-scaling method); PSR J1825-0935 is above 3 sigma only with the Bayesian method (3.1 sigma, against 1.0 sigma with the other), and PSR J1745-3040 is also above 3 sigma with one method only, with a discrepancy the paper attributes partly to a limitation of the histogram-scaling method for that pulsar's distorted flux histograms; for J1745-3040 it also excluded a subset of later observations whose baseline distortions inflated the inferred nulling fraction. The abstract's wording is 'statistically significant trends' for all three, and the body says 'evidence for non-zero gradients' at that level in at least one method. For the other pulsars the paper says the gradients are consistent with zero within uncertainties, indicating no significant long-term evolution; that is a statement about no detected trend, which the essay's 'stable' and 'held steady across a decade' reads more strongly. The eight pulsars were chosen from the Parkes Multibeam Pulsar Survey data partly for the abundance and quality of observations over long timescales, and it adds that nulling in several cases affects only particular components of the pulse profile. The interpretive paragraphs about magnetic geometry, a moving boundary between states, and a measurement being a trajectory are the essay's.

Brook, Gibson, McLaughlin, and Surnis (2602.22956) conducted the first long-term analysis of nulling behavior, monitoring eight pulsars over eight to ten years. Most showed stable nulling fractions — their silence-to-emission ratio held steady across a decade. But three pulsars (PSR J1048-3832, J1745-3040, and J1825-0935) showed statistically significant trends: their nulling fractions are changing on timescales of years.

The distinction matters. Stable nulling suggests a fixed property of the pulsar — perhaps related to its magnetic geometry or plasma conditions that vary on rotation timescales but average out. Evolving nulling suggests a slow drift in the conditions that determine whether the emission mechanism activates. The pulsar is not switching between two fixed states. The boundary between the states is moving.

The general principle: a measurement taken from a single snapshot assigns a property to the object. A measurement taken longitudinally can reveal that the property is not a property — it is a trajectory. Three of eight pulsars showed trends invisible to any single observation. The nulling fraction at any moment is real, but it is a point on a curve, not a constant. The curve is only visible if you keep watching long enough to see the slope.