Detection of a Millisecond Periodicity in BATSE Short Gamma-Ray Bursts

Abstract

Coherent oscillations at kilohertz frequencies have recently been detected in a small number of gamma-ray bursts (GRBs), suggesting quasi-periodic dynamics in their central engines. A prominent example is GRB~230307A, which exhibited a brief, highly coherent, energy-dependent periodic signal interpreted as the possible spin signature of a nascent millisecond magnetar formed after a compact binary merger. Motivated by these developments, we conducted a comprehensive search for similar signals, accounting for both temporal and spectral dependencies, in 532 short GRBs with time-tagged event data recorded by the Burst and Transient Source Experiment (BATSE) onboard the \textit{Compton Gamma-Ray Observatory}. Within this sample, we identify a single statistically significant case: GRB~960616 (BATSE trigger~5502), in which the $\sim$30~ms main emission episode is coherently modulated at 1100~Hz, with the strongest modulation above 320~keV and a fractional amplitude of $\sim$47\%. Assuming the presence of a coherent periodic modulation, we use data-driven Monte Carlo simulations to place an upper limit of $\sim$8\% on the fraction of the total radiated energy that can be modulated by the QPO. This event, exhibiting a periodicity at $\sim$0.91~ms, further supports the possibility that millisecond periodicities can arise during GRBs in merger-driven scenarios.

Figures (11)

• Figure 1: Results of the periodicity search in BATSE TTE data.a--c: Marginal and joint distributions of 186,127 candidate frequencies ($f_x$) and Rayleigh powers ($R_x$) derived from 100-ms segments extracted from different LADs and energy channels of 531 short GRBs (excluding trigger 6634). The orange curve in c shows the probability density function (p.d.f.) of the ensemble of all $R_x$. Orange dashed lines in b and c mark the median threshold power $R_{50}$ defined by $G(R_{50})=0.5$. d: Distribution of expected powers derived from 94,766 segments with $R_x \ge R_{50}$ at trial frequencies between 500 and 2500 Hz. The red curve denotes the asymptotic p.d.f. obtained from a generalized extreme-value distribution ($\mathcal{H}(R)$) fit. e: Single-trial FAPs as a function of trial frequencies. The red dashed lines indicate significance levels after accounting for the number of frequency trials. Only trigger 5502 shows a $>3\sigma$ excess at 1083 Hz, while trigger 6412 exhibits a weaker $\sim$1$\sigma$ excess at 625 Hz.
• Figure 2: WWZ spectrograms of GRB 960616 (detector resolved). Light curves recorded by different LADs are shown at a temporal resolution of 256 $\mu$s, with the corresponding incident angles indicated. Red boxes highlight the $\sim$1100 Hz signal, which is clearly detected by LAD0 and LAD1 but not prominently observed in LAD2 and LAD4.
• Figure 3: WWZ spectrograms of GRB 960616 (energy-channel resolved). Light curves from LAD0 and LAD1 are combined and shown at a temporal resolution of 256 $\mu$s for the four BATSE energy channels, with the corresponding energy ranges indicated. Red boxes highlight the $\sim$1100 Hz signal, which is clearly detected in channel 4 and marginally visible in channel 3.
• Figure 4: PDS of the main peak of GRB 960616. The PDS is normalized following 1975ApJS...29..285G. The gray band marks the red-noise regime (0--500 Hz), which is excluded from the Bayesian model fitting. The red shaded regions indicate the 1$\sigma$, 2$\sigma$, and 3$\sigma$ confidence intervals expected for the white-noise component in addition to the pure Poisson noise (mean power $=1$; $\chi^2_2$ statistics). The black curve shows the best-fit model, consisting of a Lorentzian QPO component plus an additional constant white-noise term above the Poisson baseline. A significant peak is detected at $\sim$1083 Hz with a power of $\sim$26.3. Even when treating all 186,327 trials as independent, the probability of such a peak arising by chance remains only $\sim$$7\times10^{-7}$ (from the $\chi^2_2$ distribution), indicating that the periodic signal is very significant.
• Figure 5: Pulse profile of GRB 960616 folded at 1100 Hz. The black steps show the phase-folded pulse profile of GRB 960616, using data from LAD0 and LAD1 within energy channel 4. Error bars indicate the 1$\sigma$ uncertainties. The dashed horizontal line marks the mean count ratio (unity). The red curve represents the best-fit sinusoidal model, and the red shaded region denotes its 1$\sigma$ confidence interval.