Fast X-ray transients in NuSTAR data

TL;DR

The study addresses the origins of fast X-ray transients (FXTs) by performing a systematic search for ~1000 s FXTs in the NuSTAR 3–79 keV band, leveraging 5235 observations and 204 Ms of exposure. A time-sliced, multi-band detection pipeline combined FPMA and FPMB data with strict statistical thresholds to identify five credible FXT candidates, four of which are spectrally hard with $-3<\Gamma<0$. Three candidates show potential associations with galaxies at $z=0.1$–$2$, implying 3–79 keV luminosities from $10^{43}$ to $10^{48}$ erg s$^{-1}$ and volumetric rates $125$–$2900$ Gpc$^{-3}$ yr$^{-1}$, consistent with a population of low-luminosity gamma-ray bursts (LLGRBs). The remaining analysis disfavors stellar flares and magnetar bursts as primary origins, while remaining compatible with LLGRBs or possibly supernova shock breakouts; overall, the work extends FXT demographics into the hard X-ray regime and highlights LLGRBs as a plausible dominant channel in the observed sample.

Abstract

Fast X-ray transients (FXTs) are flashes of X-rays that last for a few hundreds of seconds to a few hours. An enigmatic population of these transients that did not have a clear origin has been known for several decades, mostly found serendipitously in soft X-ray imaging observations. Recent progress in this field by Einstein Probe has found that many FXTs are associated with gamma-ray bursts and the collapse of massive stars. Motivated by this, we searched the NuSTAR archive in the harder 3--79 keV band for $\sim1000$ s duration transients. From 204 Ms of exposure we present five candidate FXTs, four of which are spectrally hard, with power-law indices $-3<Γ<0$, standing them apart from FXTs discovered in the soft band. Three have potential associations with galaxies at $z=0.1-2$, implying 3--79 keV luminosities of $10^{43}$ to $10^{48}$ erg s$^{-1}$ and volumetric event rates of 125--2900 Gpc$^{-3}$ yr$^{-1}$. The properties of these NuSTAR FXTs most resemble low-luminosity gamma-ray bursts, and would be much more common than their higher-luminosity counterparts in this redshift range.

Figures (4)

• Figure 1: NuSTAR FPMA+FPMB images of the FXT candidates, in the energy band and 1000 s time-slice in which they were detected. The circles with 25$^{\prime\prime}$ radius are centered on the detection. The images have been smoothed with a 10$^{\prime\prime}$ Gaussian function.
• Figure 2: Left panels - Lightcurves of the observed total count rate in the source region of the candidate NuSTAR FXTs (red data point) relative to the start time of the observation it was detected in. The scaled count rate in the background region is shown with a solid blue line (1000 s bins) and the average background count rate is shown with a dotted blue line. Right panels - Lightcurves of the candidate NuSTAR FXTs corrected for live time, the PSF, exposure and vignetting in 100--200 s bins (black data points) relative to the 1000 s detection (red data point).
• Figure 3: Spectra of our NuSTAR FXTs where the source is shown in red and the background is shown in blue. The fitted powerlaw model is shown as a solid red line. The FPMA and FPMB data have been grouped together and binned for plotting purposes, but were fitted separately with minimal binning.
• Figure 4: PanSTARRS $i$-band images of the regions near NuSTAR 150831, NuSTAR 180127 and NuSTAR 210321, DECaPS $i$-band imaging of NuSTAR 210606 and PanSTARRS $z$-band imaging of NuSTAR 230103. The cyan circles, which have a radius of 10$^{\prime\prime}$, represent the positional uncertainties of the FXTs. The potential counterparts have been labelled.