NuSTAR observations of Swift J1756.9-2508 during its April 2018 outburst

ATel #11603; Lucien Kuiper (SRON-The Netherlands), Sergey Tsygankov (Univ. of Turku-Finland), Maurizio Falanga (ISSI-Switzerland), Duncan Galloway (Monash Univ.-Australia), Juri Poutanen (Univ. of Turku-Finland)
on 3 May 2018; 17:03 UT
Credential Certification: Lucien Kuiper (L.M.Kuiper@sron.nl)

Tweet

We report the results of spectral and timing analyses of NuSTAR (3-79 keV) and Swift (0.3-10 keV) observations of the accretion-powered millisecond pulsar Swift J1756.9-2508, performed during its recent April 2018 outburst (ATel #11497, #11502, #11505, #11523, #11566 and #11581).

On our request the source was observed twice with NuSTAR on April 8 and 14, 2018 for 39.5 ks and 61 ks, respectively, and was monitored with the XRT aboard The Neil Gehrels Swift Observatory every other day during the whole outburst (April 2-24, 2018; 10.2 ks exposure).

The first NuSTAR observation was performed on April 8, 2018 starting at 08:31:09, when the source was already in the fading phase of the outburst. The source was significantly detected up to 70 keV allowing us to construct a broad-band spectrum for the first time. The closest Swift-XRT observation was performed on the same day (ObsID 00088662001; 2.2 ks) and was used to cover energies below 3 keV. The joint spectrum was fitted with an absorbed comptonized spectrum (phabs*compPS) under the assumption of a slab geometry. We obtained an excellent fit (red. chi^2=1.01; 1395 dof; NuSTAR FPM-A,B and Swift-XRT) with the following parameters: N_H=(3.9 +/- 0.1)e22 cm^(-2), electron temperature kTe=(29.5 +/- 1.7) keV, Thomson optical depth tau_T=(2.0 +/- 0.1) and black body temperature frozen at kT_bb=0.1 keV. Unabsorbed fluxes in 2-10 and 20-60 keV energy bands are F_2-10=(2.20 +/- 0.5)e-10 and F_20-60=(2.60 +/- 0.05)e-10 erg/(s cm^2), respectively. No evidence for an iron emission line was found.

The second NuSTAR observation was taken about one week later on April 14, 2018 starting at 02:56:09, and the source was undetectable in the NuSTAR data in agreement with quasi-simultaneous Swift-XRT data.

We also performed a timing analysis using NuSTAR data of the first April 8, 2018 observation. For this purpose a phase coherent timing model, composed of updated pulsar and orbital ephemerides, was constructed from NICER data covering the 3--11 April, 2018 period (see P.M. Bult et al., ATELs #11502 and #11581). Adopting the DE405 solar system ephemeris in the barycentering process we obtained in our 2 parameter optimisation procedure a pulse frequency of 182.06580384(4) Hz at Epoch MJD 58216 (TDB) and a time of ascending node Tasc of 58211.01753(6) MJD (TDB). Errors are at 1 sigma confidence level. In this procedure we kept the orbital period, Porb, projected orbital size, asini,fixed to the values given in MNRAS 403, 1426 (Patruno et al. (2010)). A 3 parameter optimisation with also Porb free gives a best fit value of 3282.35(10) s, fully consistent with the value given in MNRAS 403, 1426.

Folding the NuSTAR timing data (using the latest clock correction information) on the updated ephemeris yielded highly significant pulsed emission up to ~50 keV: 3-10 keV 25.8 sigma, 10-20 keV 15.2 sigma and 20-60 keV 7.5 sigma, adopting a Z_2^2-test. The (background corrected) pulsed fractions are 6.2% for all three energy bands, consistent with the NICER findings (ATEL #11502). The pulse shape does not seem to be dependent on energy, and is fully compatible with the 3-10 keV NICER profile, though the NuSTAR pulse arrives +0.37(4) ms later than the NICER pulse (see http://www.sron.nl/~kuiper/ATEL/SwiftJ1756.9-2508/ for the NICER and NuSTAR pulse profiles).

We thank the NuSTAR and Swift teams for scheduling our ToO observations.