2−10

1−10

1

10

-1 s

-2Ph

oton

s cm

2-4 keV4-10 keV10-20 keV

1

Har

dnes

s ra

tio

HR1HR2

58000 58050 58100 58150 58200 58250 58300 58350 584002017-09-04 2018-10-09

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2Pulse phase

1

10-1 s

-2Ph

oton

s cm

MJD58064580745808458094581045811458124581345814458154

3810

3910

)-1

Lum

inos

ity (

erg

s

Hardness ratio

1

10-1 s

-22-

20 k

eV p

hoto

ns c

m

3810

3910

)-1

Lum

inos

ity (

erg

s

0.2 0.5 1.0 2.0

HR1HR2HR3

Swift J0243.6+6124 (Swft J0243) is a new Be/X-ray binary pulsar discovered on October 3, 2017. The source was first identified as a new X-ray object by the Swift BAT survey. The coherent pulsation with the ~9.86 s period was confirmed by the Swift XRT observation and the Fermi-GBM pulsar analysis. The MAXI all-sky survey had also recognized the X-ray activity, but could not resolve the source from the known nearby X-ray object, LS I +61 303. After the discovery, the source has been observed by many X-ray satellites. X-ray monitoring by MAXI, Swift, Fermi/GBM revealed that the outburst continued for more than 4 months, and the flux reached over 5 Crab at the peak. On the other hand, the recent GAIA result suggests the source distance to be ~7 kpc, meaning that the X-ray luminosity reached the 10 times of the Eddington luminosity (10 LEDD) at the peak. We performed detailed MAX-GSC data analysis taken at around the flux peak, using the Fermi/GBM pulsar ephemeris. X-ray spectra in the 2-30 keV band were largely approximated by power-law functions with a higher-energy cutoff and an iron-K emission line, which is typical in X-ray binary pulsars. The results clarified the spectral softening due to the flux increase, as observed by NICER and Fermi/GBM. The folded pulse profiles show the energy-dependent change due to the intensity. We also investigated the relation between the luminosity and the spin-up rate and found that the positive correlation close to the proportionality smoothly extends up to the 10 LEDD. Assuming the accretion-torque model proposed by Ghosh & Lamb, the surface magnetic filed is estimated to be in a order of 1012 G.

MAXI results of the Galactic ultra-luminous X-ray pulsar Swift J0243.6+6124

M. Sugizaki, M. Oeda, N. Kawai (Tokyo Tech.), T. Mihara, K. Makishima (RIKEN), M. Nakajima (Nihon Univ.) and MAXI Team

4. Luminisity – spin-up relation

1. Light curves and hardness ratios by MAXI GSC

3. X-ray spectra

2. Pulsar analysis using Fermi/GBM ephemeris data

5. Summary

10Energy (keV)

3−10

2−10

1−10

1

10

-1 k

eV-1

s-2

Phot

ons

cm

MJD 58060-58070MJD 58070-58080MJD 58080-58090MJD 58090-58100MJD 58100-58110MJD 58135-58145MJD 58145-58155MJD 58155-58165MJD 58165-58175

)-1 erg s39 (10XL

)-1

Hz

s-1

0 (

10ν

1.0 2.0 3.0

0.5

1.0

1.5

2.0

Ghosh & Lamb model G121.0 x 10 G122.0 x 10

101.5101.6101.7101.8101.9

102102.1

Spin

freq

. (m

Hz)

Ferm

i/GBM

1

10

-1 s-1

Phot

ons

cmPh

ase-

aver

age

58040 58060 58080 58100 58120 58140 58160MJD

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

Puls

e ph

ase

1

10

-1 s-1

Phot

ons

cm

Swift J0243 X-ray light curves obtained by MAXI GSC in 2-4 keV, 4-10 keV, and 10-20 keVbands from 2017 September to 2018 July for about a year. Data in shaded area were taken by degraded GSC instruments which have large systematic uncertainty.

Time evolution of hardness ratios (HRs) HR1 = 4-10 keV / 2-4 keVHR2 = 10-20 keV / 4-10 keV.

HR-intensity diagrams for 3 HRs of HR1 = 4-10 keV / 2-10 keVHR2 =10-20 keV/ 4-10 keVHR3 = 8-16 keV / 4-8 keV.The ordinate on the right-side vertical axis represents the luminosity scale calculated from the source distance of 7 kpc and the average spectrum.

All of the three diagrams show the negative correlations, as reported by Wilson+2018.The detail profiles are different among the energy bands.

Fermi/GBM pulsarspin-frequency data

MAXI/GSC pulse-phase-average flux

MAXI GSC event data are folded with the pulsar spin period measured by the Fermi/GBM pulsar project (Wilson+2018). The period changes between the Fermi/GBM measurements are interpolated by the cubic-spline function.

Time evolution of the obtained MAXI-GSC 2-20 keV pulse profile during the outburst from MJD 58040 to 58160.The transition from the simple sinusoidal double-peak to the broad single-peak at around MJD 58120 during the outburst decay phase is confirmed.

Sample of MAXI-GSC 2-20 keVpulse profiles for every 10 days during the outburst decay phase.

Pulsed fraction f,

! = #max − #min#min

against X-ray intensity.

f is larger in the higher energy band.

X-ray spectra were extracted for 10-day intervals from the outburst peak. They fitted well to a model consisting of a cutoff power-law plus a blackbody (BB) and a gaussian for Fe-K line (Gaus), * + = ,+-. exp − 1

1c + 44 + 5678,which is typical in X-ray binary pulsars.The figure shows the unfolded photon spectra. The spectral softening due to the flux increase is seen.

The luminosity – spin-up relation extracted from the MAXI/GSC X-ray light curve, spectra, and the Fermi/GBM pulsar data. The source distance of 7 kpc is employed from the GAIA DR2. The figure compares the observed relation with those expected from the Ghosh & Lamb 1979 model with the canonical neutron-star (NS) mass (1.4 solar mass), radius (10 km), and the surface magnetic fields B=1x1012, 2x1012 G. The best-fit suggest B=1.2x1012 G.

Swift J0243.6 is considered as the first Galactic ultra-luminous X-ray pulsar which exhibits the extraordinary large luminosity (>10 LEDD), from the X-ray flux and the source distance determined by GAIA. Using the MAXI GSC monitoring data and the Fermi/GBM pulsar ephemeris, we confirmed and found the following facts, 1. Spectral softening due to the flux increase in the high-luminosity regime above the critical luminosity Lcrit~ LEDD, as observed by NICER and Fermi/GBM (Wilson+2018)2. Pulse-profile transition from the single-peak to the double-peak at around the critical luminosity (Tsygankov+2018, Wilson+2018).3. X-ray spectra represented by a cutoff power-law plus a blackbody and a gaussian for the Fe-K line.4. The positive correlation between the luminosity and spin-up rate, smoothly extended up to the maximum of ~ 10 LEDD. Assuming the canonical NS mass (1.4 solar mass),

radius (10 km), and the Ghosh & Lamb 1979 disk-accretion model, the best-fit function suggests B=1.2x1012 G.As for points 1. and 2, the similar results have been observed in other Galactic XBPs. These behaviors are interpreted as the transition between two mass-accretion regimes dominated by Coulomb collision and radiation, respectively. Although the estimated Lcrit suggests a possible strong magnetic filed of B~1013 G, it requires several assumptions (Tsygankov+2018, Wilson+2018). The points 3. and 4. are typical in XBPs. Thus, the difference to cause the extraordinary large luminosity is still under debate. We consider further detailed data analysis and observations.

-1 s-22-20 keV Photons cm

Puls

e fra

ctio

n

1.0 2.0 5.0 10.0 20.0

0.2

0.5

1.0

2-20 keV

10-20 keV

3810 3910 )-1Luminosity ( erg s

B =1.2x1

012 G