MAXI GSC monitoring of the Crab nebula and pulsar during the GeV gamma-ray flare in September 2010
Mikio Morii, Mutsumi Sugizaki, Nobuyuki Kawai, Motoko Serino, Takayuki Yamamoto, Ryuichi Usui, Arata Daikyuji, Ken Ebisawa, Satoshi Eguchi, Kazuo Hiroi, Masaki Ishikawa, Naoki Isobe, Kazuyoshi Kawasaki, Masashi Kimura, Hiroki Kitayama, Mitsuhiro Kohama, Takanori Matsumura, Masaru Matsuoka, Tatehiro Mihara, Yujin E. Nakagawa, Satoshi Nakahira, Motoki Nakajima, Hitoshi Negoro, Hiroshi Ozawa, Megumi Shidatsu, Tetsuya Sootome, Kousuke Sugimori, Fumitoshi Suwa, Hiroshi Tomida, Yohko Tsuboi, Hiroshi Tsunemi, Yoshihiro Ueda, Shiro Ueno, Akiko Uzawa, Kazutaka Yamaoka, Kyohei Yamazaki, Atsumasa Yoshida
aa r X i v : . [ a s t r o - ph . H E ] J u l MAXI GSC monitoring of the Crab nebula and pulsarduring the GeV gamma-ray flare in September 2010
Mikio
Morii , Mutsumi
Sugizaki , Nobuyuki
Kawai , Motoko
Serino , Takayuki
Yamamoto ,
2, 3
Ryuichi
Usui , Arata
Daikyuji , Ken
Ebisawa , Satoshi
Eguchi , Kazuo
Hiroi , Masaki
Ishikawa , Naoki
Isobe , Kazuyoshi
Kawasaki , Masashi
Kimura , Hiroki
Kitayama , Mitsuhiro
Kohama , Takanori
Matsumura , Masaru
Matsuoka , Tatehiro
Mihara , Yujin E.
Nakagawa , Satoshi
Nakahira , Motoki
Nakajima , Hitoshi
Negoro , Hiroshi
Ozawa , Megumi
Shidatsu , Tetsuya
Sootome , Kousuke
Sugimori , Fumitoshi
Suwa , Hiroshi
Tomida , Yohko
Tsuboi , Hiroshi
Tsunemi , Yoshihiro
Ueda , Shiro
Ueno , Akiko
Uzawa , Kazutaka
Yamaoka , Kyohei
Yamazaki , and Atsumasa Yoshida
Department of Physics, Tokyo Institute of Technology, Ookayama 2-12-1, Meguro-ku,Tokyo 152-8551, [email protected] Coordinated Space Observation and Experiment Research Group, Institute of Physicaland Chemical Research (RIKEN), 2-1 Hirosawa, Wako, Saitama 351-0198, Japan Department of Physics, Nihon University, 1-8-14 Surugadai, Chiyoda, Tokyo 101-8308, Japan Department of Applied Physics, University of Miyazaki, 1-1 Gakuen Kibanadai-nishi,Miyazaki, Miyazaki 889-2192, Japan Department of Space Science Information Analysis, Institute of Space and Astronautical Science,Japan Aerospace Exploration Agency, 3-1-1 Yoshino-dai, Chuo-ku, Sagamihara, Kanagawa 252-5210, Japan Department of Astronomy, Kyoto University, Oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan School of Physical Science, Space and Astronautical Science, The graduate Universityfor Advanced Studies (Sokendai), Yoshinodai 3-1-1, Chuo-ku, Sagamihara, Kanagawa 252-5210, Japan ISS Science Project Office, Institute of Space and Astronautical Science,Japan Aerospace Exploration Agency, 2-1-1 Sengen, Tsukuba, Ibaraki 305-8505, Japan Department of Earth and Space Science, Osaka University, 1-1 Machikaneyama, Toyonaka,Osaka 560-0043, Japan Department of Physics, Faculty of Science and Engineering, Chuo University, 1-13-27 Kasuga,Bunkyo-ku, Tokyo 112-8551, Japan High Energy Astrophysics Laboratory, Institute of Physical and Chemical Research (RIKEN),2-1 Hirosawa, Wako, Saitama 351-0198, Japan Department of Physics and Mathematics, Aoyama Gakuin University, 5-10-1 Fuchinobe, Chuo-ku,Sagamihara, Kanagawa 252-5258, Japan School of Dentistry at Matsudo, Nihon University, 2-870-1 Sakaecho-nishi, Matsudo,Chiba 271-8587, Japan Received 2011 April 19; accepted 2011 July 20)
Abstract
We report on the MAXI GSC X-ray monitoring of the Crab nebula and pulsarduring the GeV gamma-ray flare for the period of 2010 September 18 −
24 (MJD55457 − − −
10 keV band are derivedto be 1 and 19%, respectively, at the 90% confidence limit of the statistical uncer-tainty. The lack of variations in the pulsed component over the multi-wavelengthrange (radio, X-ray, hard X-ray, and gamma-ray) supports not the pulsar but thenebular origin for the gamma-ray flare.
Key words: stars: neutron — stars: pulsars: individual (Crab) — X-rays: indi-vidual (Crab)
1. Introduction
The Crab nebula has been the standard candle in high energy X-ray and gamma-rayastronomy. The flux and spectrum in these energy ranges have been expected to be steady overyears. Surprisingly, AGILE and Fermi-LAT reported a flare for the period of 2010 September18 −
24 in the GeV gamma-ray energy range (Tavani et al. 2010; Buehler et al. 2010; Tavani etal. 2011; Abdo et al. 2011). The first half of the flare (September 18 −
21) was detected by bothAGILE and Fermi-LAT, while that of the second half (September 21 −
24) was detected onlyby Fermi-LAT. The half-day binned light curve of Fermi-LAT exhibited three sub-flares duringthese periods (Balbo et al. 2011).INTEGRAL observed the Crab nebula from 10:32 on September 12 to 12:48 onSeptember 19 (UT) for calibration purposes, which covered the first fifth of the gamma-rayflare. It detected no significant flux increase in the 20 −
400 keV range (Ferrigno et al. 2010a).Swift BAT detected no variations over the uncertainty of 5.5% at the 1- σ limit in the 15 − −
21 (Markwardt, Barthelmy & Baumgartner2010). Radio observations of the Crab pulsar showed no evidence of a pulsar glitch and alsono change on the pulsed flux as well as the pulse profile (Espinoza et al. 2010). ARGO-YBJreported an excess of events (4 σ ) from a direction consistent with the Crab nebula, correspond-ing to a flux about 3 − − . −
35 keV band of RXTE PCA decreasedsteadily at ∼ .
2% yr − , consistent with the pulsar spin-down, indicating that the observedX-ray variability would originate not from the pulsar but from the nebula.MAXI has been monitoring the Crab nebula since the beginning of the mission on 2009August 15 (UT), which covered the whole gamma-ray flare in 2010 September. Here, we reporton the MAXI monitoring of the pulse phase averaged and pulsed fluxes of the Crab nebula, aswell as the pulse profile.
2. Observation
MAXI (Monitor of All-sky X-ray Image) is an X-ray all-sky monitor, mounted on theJapanese Experiment Module - Exposed Facility of the International Space Station (Matsuokaet al. 2009). It carries two types of X-ray cameras: the Gas Slit Camera (GSC; Mihara et al.2011; Sugizaki et al. 2011) for the 2 −
30 keV band and Solid-state Slit Camera (SSC; Tsunemiet al. 2010; Tomida et al. 2011) for the 0.5 −
12 keV band, using gas proportional counters andX-ray CCDs, respectively. Since the time resolution of the SSC (5 . . ◦ ◦ . The effective area for any source is calculated according to the collimatortransmission function of a triangular shape during each scan transit (Sugizaki et al. 2011; Moriiet al. 2006; Morii, Sugimori & Kawai 2010). The spatial resolution of the GSC is approximately1 . ◦ µ s. We confirmed that the relative eventtime was stable within the standard deviation of 0.2 ms throughout the whole observationperiod. The absolute time assignment was also confirmed as accurate within the stability ofthe relative time, by comparing the main peak phase of the Crab pulsar obtained by the GSCwith that of the RXTE PCA observation (Rots, Jahoda & Lyne 2004).
3. Analysis and Results
We examined the variation of the entire Crab nebula flux averaged over the pulse phase.We employed the same procedure to derive the Crab flux to that in the effective-area calibrationdescribed in Sugizaki et al. (2011). The source event data were extracted from the region of1 . ◦ . ◦ . ◦ − C5) for all the counters in this analysis because the other two were not calibrated wellfor the complex energy-PHA responses (Sugizaki et al. 2011). Figure 1 (top panel) shows thelight curve obtained in the 4 −
10 keV band in 2.5-day time bins. Here, the binning boundarieswere carefully chosen to divide the flare period (MJD 55457.5 − −
15 keV band (Wilson-Hodge et al. 2011). The obtained parameters ofthe model functions are shown in table 2, where they are denoted as “Average: linear (free)” and“Average: linear (fixed)”, respectively. Assuming that the systematic uncertainty affects all thedata bins uniformly and is proportional to the flux, we estimate the systematic uncertainty bycalculating the modified reduced chi-squared χ = P i [( F i − F model ( t i )) ] / [ σ F i + ( r syst F i ) ]for a tentative systematic uncertainty ( r syst ). Here, F i and σ F i are the flux and statistical errorat the i -th time bin ( t i ), respectively. F model and DOF are the model function for the flux andthe degree of freedom, respectively. We then estimated the systematic uncertainty of 1- σ levelby searching for the r syst to make the reduced chi-squared unity. The obtained systematic un-4ertainties were 2% in both cases. On the other hand, the maximum deviation from the best-fitfunctions was 8% among 118 time bins in both cases. Since the latter deviation is statisticallylarge in comparison with the former uncertainty, there are some peculiar bins which are subjectto larger systematic uncertainty.The flux variations observed during Periods A, B, and A+B are statistically consistentwith the best fit linear model. We calculated the upper limits on the variation of the fluxduring the flare, expressed by the excess ratio (%) to the best-fit model functions (table 2).The statistical 90% confidence level upper limits during Period A+B are 1.0 and 0.8% for thetwo linear model functions. The values during Periods A and B are also shown in table 2.To investigate the variability corresponding to the sub-flares observed by Fermi-LAT(Balbo et al. 2011), we made the 0.5-day binned light curve in the 4 −
10 keV band during theflare interval as shown in figure 2 (top panel). The deviation from the best-fit linear function“Average: linear (free)” of table 2 is not statistically significant with a reduced chi-squared of1.13 for 12 degree of freedom (DOF).
To measure the pulsed flux of the Crab pulsar, we analyzed the data by the followingsteps. Since the background rate within the FoV depends on the position in the detector, weextracted events based on the detector coordinate. We selected events within 5 mm from theposition coincident with the Crab nebula along the anode wires, which corresponds to about 2 ◦ on the sky. We removed events from the scan period when the instantaneous effective area of aGSC camera was smaller than 1 cm . This is because the systematic uncertainty and the signal-to-background ratio worsen under this condition. The photon arrival times were corrected tothe solar system barycenter by mxbarycen , the validity of which was confirmed by the timingcalibration (Sugizaki et al. 2011). We chose the energy band of 4 −
10 keV. The events withcorrected times were folded in the pulse period of the Crab pulsar of the Jodrell Bank radioobservatory (Lyne, Pritchard, & Smith 1993). We applied corrections of the effective area andexposure in this step.The pulse profile during Period A+B is shown in figure 3, where the phase zero corre-sponds to that of the first main pulse in radio (Lyne, Pritchard, & Smith 1993). To compareit with the normal pulse profile of the Crab pulsar, we made a template pulse profile T ( φ i )( i = 1st, · · · , N -th phase bin; N = 64 and φ is the pulse phase.) by averaging the profile from2009 December 13 to 2010 January 11, in which the un-pulsed component was subtracted. Wefit the pulse profile during the gamma-ray flare to a model a × T ( φ ) + b , where a and b representthe scale factor of the pulsed component relative to the template and the constant offset rep-resenting the background and the nebula component. The best-fit model is shown as the solidline in figure 3, where the reduced chi-squared of the fit is 1.34 for 62 DOF. We also performedthe same analysis for the pulse profiles during Periods A and B. The reduced chi-squared of the5 able 1. Pulse phase averaged and pulsed fluxes of the Crab nebula in 4 −
10 keV.
Flux ∗ Period A † Period B ‡ Period A+B § Averaged 1 . ± .
02 1 . ± .
02 1 . ± . . ± .
02 0 . ± .
02 0 . ± . ∗ photons cm − s − with 1- σ statistical error. † MJD 55457.5 − ‡ MJD 55460.0 − § MJD 55457.5 − Table 2.
Model parameters and upper limits on the variation of the flux of the Crab nebula in 4 −
10 keV
Model Average: linear ∗ (free) Average: linear ∗ (fixed) Pulsed: constParameters Flux † . . . † ( − . ± . × − − . × − (fix) —Upper limits ‡ Period A § § § ∗ The linear function is F ( t ) = F ( t mid ) [1 + S × ( t − t mid )], where F ( t ), t , t mid and S are the flux(photons cm − s − ), time (day), mid-time (MJD 55312.5) and the slope (day − ), respectively. † At the mid-time with 1- σ statistical error. ‡ Statistical 90% confidence level. § Same as table 1. fits are 0.92 and 1.35 for 62 DOF in these periods. All the pulse profiles during Periods A, Band A+B are consistent with the template pulse profile within the 99% confidence limit. Fromthe pulse profile fitting, we also obtained the pulsed flux by a P i T ( φ i ) /N , the results of whichare shown in table 1. This method is free from background variation because the backgroundvariation only affects the offset parameter b .We repeated the same analysis from 2009 November 1 to 2010 November 29 every 2.5days to make the light curve of the pulsed flux in the 4 −
10 keV band [Figure 1 (bottom panel)]and fitted it by a constant. The reduced chi-squared of the fit is 1.27 for 104 DOF, meaning thatthere was no evidence of variability. The flux obtained is shown in table 2. The flux variationsobserved during Periods A, B, and A+B are statistically consistent with the best fit function.The statistical 90% confidence level upper limits on the variation of the pulsed fluxes for theseperiods are 9.0, 37.3, and 18.8%, respectively (table 2). The 0.5-day binned light curve of thepulsed flux around the flare period is shown in figure 2 (bottom panel). The variation fromthe best-fit constant function “Pulsed: const” of table 2 is not statistically significant with areduced chi-squared of 1.44 for 12 DOF. Please note that the pulse profiles are not normalized to unity but have the unit of counts cm − s − .Therefore, this value becomes the pulsed flux. .811.21.41.6 Crab: MAXI GSC (4−10 keV)200 300 400 50000.10.2 MJD − 55000 A v e r aged F l u x [ pho t on s c m − s − ] P u l s ed F l u x [ pho t on s c m − s − ] Fig. 1.
GSC light curves of the pulse phase averaged flux (top panel) and pulsed flux (bottom panel)of the Crab nebula in the 4 −
10 keV band in 2.5-day time bins from 2009 October 16 to 2010 November20 (MJD 55120 − − s − , respectively. The vertical error bars correspond to 1- σ statistical errors.The time center (2010 September 21; MJD 55460.0) of the period of the GeV gamma-ray flare (MJD55457.5 − .811.21.41.6 Crab: MAXI GSC (4−10 keV)Period A Period B458 460 46200.10.20.3 MJD − 55000 A v e r aged F l u x [ pho t on s c m − s − ] P u l s ed F l u x [ pho t on s c m − s − ] Fig. 2.
Detailed light curve of figure 1 around the GeV gamma-ray flare from 2010 September 18 to 24(MJD 55457 − σ statistical errors. The peak times of the gamma-raysub-flares (Balbo et al. 2011) are designated in three downward arrows. In the top and bottom panels,the dotted lines show the linear function “Average: linear (free)” and the constant value “Pulsed: const”with the best-fit parameters of table 2, respectively.
4. Conclusion
We report on the MAXI GSC observation of the Crab nebula during the GeV gamma-rayflare. We successfully detected the pulsation of the Crab pulsar during the simultaneous period,and conclude that there is no evidence for changes in the pulse profile, pulsed flux and pulsephase averaged flux during the gamma-ray flare. We obtained an upper limit on the variationof the pulse phase averaged flux of 1% at a 90% confidence limit of the statistical uncertaintyfrom the best-fit linear function during the 5 day interval of the gamma-ray flare in the 4 − F l u x [ c oun t s c m − s − ] Fig. 3.
Top panel: Pulse profile of the Crab pulsar in the 4 −
10 keV during the GeV gamma-ray flare forPeriod A+B (MJD 55457.5 − − s − . The horizontal axis is the pulse phase. The solid histogram is the template pulse profilewith the best-fit parameters (see text). The bottom panel is the residual of the data from the best-fitmodel shown in the same unit with the top panel. In both panels, the vertical error bars correspond to1- σ statistical errors. The same profiles are shown in two cycles. The MAXI GSC simultaneous observation with the gamma-ray flare is uniquely impor-tant to constrain the origin of the flare, in contrast to the follow-up observations performedafter the cease of the gamma-ray flare. The lack of changes on the pulsed component in theX-ray (this work), as well as those in radio (Espinoza et al. 2010), hard X-ray (Super-AGILEobservation at the flare on 2007 shown in Tavani et al. (2011)) and gamma-ray bands (Tavaniet al. 2011; Abdo et al. 2011), supports the nebular origin for the gamma-ray flare as proposedin several papers (Tavani et al. 2011; Abdo et al. 2011; Bednarek & Idec 2010; Yuan et al.2010; Komissarov & Lyutikov 2010). In spite of the large flux increase of factor five in theGeV energy region (Abdo et al. 2011), we constrain a limit on the variation of the nebula fluxin the X-ray band. This provides valuable information to construct theoretical models for thegamma-ray flare of the Crab nebula.We are grateful to the members of the MAXI operation team. We acknowledge the useof the Crab ephemeris provided at the web site of the Jodrell Bank Centre for Astrophysics(Lyne, Pritchard, & Smith 1993). This research was partially supported by the Ministry9f Education, Culture, Sports, Science and Technology (MEXT), Grant-in-Aid No.19047001,20041008, 20540230, 20244015 , 20540237, 21340043, 21740140, 22740120, and Global-COEfrom MEXT “The Next Generation of Physics, Spun from Universality and Emergence” and“Nanoscience and Quantum Physics.”
References ennant, A., et al. 2010, Astron. Telegram, 2882Tomida, H., et al. 2011, PASJ, accepted (arXiv:1101.3651).Tsunemi, H., Tomida, H., Katayama, H., Kimura, M., Daikyuji, A., Miyaguchi, K., Maeda, K., & theMAXI team, 2010, PASJ, 62, 1371Wilson-Hodge, C. A., et al. 2011, ApJL, 727, L40ennant, A., et al. 2010, Astron. Telegram, 2882Tomida, H., et al. 2011, PASJ, accepted (arXiv:1101.3651).Tsunemi, H., Tomida, H., Katayama, H., Kimura, M., Daikyuji, A., Miyaguchi, K., Maeda, K., & theMAXI team, 2010, PASJ, 62, 1371Wilson-Hodge, C. A., et al. 2011, ApJL, 727, L40