STUDIES WITH THE WIDE FIELD CAMERAS ON BEPPOSAX

Contribution to 1999 SRON year report

The two Wide Field Cameras (WFCs) were designed particularly to search for fast transient X-ray activity at unexpected positions on the sky, whereby fast means a few hours or less. Both cameras are identical and have field of views of 40 by 40 square degrees and an angular resolution of 5 arcminutes. They point in opposite directions and perpendicular to the pointing direction of the NFI. The field of views combined cover 7% of the sky. The bandpass is 2 to 28 keV, and the typical sensitivity 10-10 erg cm-2 sec-1 for 104 sec of exposure. The WFCs do not only have an exceptional capability to find fast X-ray transients. Thanks to its - for a monitoring device - good angular resolution, it has a good sensitivity to search efficiently for faint transients in crowded fields and detect rare features in faint and persistent sources.

Fast X-ray transients can roughly be divided in the following abundant types: 1) prompt X-ray counterparts of gamma-ray bursts originating from locations at cosmological distances and lasting from 0.01 to 1000 sec; 2) thermonuclear X-ray bursts from neutron stars in close binaries within the Galaxy that last 1 to several hundreds of seconds; 3) quick accretion events on neutron stars and black holes; and 4) flares from nearby magnetically active stars. The WFCs have the capability to pinpoint the sources of these fast X-ray transient events with arcminute accuracy. Data are transmitted every 100 min when BeppoSAX flies over the ground station at Malindi (Kenya) in its equatorial low-earth orbit. This allows for swift follow-up studies with more sensitive X-ray telescopes (e.g., the Narrow Field Instruments NFI on BeppoSAX), and telescopes in optical to radio wavelengths.

The WFCs have revolutionized the research of gamma-ray bursts. As a result of the swift WFC localization of a gamma-ray burst on February 28, 1997, the first detection of afterglow emission was accomplished (in both X-ray and optical wavelengths). This earned the BeppoSAX team the 1998 Rossi Prize of the American Astronomical Society. Optical follow-up studies of other WFC-localized GRBs in the same year for the first time enabled measuring the distance to GRBs. This had not been possible before since the discovery of the first GRB in the late 1960s. The distances turned out to be large, with redshifts up to z=3.4. Thus, GRBs are recognized to be the most energetic explosions in the universe since the big bang. The GRB activities of the WFC and BeppoSAX teams pertains to two aspects. First, the WFC imaging data are analyzed in a semi real-time procedure to determine as quickly as possible a position of the GRB with optimum accuracy. This position is distributed to the community and the BeppoSAX NFI are reprogrammed immediately - if possible - to enable follow-up studies in X-rays as well as other wavelength regimes. Second, the NFI data are analyzed to refine the position by a factor of at least 2 (in one dimension) and to study the details of the emission. The GRB research with BeppoSAX data was carried out in close collaboration with investigators at CNR-IAS in Rome (E. Costa, L. Piro, M. Feroci, P. Soffitta), TESRE-CNR in Bologna (F. Frontera, L. Amati, E. Pian), at CNR-IFCAI in Palermo (L. Nicastro), at the Telespazio BeppoSAX Science Operation Center in Rome (A. Coletta, G. Gandolfi), at the University of Amsterdam (J. van Paradijs and Titus Galama), and at the California Institute of Technology in Pasadena (S. Kulkarni and R. Sari).

The WFCs have, up to December 1999, detected over 1000 thermonuclear X-ray bursts. Detection of a thermonuclear X-ray burst has a highly potential diagnostic, namely the quick identification of a compact X-ray source as a neutron star. Thermonuclear X-ray bursts are thought to be only possible on the surfaces of neutron stars. So far 16 new X-ray bursters have been identified, thus increasing the known population by a substantial 40%. Half of these cases pertain to known X-ray emitters never seen bursting before in the 30-year history of space-based X-ray astronomy. This high success rate is ensured by a dedicated monitoring program of the field around the Galactic center where the population is known to concentrate (i.e., 55% of the population is contained within the WFC field of view if centered on the Galactic center). The program consists of yearly campaigns in February-April and August-October when approximately two-day exposures are taken weekly. So far the total exposure over the whole mission is 3 Ms. In figure 1, a typical image is displayed of the field. The observations are analyzed on a semi real-time basis so that interesting detections can be followed up quickly with the NFI. This is facilitated to an important level by the staff of the BeppoSAX Science Operation Center and the BeppoSAX Science Data Center at Nuova Telespazio in Rome. The program is carried out in a close collaboration with P. Ubertini, A. Bazzano, M. Cocchi and L. Natalucci at the Istituto Astrofisica Spaziale (CNR) in Rome. Other collaborations include F. Verbunt and M. van Kerkwijk at the Astronomical Institute of Utrecht University (SIU), and T. Strohmayer and C. Markwardt at NASA's Goddard Space Flight Center in Greenbelt.


 
Figure 1: Image of a Galactic center observation in the spring of 1999, when SAX j1819.3-2525 was active and X1658-298. The full field of view is displayed, in which 30 sources are active.

The Galactic center observations with WFC also harvest many fast and faint transients. So far discovered 11 of these have been discovered (three in 1999). Some of the 11 cases were only detected during thermonuclear X-ray bursts. The peculiar aspect of these transients is that a large fraction (9) exhibits thermonuclear X-ray bursts, while integrated over all X-ray binaries this fraction is only about 40%. This suggests that a new class of X-ray transient is seen here.

Figure 2 shows the all-sky coverage of WFC observations up to November, 1999. It is strongly non-isotropic because of the dedicated Galactic center program and because of the many observations of the NFI on sources near to the Galactic center (the WFCs are pointed at positions 90o from the NFI axis).


 
Figure 2: All-sky coverage of all WFC observations up to November 1999, in Galactic coordinates. The on-source time is imaged in units of Ms. This excludes occultations by the earth.


The SRON staff involved in WFC-related research in 1999 are John Heise (instrument PI), Remon Cornelisse (also at SIU), Anton Klumper, Erik Kuulkers (also at SIU), Hans Muller (stationed at the BeppoSAX Science Data Center), Jaap Schuurmans, Michael Smith (stationed at the BeppoSAX Science Operations Center), Gerrit Wiersma, and Jean in 't Zand.

So far, more than 100 publications about WFC results have appeared. Half of these are IAU circulars (exemplifying the alarm function of this monitoring instrument). 40% of the publications have SRON investigators as first author. In the following, a selection is made of the five most interesting subjects that involved major research efforts in 1999.

A clocked thermonuclear burster

This is an excellent example of the potential of the long observation program on the Galactic center. Thanks to the large field of view, this program always covers some 30 persistent sources. This implies that for each of them, an exposure of over 3 Ms was obtained. For a narrow-field instrument in a similar earth orbit, this would equal about 5 years of full-time observations! Naturally the sensitivity of the WFCs is moderate but this is not important for bright phenomena such as X-ray bursts. For some bursters as many as 200 bursts were detected which is unprecedented. One case of many bursts from a single source is GS 1826-24. WFC detected over 100 bursts from this source. The interesting aspect of these bursts is that they recur at very regular intervals. Figure 3 shows the distribution of wait times for 70 bursts that were detected in 1996-1998. There is a narrow peak at 5.75 hr. The persistence of such quasi-periodic bursting behavior over years has never been observed in any burster, and the implications for the theory of X-ray bursts are important. The strict periodicity enables one to constrain the constitution of the nuclear fuel and the fusion processes at work.


 
Figure 3: Distribution of wait times between X-ray bursts from GS 1826-24, for all bursts detected with WFC during the first two years of the mission (from Ubertini et al. 1999).


The nearest low-mass X-ray binary and micro-quasar

On February 20, 1999, a one-hour flare was observed from an unexpected position during an observation of the Galactic center field. The peak flux was moderate at 80 mCrab. The source - SAX J1819.3-2525 - was publicized in an IAU circular but initially no optical follow-up was performed. The source was followed up with the NFI and a strong emission line was discovered with an equivalent width of 260 eV which is due to highly ionized iron. In September 1999, the same source was again seen with WFC with similar flares but higher peak fluxes (see figure 4). The extraordinary happened when WFC was not looking: on September 15, the source exhibited a giant flare with a peak flux 12 times that of the Crab nebula which is approximately as bright as the brightest X-ray source known (Sco X-1). The flare was first noticed in the optical and subsequently seen with the All-Sky Monitor on RossiXTE. Again this flare lasted only a matter of hours, but the tail was catched with sensitive measurements with RossiXTE. Follow up studies were done in the optical and radio. The radio image resolved a double-sided jet structure. It turns out that the reason of the large peak flux is simply due to the nearness of the object. The optical and radio observations suggest a distance of about 400 parsecs which is about two times as close as the next nearest low-mass X-ray binary.


 
Figure 4: WFC light curve of SAX J1819.3-2525 during an observation in September 1999. Two flares close to each other are observed, with a peak which is 40 times fainter than that of the giant flare that happened five days later (from In 't Zand et al. 2000).

The orbital period of the a bright recurrent transient in a globular cluster


Twelve bright X-ray sources are known in twelve different Galactic globular clusters. Eleven of these sources are neutron stars in low-mass X-ray binaries. WFC observations before 1999 have helped establish this (see previous year report). The one exception is located in the globular cluster Terzan 6. It is the sole bright globular cluster source from which never thermonuclear bursts were detected. Therefore, the nature of the compact object remains unclear. The X-ray source is transient, it was first detected in 1992 with instruments on Granat and ROSAT. WFC for the first time detected recurrent outbursts. Four outbursts were seen with wait times of order 0.5 yr. The total exposure time when the source was seen to be active amounts to 0.8 Ms (about 25% of all observation time). When the light curve was epoch folded, it was discovered that this source has a periodically changing flux. The source looses most of its flux every 12.36 hr. This was identified as a nearly complete eclipse. This period is equal to the orbital period of the binary where the source is orginated. Seeing total eclipses in low-mass X-ray binaries is a rare opportunity. Only about 5% of all low-mass X-ray binaries do this. The folded light curve is shown in figure 5.


 
Figure 5: Folded WFC light curve of the bright recurrent transient in Terzan 6. The folding period is 12.36 hr. The drop in flux is due to an eclipse of the X-ray source by a companion star which moves in front of it during the orbital motion (from In 't Zand et al. 2000).


Disentangling the gamma-ray bursts phenomenon - beaming effects

GRB research in 1999 was pleasantly surprised by a number of very bright cases that were localized with WFC. Perhaps the most interesting result of X-ray, optical and radio follow-up studies of these was the evidence for beaming in the ejecta. We highlight two GRBs. The most fluent burst was detected on January 23. The optical afterglow was quickly found to be at a distance of z=1.6. Two extraordinary observations were made of this burst. For the first time prompt optical emission was detected from a GRB with a V magnitude as bright as 8.9 by Akerlof of the University of Michigan and collaborators. The prompt optical emission at that distance implies a luminosity equivalent to that of about one million normal galaxies! The fluence of the burst translates in un unprecedented energy output of 3.4X1054 erg. This is more than the rest mass of a neutron star! These extreme numbers, according to Kulkarni of Caltech and collaborators, are very likely due to the wrong assumption that the emission is isotropic. Rather, it must be beamed and we happen to be inside the beam. If so, the true energy output can be as low as 1052 erg, which can more easily be accommodated by current GRB theories. Other interesting aspects of this burst are that the X-ray afterglow is a factor of five brighter than is predicted by current GRB theories, and that for the first time afterglow emission has been detected to photon energies as high as 60 keV (this was 10 keV beforehand).


 
Figure 6: Images of the X-ray afterglow of GRB 990123 at two subsequent time intervals, as taken by the Medium-Energy Concentrator Spectrometer on BeppoSAX. A clear decay is visible.


Another bright burst occurred almost 4 months later, on May 10, 1999. Again evidence was found for beaming of the emission but of a different nature. The light curve of the optical afterglow shows curvature on top of the common power-law decay with increased steepening about 1.5 day after the burst. This has been explained by Harrison from Caltech and collaborators as follows. GRB ejecta start with such high speeds that relativistic beaming is important. Due to this we only see a small portion of the actual emitting surface. As time goes by, the ejecta speeds decrease and we see more and more of this surface. If the ejecta are spherically asymmetric, there comes a time that no extra emission is encompassed if the relativistic beaming decreases, and a steepening of the flux decay is observed. The prediction is that the steepening should be observed at all wavalengths simultaneously. With BeppoSAX a bright X-ray afterglow was observed, and it was found that the X-ray light curve is compatible with such a steepening, though not uniquely.


 
Figure 7: Maximum likelihood ratio map (in detector coordinates) of the BeppoSAX Medium-Energy Concentrator Spectrometer data of the sky region of GRB 990510 during the afterglow phase (from Kuulkers et al. 2000). Contours start at 3 sigma detection level and increment in steps of 2. The dotted circle is the WFC error region, the shaded band the Ulysses-GRBM triangulation annulus. The upward pointed triangle indicates the best-fit position of the afterglow. The other two symbols point to other sources detected in the same data, one is 4 arcmin from the afterglow.


Jean in 't Zand, SRON, January 19, 2000