OVERVIEW

1
A Search for Radio Transients and Variables in
the Galactic Center
Jennifer L. Neureuther, Scott D. Hyman (Sweet
Briar College), T. Joseph W. Lazio, Michael E.
Nord, Namir E. Kassim (NRL)
OVERVIEW
TRANSIENTS
VLA OBSERVATIONS
We report on a search for variable and
transient radio sources in the Galactic center
using a number of high resolution, wide-field
(2.5) 330 MHz VLA observations made during the
1990's and monthly during Spring and Summer 2002.
Known classes of variable and transient radio
sources include radio counterparts of X-ray
sources and microquasars. The Galactic center
(GC) is a promising region to search as the
stellar densities are high, and neutron- and
black hole binaries appear as X-ray transients
concentrated toward the GC. Two previous radio
transients detected toward the GC, A1742-28
(Davies et al. 1976) and the Galactic Center
Transient (Zhao et al. 1992), had similar radio
properties, but only the former was associated
with an X-ray source. Previous radio surveys
have suffered from confusion in the GC or have
been limited to a single epoch. We have
detected at least one new radio transient, GCRT
J1746-2757, located only 1.1 degrees north of the
Galactic center, and 5 variable candidates. We
discuss the implications of this work for
constraining the Galactic center population of
variables and transients.

• GCRT J1746-2757
• (Hyman et al. 2002, Astron. J., 123, 1497)
• Detected only on 1998 September 25 at 330 MHz.
• Flux Density 216 20 mJy (20 ? detection).
• Unresolved (lt 9) Located 1.1º (150 pc) north
  of
• Sgr A.
• Nonthermal brightness temperature gt 5 x 104 K.
• No X-ray counterpart detected source either
  Doppler
• boosted in radio or a fast X-ray transient or
  a member of a class of radio transients with no
  associated X-ray emission.
• XTE J1748-288
• (Hjellming, Rupen, et al. 1998, Hyman et al.
  2002)
• 330 MHz detection in 1998 September and 1998
  November.
• High frequency monitoring from radio peak in 1998
  June to present.
• Classified as a black hole binary by Fender
  Kuulkers 2001.
• Of the 30 Low-Mass X-ray Binary sources reported
  within our field-of-view, we detect only one
  radio transient counterpart (XTE J1748-288).
  Hjellming et al. find one or two other radio
  counterparts located near the edge of our
  field-of-view where our sensitivity is poor.
• However, we do detect the following source in a
  quiescent state
• GRS 1734-292 (Marti et al. 1998)
• X-ray burst in 1992 September.

Epoch VLA Resolution Duration (hr) rms (mJy/bm)
1989 March 20" x 40 " 5.5 5
1995 August 5" x 10" 1.0 11
1996 October 5" x 10" 5.8 3
1997 February 10" x 20" 1.3 14
1998 March 5" x 10" 5.5 3
1998 September 20" x 40" 6.7 3
2002 March 5" x 10" 1.1 7
2002 April 5" x 10" 1.4 7
2002 May 10" x 20" 1.4 5
2002 June 15" x 60" 0.6 14
2002 July 15" x 40" 1.0 9
Below Detection thresholds for transient
detection as a function of timescale, ?. The
blue line is our actual threshold for
observations of 5 hr duration. The red curve
is our threshold for timescales lt 5 hr. It is
determined by adding simulated transient sources
to the 1998 March observations. We are sensitive
to detecting sources with flux densities in the
region above the red and blue lines. Constant
brightness temperature curves are shown based on
source size upper limits (c?). The temperature
upper limit for an incoherent synchrotron emitter
is 1012 K.
SEARCH METHODOLOGY

• TRANSIENTS
• The 1998 epoch images are used as model images
  to compare to the images at other epochs.
• Run the task UVSUB in AIPS to subtract CLEAN
  components of the model images from each epochs
  uv-data.
• Image the residual data using 512 (55) fields
  for the higher (lower) resolution epochs.
• Search each residual field for bright sources
  and "hole" sources corresponding to possible
  transients.
• The 1989 image below clearly shows a "hole" at
  the location of the 1998 transient GCRT
  J1746-2757. Other sources with constant flux
  density are effectively removed from the field,
  although the source in the northeast corner is
  still visible in the 1989 image due to
  insufficient CLEANing of the 1998 image.

CONCLUSIONS

• VLA observations at 330 MHz of the GC allow for
  efficient transient searching because they
  provide high-resolution (lt20), high dynamic
  range, large field of view (2.5º) images.
• Low-frequency observations exploit the apparent
  steep-spectrum nature of these sources.
• From 330 MHz VLA observations over the past
  decade and Summer 2002, we conclude that radio
  transients above 100 mJy are either very
  infrequent ( one every few years) or have
  timescales much shorter than a month.
• Thus far, we have found only a few radio
  variable candidates out of 250 sources detected,
  but our observations may yet reveal additional
  variables. The task is more difficult than
  detecting transients e.g., a 100 mJy source
  visible in only one epoch is a more reliable
  detection than a 100 mJy variation in a 500 mJy
  source. The latter could be due to a number of
  systematic uncertainties.
• The scarcity of radio transients/variables
  detected in this survey so far underscores the
  need for a far more extensive monitoring program,
  with more sensitive and frequent observations, in
  order to detect, monitor, and identify a large
  number of radio sources.

POSSIBLE VARIABLES
1989
1998
1998
1989

• VARIABLES
• Automated source detection program (AIPS task
  Search and Destroy SAD) used to find sources in
  combination image from the 1996 and 1998 epoch
  observations, produced reference database of 250
  sources (Nord et al. 2003).
• Run SAD on each epoch image. Select measurements
  of sources corresponding to those in the
  reference database. This saves the time of
  "weeding" out false detections by eye for each
  epoch. Also, sources reliably detected on the
  combined 19961998 image lends credence to only
  marginal detections made on individual epoch
  images.
• Confirm all sources whose flux densities differ
  by more than 5? source flux densities confirmed
  by hand. 5 variable candidates detected thus
  far.
• Sources that differ due to unmatched resolutions
  and/or confusion with extended emission are
  removed from the list of variable candidates.

G0.490-1.043 The light curve shows consistent
detections with one notable non-detection in
March 1989 and possible variability in 2002.
G358.638-1.161 The light curve shows a
significant increase in March 1998 and a
non-detection in August 1995. The NVSS detects
the source at 1.4 GHz with flux density 21.1
0.8 mJy.
ACKNOWLEDGEMENTS This research has been helped
considerably by students at Sweet Briar College.
Jennifer Neureuther, Suzanne Bollinger, Ashlee
Bartleson, and Mariana Lazarova have participated
under various SBC student research programs.
S.D.H. thanks Grant Denn for further assistance.
The Very Large Array of the National Radio
Astronomy Observatory is a facility of the
National Science Foundation operated under
cooperative agreement with Associated
Universities, Inc. Basic research in radio
astronomy at NRL is supported by the Office of
Naval Research, and at SBC by Research
Corporation, the Jeffress Memorial Trust, and the
National Science Foundation.
Contact shyman_at_sbc.edu