Low Frequency VLBI

1
Low Frequency VLBI

• Astrometry through beer goggles
• Adam Deller
• Swinburne University

2
Outline

• What you get (resolution, FOV, sensitivity)
• Previous low frequency VLBI
• What you can do with low frequency VLBI
• Localise transients
• ISM studies
• Nearby astrometry
• Why its hard (ionosphere) and what can be done
  to counter this

3
Disclaimer

• Im not a low-frequency expert!
• This is (mainly) not VLBI in the current,
  disconnected/narrow field sense
• Simply (connected) long baselines
• Focus on E-LOFAR since this is the only truly
  long-baseline instrument (LWA and VLAVLBA shown
  for comparison)

4
Sensitivity

• LWA 1 mJy in 1 hour at 20 MHz
• E-LOFAR 0.1 mJy in 1 hour at 200 MHz
• VLBA phased VLA 10 mJy in 12 hours at 74 MHz
• By way of comparison EVLA, E-MERLIN, EVN,
  upgraded VLBA all heading towards micro-Jy
  sensitivity in single pointings
• PSR 193721 in M31 several mJy _at_30 MHz!! c.f.
  lt1?Jy brightest pulsars _at_2 GHz

5
Resolution

• LWA (400km baselines) gt2
• VLAVLBA (8000 km baselines) gt100 mas
• E-LOFAR (1000km baselines) gt200 mas
• Cf. Existing low frequency surveys VLSS (74 MHz)
  80, WENSS (327 MHz) 54
• Cf. other widefield surveys FIRST 5, NVSS 45,
  ATLAS 6
• Optical surveys 1

6
Station FOV

• LWA 2.5-10 FWHM (80-20 MHz) x8
• VLA VLBA Arcminute(s) (phased VLA) x1
• E-LOFAR 1-10 FWHM (300-30 MHz) x8
• Full FOV unlikely to be available to E-LOFAR due
  to time/bandwidth smearing

7
Reasons for low-freq VLBI

• The love of the game very little is known
  about the low-frequency sky at high resolution
• E.g. accurate fluxes for compact sources
• Solar observations (LOFAR)
• High-Z AGN/starburst discriminant
• Studying the ISM via pulsars
• Interferometric scintillation measurements
• Angular broadening
• Astrometry (?!?!) Transients, pulsar kinematics

8
Previous low-freq VLBI

• Lenc et al. 2008 looked for VLBI counterparts to
  NVSS/WENSS sources using the VLBA some EVN at
  327 MHz, piggyback on targeted observation
• Detected 27 WENSS sources (expect 270 if all
  unresolved) from 2x13 sq. deg. fields
• Sensitivity several mJy/beam

9
Previous low-freq VLBI
10
Previous low-freq VLBI

• The Lenc et al. imaging took months!!!
• All other previous low-freq VLBI surveys combined
  yield a similar number of published images
  (Altschuler et al. 1995 Lazio Cordes 1998
  Chuprikov et al. 1999 Cai et al. 2002)
• Thus both the techniques and understanding of the
  source population (starburst/AGN/Galactic) are
  fledgling

11
ISM studies with VLBI

• Angular broadening for turbulence measurement
  imaging pulsars
• Interferometric measurements of scintillation
  Brisken et al. (Arecibo, GBT, phased Westerbork,
  Jodrell Bank)
• Measurement of the transfer function of the ISM
  for coherent de-scattering?

12
Astrometry

• Will never be as good as higher frequency
  observations - obviously!!
• Single epoch accuracies mas at best?
• Refractive scintillation may be short-term limit?
• Three reasons why this is still absolutely
  crucial for low frequency instruments
• Often we just want arcseconds eg transients
• Many sources are nearby, large motions
• If this is the only possible ?, we must!

13
Astrometry

• Spatial localisation of transients
• LOFAR transient follow up TBB vs lower freq and
  wait for dispersed signal to arrive?
• Faint steep spectrum objects
• Kinematics of pulsars, other neutron stars
• Might be tough to do parallax unless very nearby
  but proper motion should be ok
• Because of wide fields we may be able to observe
  many times/year c.f. currently

14
Extrapolating from 1600 MHz

• 7 observations with the LBA (1400 km baseline, 5
  antennas) phase referenced, 1 out-of-beam cal,
  primitive ionosphere corr.
• 40 mJy pulsar, Thermal SNR errors 0.25 mas,
  systematic iono errors 0.3 mas
• Scale by 10 and 102 for resolution and
  ionosphere to 140 MHz 2.5,30 mas, but
  sensitivity/calibration will be much better!

15
Challenges for low-freq VLBI

• Calibration, calibration, calibration
• Ionosphere is biggest problem, but most VLBI
  sources are weak so dynamic range/confusion is
  also a big deal
• Compared to shorter baselines we have less
  calibrators and bigger fluctuations
• Multiple calibrators and good a priori ionosphere
  (field calibration, Lazio) crucial

16
Multi-calibrator approach

• The LOFAR ionosphere model must be refined
  further within the long-baseline FOV to get
  accurate astrometry
• With more than one calibrator, can refine
  calibration across the beam with increasing
  accuracy
• Bright calibrators rapid corrections
• Can ?2 selfcal on target (Brisken)

17
Multi-calibrator approach

• Expect 3 good calibrators per square degree (gt10
  mJy flux, lt200 mas in extent, Lenc et al. 2008)
• What FOV will E-LOFAR have - will 3 calibrators
  per degree be enough?
• Maybe lower frequency will be better
  astrometrically!!! (More calibrators, brighter)

18
Conclusions

• VLBI is possible at low frequencies!
• VLBI is useful at low frequencies!
• Pulsars are one of the prime targets for low-freq
  VLBI, both as targets and to illuminate the ISM
• Integrated VLBI will dominate the science
• Maximising FOV and sensitivity is crucial in
  order to get the best possible calibration