Very Long Baseline Interferometry
R. W. Porcas
Very Long Baseline Interferometry (VLBI)
R. W. Porcas
- Very Long Baseline Interferometry
- Longer baselines on Earth (and into Space)
==> higher angular resolution for finer structural details and higher positional accuracy
- Traditionally realized without realtime links
- Independent local oscillators
- Local sampling and recording of data on tapes, disks
- Interferometry achieved by off-line replay and correlation of the recorded data streams at correlator centre (JIVE, Bonn, VLBA-Socorro)
- Modern developments with realtime transmission of the data streams and realtime correlation
- Longer baselines on Earth (and into Space)
- Principles of VLBI exactly the same as for shorter baselines;
- in reality there are practical differences
- Differences due to longer baselines
- Often cross international borders
==> naturally leads to international cooperation !
- Higher angular resolution
- Most calibration sources resolved, so need a priori calibration, using telescope gains and system noise measurements
- More missing flux due to large holes in uv plane - sources must be even more compact to be detected
- Delay and fringe-rate beams are narrower
==> Reduction of the field-of-view
- Need more accurate source positions
- Need more accurate Earth geometry (UT1-UTC, motion of pole)
- Need shorter visibility sampling times, and finer frequency resolution
==> Fringe fitting (search for source response in delay/rate space)
- leading to a detection threshold
- Often cross international borders
- Differences due to the spherical Earth
- Source has different elevations at different antennas
- Restriction of the amount of time for mutual visibility
- (can't see Galactic Centre on Europe-USA baseline !)
- For a VLBI array, relatively more low uv-coverage than high
- Non-equal path lengths through the troposphere and ionosphere leads to non-cancellation of their contributions to the signal delay and phase
==> need atmospheric and ionospheric model for phase-referencing
- Restriction of the amount of time for mutual visibility
- Source has different elevations at different antennas
- Use of independent local oscillators
- Need for high temporal coherence local oscillator
==> Hydrogen Maser frequency standard (stability ~10**14)
- Need for high temporal coherence local oscillator
- Use of independent clocks (to label the sampled data)
- Need to have accurate monitoring of clocks (with GPS)
- Residual clock (and clock rate) uncertainty needs to be determined using fringe-fitting
- Use of unconnected elements
- Can make "ad hoc" VLBI arrays with an arbitrary set of antennas with compatible observing and recording equipment. This leads to arrays (e.g. the EVN) where the individual telescopes have very different properties - range of sensitivities and range of primary beamwidths
==> visibilities have baseline-dependent errors, which must be considered when imaging
- Recording data naturally leads to digitization which historically resulted in multiple, relatively-narrow, data channels
==> need to calibrate relative phase between channels with "phase-cal" signals
- Can make "ad hoc" VLBI arrays with an arbitrary set of antennas with compatible observing and recording equipment. This leads to arrays (e.g. the EVN) where the individual telescopes have very different properties - range of sensitivities and range of primary beamwidths
- Use of "Off-line" rather than "realtime" correlation
- Correlation need not be made in strict observing sequence
Correlation may be done in "multiple passes" (e.g. if number of elements is > number of playback units, or if multiple field centres are needed)
==> Data from correlator may not be in time-baseline order
VLBI-specific AIPS analysis tasks
- MSORT - sort data if correlator output not in time/baseline order (ANTAB) - Writes TY (system temperature) and GC (gain-curve) tables if amplitude calibration data supplied separately from the visibilty data APCAL - extracts amplitude calibration from TY and GC tables and writes an SN table (PCLOD) - Writes PC (phase-cal) table if phase-cal data is supplied separately from the visibilty data PCCOR - extract phase-cal information to an SN table ACCOR - determine 2-bit sampling amplitude corrections for each antenna and write to an SN table CLCAL - uses data from SN tables to update total calibration in a CL table FRING - self-calibration of (antenna) residual phase, delay and fringe-rates arising from errors in interferometry geometry, station clocks, tropospheric and ionospheric propagation CALIB - self-calibration of "antenna" phases
