Radio interferometer — 2D uv-plane and imaging lab

Switch to 3D
Controls
Live readouts
Numerical checks
What to try
  1. Press "East-West line" then "Equatorial source": at low declination Earth rotation only draws a thin line in the uv plane and the image smears into a ridge.
  2. Move declination towards 90 deg: uv tracks become circles, the beam round.
  3. Compare "Two dishes" (a single stripe pattern) with the Y array (a rich, nearly filled uv plane).
  4. Halve the observing time: the uv arcs get shorter and the dirty beam gains sidelobes.
The science

An interferometer correlates the signals of a pair of dishes. For a baseline vector b and a source direction s the correlator measures one Fourier component of the sky brightness I(l,m) at the spatial frequency (u,v) = bproj/λ:

V(u,v) = ∫∫ I(l,m) e−2πi(ul + vm) dl dm u = (X sin H + Y cos H)/λ   v = (−X sinδ cos H + Y sinδ sin H + Z cosδ)/λ Dirty image = I ∗ Bdirty,   Bdirty(l,m) = (1/N) Σ cos[2π(uk l + vk m)],   resolution ≈ λ / Bmax

(X,Y,Z) is the baseline in the equatorial frame; X = −N sinφ + U cosφ, Y = E, Z = N cosφ + U sinφ for latitude φ. As the Earth turns, the hour angle H sweeps each baseline through an ellipse, filling the uv plane without moving any dish. The page samples every 5 minutes of sky time, adds the conjugate (−u,−v) points, computes the true visibilities of the chosen sky model and accumulates the dirty image and beam exactly.

Checks: constant projected baseline for a polar source, the elliptical uv track of an east–west baseline, recovery of a point source position and flux, and the uv aspect ratio sinδ.

Simplifications: no deconvolution (CLEAN), no primary beam, noise, delay tracking, bandwidth smearing or w-term; a small flat field of view; coordinates are in units of λ/Bmax so the same picture scales to any wavelength and array size.