Understanding Fourier Transforms and Their Role in Interferometry

Introduction

Adam Evans from the University of Manchester gives a concise, hands‑on introduction to Fourier transforms and explains why they are fundamental to radio interferometry.

What is a Fourier Transform?

• A mathematical operation that converts a waveform (function of time or space) into a representation as a sum of sinusoidal components.
• The result is expressed in terms of frequencies (temporal or spatial) and includes amplitude and phase for each component.
• The inverse transform reconstructs the original waveform from its frequency components.

A Food Analogy

Imagine trying to discover the ingredients of a cake without a recipe. Running the cake through a “Fourier kitchen” would separate it into its constituent ingredients and their quantities – this is the Fourier domain. The inverse operation would take those ingredients and bake the original cake again. The analogy highlights two practical benefits: - Easy comparison between different “cakes” (signals) by looking at their ingredient lists. - Verification that the right ingredients are present before attempting to recreate the cake.

The 2‑D Fourier Transform in Astronomy

• In interferometry we are interested in the sky brightness distribution (f(x,y)) on the two‑dimensional plane of the sky.
• The 2‑D Fourier transform converts this spatial image into a complex function (F(u,v)) where u and v are spatial frequencies.
• The real and imaginary parts of (F(u,v)) correspond to measurable amplitudes and phases in an interferometer.

Visualising Fourier Transforms of Everyday Objects

Adam demonstrates the concept by transforming pictures of a dog wearing glasses and a foamy drink: - Amplitude maps reveal where the signal is strongest (large‑scale structures appear as bright central spots). - Phase maps encode the distribution of the emission; subtle ripples indicate finer details. - The examples show that looking at amplitude and phase together is necessary to interpret the original image.

How an Interferometer Works

1. Collect signals with an array of antennas.
2. Correlate the signals in a super‑computer (the correlator) to obtain visibilities – the measured amplitudes and phases.
3. Feed the visibilities into imaging software to reconstruct the sky image.

Incomplete UV Sampling

• Because only a finite number of antennas are available, only a limited set of (u,v) points are sampled.
• The resulting UV coverage plot shows gaps, which leads to artefacts and loss of information in the reconstructed image.

Impact of Antenna Configuration

• Two antennas (a single baseline) sample two points in the UV plane, producing a simple sinusoidal pattern in the image.
• Increasing the baseline length shifts the sampled spatial frequencies to higher values, making the interferometer sensitive to smaller angular scales.
• Adding more antennas creates many baselines, filling the UV plane more densely and improving image fidelity. Adam shows a progression from 3 to 5, 10, and 30 antennas, illustrating how the reconstructed image evolves from a checkerboard of overlapping sinusoids to a recognizable picture.

Earth‑Rotation Synthesis

• As Earth rotates, the projected baselines change, providing new UV points for each time step.
• Even with a modest number of antennas, long‑duration observations dramatically improve UV coverage and thus image quality.

Strategies to Enhance UV Coverage

• Compact core: placing some antennas close together captures large‑scale (low‑frequency) information.
• Extended baselines: spreading antennas far apart accesses fine details (high‑frequency information).
• Combining both approaches yields a well‑sampled UV plane across a wide range of spatial scales.

Summary of Key Points

• Fourier transforms decompose time/space signals into sinusoidal components (amplitude + phase).
• In interferometry the 2‑D Fourier transform links the sky brightness distribution to measured visibilities.
• Sparse UV sampling limits image reconstruction; more antennas and Earth‑rotation synthesis mitigate this.
• Proper antenna layout (short and long baselines) ensures coverage of both large and small spatial frequencies.

Fourier transforms turn complex sky signals into measurable amplitudes and phases; the quality of interferometric images depends on how completely we sample the corresponding spatial frequencies, which is achieved by using many antennas, clever array layouts, and Earth‑rotation synthesis.