Transcript the CMB

Interferometry



Jack Replinger
Observational Cosmology Lab
Professor Peter Timbie
University of Wisconsin-Madison
Interferometry uses an array of small dishes to gain the resolution of a
much larger single dish
The purpose of this tutorial is to give a viewer a basic understanding of
the physics of interferometry
We proceed by discussing interference and diffraction.
picture courtesy of http://aries.phys.yorku.ca/~bartel/GPBmovie/Vla.jpg
Two Slit Interference
Interference pattern
λ/2d
constructive
interference
destructive
interference
light sources
Two Slit Interference
Reversed
“Interference pattern”
λ/2d
detected
in phase
detected
out of phase
The detectors are
therefore sensitive to the
interference pattern
detectors
Adding Interferometer
signal emitted from source
reaches right antenna δt
sooner than left antenna
signal detected by right
antenna delayed by δt
such that at the tee, two
corresponding signals are
interfered
diagram courtesy of http://www.geocities.com/CapeCanaveral/2309/page3.html
Antenna Optics


Antenna is designed such that parallel
rays converge at focus
Use reversibility: Analogous to single
slit diffraction
Diffraction Reversed
diffraction
pattern
detectors
sources
λ/D
undetectable
total
destructive
interference
maximum
response
side lobe
The single antenna is
therefore sensitive to the
diffraction pattern
light source
antenna
Two Slit Diffraction



Envelope due to
antenna sensitivity
(Diffraction)
Peaks due to baseline
(Interference)
Angular Resolution
(Rayleigh Criterion)
is λ/2d from baseline,
instead of λ/D from
diffraction limit
Image courtesy of http://img.sparknotes.com/figures/C/c33e2bffc162212e1d9aa769ad3ae54f/envelope.gif
Image courtesy of http://www.ece.utexas.edu/~becker/diffract.pdf
Interferometer Sensitivity

Interferometer is like
diffraction in reverse, in 2D



Each antenna is analogous
to a circular aperture
Example at left is the
Diffraction pattern from two
circular apertures (shown at
upper left)
The interferometer is
sensitive to projection of the
diffraction pattern on the
sky


Sources in light regions are
detected, signal strength
varies with intensity
Sources in dark regions are
undetectable
contributes weak signal
contributes strong signal
contributes no signal
courtesy of http://www.ee.surrey.ac.uk/Personal/D.Jefferies/aperture.html
Fourier Analysis:
Background
r

In order to understand how to reconstruct
an image it is important to understand the
mathematics of diffraction and interference
θ
0θ
x
dx
a-
P

Notation:
 x (and y) describe the plane of the aperture
 θ (and Ф) describe the “plane” of the
diffraction pattern
Equation describing the E-field from a point source of the wave at any distance d
E = Eoei(kd-wt)
Fourier Analysis:
Single Slit Diffraction
r
θ
0θ
x
dx
a-
at P, the electric field due to a small dx is
at P, the electric field is
dE = Eoei(k(r+xsinθ)-wt)dx
E(θ) = ∫oa Eoei(k(r+xsinθ)-wt)dx
generally, for any aperture
= Eo ∫aperture eikxsinθdx
let A(x) describe the aperture
= Eo ∫all space eikxsinθA(x)dx
Generalizing this to two dimensions, and any aperture, the E field at P is
E(θ,Ф) = Eo ∫all space eikxp+ikyq A(x,y)dxdy
p and q are functions of θ and Ф, describing the phase shifts
P
Fourier Analysis:
Generalized Diffraction
P
r
detector
θ
0-
sky
θ
x
dx
a-
The intensity (proportional to E2) on the sky is therefore
I(θ,Ф) = C*(E(θ,Ф))2
= C*(∫all space eikxp+ikyq A(x,y)dxdy)2
If the aperture is the detector, then the diffraction pattern describes the sensitivity of the
instrument
G(θ,Ф) = I(θ,Ф)
The power P recorded by the detector is the product of the sensitivity function and the intensity of
the sky S(θ,Ф) integrated over the sky
P = ∫sky G(θ,Ф)*S(θ,Ф )dθdФ
= C*∫sky(∫all spaceeikxp+ikyqA(x,y)dxdy)2*S(θ,Ф)dθdФ
Fourier Analysis:
P
Image Reconstruction
r
detector
θ
0-
sky
θ
x
dx
a-
P = C*∫sky(∫all spaceeikxp+ikyq A(x,y)dxdy)2*S(θ,Ф) dθdФ
P is recorded each time the detector is pointed
A(x) is determined by the aperture, p and q are known functions that describe the phase
shifts
By recording P and varying A, by changing the baselines or using multiple baselines if
there are enough detectors, we obtain enough information to solve for S(θ,Ф) for that
patch of the sky, fortunately this can be done on computers with existing software. For
MBI the program will be written by Siddharth Malu.
It is important to note that this is an oversimplification of the situation, when applying this
is a diffuse source the coherence of the light from different regions of the source must be
addressed with a coherence function. This is addressed by the Van Cittert-Zernike
theorem, which is beyond the scope of this presentation.