Transcript L02.Telescopes

Telescopes and
Astronomical Observations
Ay16 Lecture 5
Feb 14, 2008
Outline:
What can we observe?
Telescopes
Optical, IR, Radio, High Energy ++
Limitations
Angular resolution
Spectroscopy
Data Handling
A telescope is an instrument designed
for the observation of remote objects
and the collection of electromagnetic
radiation. "Telescope" (from the Greek
tele = 'far' and skopein = 'to look or see';
teleskopos = 'far-seeing') was a name
invented in 1611 by Prince Frederick
Sesi while watching a presentation of
Galileo Galilei's instrument for viewing
distant objects. "Telescope" can refer to
a whole range of instruments operating
in most regions of the electromagnetic
spectrum.
Telescopes are “Tools”
By themselves, most telescopes
are not scientfically useful. They
need yet other tools a.k.a.
instruments.
What Can We Observe?
Brightness (M)
+ dM/dt = Light Curves, Variability
+ dM/d = Spectrum or SED
+ dM/d/dt = Spectral Variability
Position
+ d(,)/dt = Proper Motion
+ d2(,)/dt2 = Acceleration
Polarization
“Instruments”
• Flux detectors
Photometers / Receivers
• Imagers
Cameras, array detectors
• Spectrographs + Spectrometers
“Spectrophotometer”
Aberrations
•
•
•
•
•
Spherical
Coma
Chromatic
Field Curvature
Astigmatism
Mt. Wilson
& G. E. Hale
60-inch 1906
100-inch 1917
• Edwin Hubble
at the
Palomar
Schmidt
Telescope
circa 1950
Telescope Mirrors
Multiple designs
Solid
Honeycomb
Meniscus
Segmented
Focal Plane Scale
Scale is simply determined by
the effective focal length “fl” of
the telescope.
= 206265”/fl(mm) arcsec/mm
* Focal ratio is the ratio of the
focal legnth to the diameter
Angular Resolution
The resolving power of a telescope (or any
optical system) depends on its size and
on the wavelength at which you are
working. The Rayleigh criterion is
sin () = 1.22  /D
where  is the angular resolution in
Radians
Airy Diffraction Pattern
•
* more complicated as more
optics get added…
Encircled Energy
Another way to look at this is to calculate how
much energy is lost outside an aperture.
For a typical telescope diameter D with a
secondary mirror of diameter d, the excluded
energy is
x( r) ~ [5 r (1- d/D)] -1
where r is in units of  /D radians
 a 20 inch telescope collects 99% of the
light in 14 arcseconds
2 Micron AllSky Survey
•
3 Channel
Camera
Silicon Arrays --- CCDs
CCD Operation
Bucket Brigade
•
•
FAST Spectrograph
• Simple Fiber fed Spectrograph
Hectospec (MMT)
Holmdel Horn
GBT
•
Astronomical Telescopes
& Observations, continued
Lecture 6
The Atmosphere
Space Telescopes
Telescopes of the Future
Astronomical Data Reduction I.
Atmospheric transparency
Hubble
Ground vs Space
•
Adaptive
Optics
Chandra X-Ray Obs
Grazing Incidence X-ray Optics
Total External Reflection
X-Ray Reflection
Snell’s Law
sin11 = sin22
2/1 = 12
sin2 = sin1 /12
Critical angle = sin C = 12
--> total external reflection, not refraction
GLAST
A Compton
telecope
Compton Scattering
LAT
GBM
The Future?
Space
JWST, Constellation X
10-20 m UV?
Ground
LSST, GSMT (GMT,TMT,EELT….)
TMT
TMT
GMT
EELT = OWL
OWL
Optical
Design
JWST
ConX
Chinese Antarctic Astronomy
Astronomical Data
Two Concepts:
1. Signal-to-Noise
2. Noise Sources
Photon Counting
Signal O = photons from the astronomical
object. Usually time dependent.
e.g. Consider a star observed with
a telescope on a single element
detector
O = photon rate / cm2 / s / A
x Area x integration time x bandwidth
= # of photons detected from source
Noise N = unwanted contributions to
counts. From multiple sources
(1) Poisson(shot) noise = sqrt(O)
from Poisson probability distribution
(Assignment: look up
Normal = Gaussan and
Poisson distributions)
Poisson Distribution
•
Normal=Gaussian Distribution
The Bell Curve
Normal = Gaussian
50% of the area is inside +/- 0.67 
68%
“ “
“
+/- 1.00 
90%
“ “
“
+/- 1.69 
95 %
“ “
“
+/- 1.96 
99 %
“ “
“
+/- 2.58 
99.6% “
“
“
+/- 3.00 
of the mean
(2) Background noise from sky +
telescope and possibly other sources
Sky noise is usually calculated from the
sky brightness per unit area (square
arcseconds) also depends on telescope
area, integration time and bandpass
B = Sky counts/solid angle/cm2/s/A
x sky area x area x int time x bandwidth
Detector Noise
(3) Dark counts = D
counts/second/pixel
(time dependent)
(4) Read noise = R
(once per integration so
not time dependent)
So if A = area of telescope in cm2
t = integration time in sec
W = bandwidth in A
O = Object rate (cts/s/cm2/A)
B = Sky (background) rate
D = dark rate
R = read noise
S/N =
OAtW/((O+B)AtW + Dt + R2)1/2
Special Cases
Background limited (B >> D or R)
S/N = O/(O+S)1/2 x (AtW)1/2
Detector limited (R2 >> D or OAtW or BAtW)
S/N = OAtW/R
(e.g. high resolution spectroscopy)
CCD Data
Image data
cts/pixel from object, dark, “bias”
Image Calibration Data
bias frames
flat fields
dark frames (often ignored if detector
good)
Image
Display
Software
SAODS9
Format
.fits
NGC1700 from Keck
Spectra with
LRIS on Keck
Bias
Frame
gives the DC
level of the
readout
amplifier,
also gives the
read noise
estimate.
Flat Field
Image
through filter
on either
twilight sky or
dome
Image Reduction Steps
Combine (average) bias frames
Subtract Bias from all science images
Combine (average) flat field frames filter
by filter, fit smoothed 2-D polynomial,
and divide through so average = 1.000
Divide science images by FF, filter by
filter.
Apply other routines as necessary.
Astronomical Photometry
For example, for photometry you will want
to calibrate each filter (if it was
photometric --- no clouds or fog) by doing
aperture photometry of standard stars to
get the cts/sec for a given flux
Then apply that to aperture photometry of
your unknown stars.
NB. There are often color terms and
atmospheric extinction.
Photometry, con’t
v = -2.5 x log10(vcts/sec) + constant
V = v + C1(B-V) + kVx + C2 ……
x = sec(zenith distance) = airmass
(B-V) = C3(b-v) + C4 + kBVx + ….