Why Spectroscopy?

Download Report

Transcript Why Spectroscopy?

AAVSO Citizen Sky
Spectroscopy
Workshop
07August 2009
Presented
by
Hopkins Phoenix Observatory
Background
.
HPO Research
Research at the Hopkins Phoenix Observatory
UBV Photon Counting Photometry
(Since 1980)
.
High Resolution Spectroscopy
(Since 2008)
HPO-1
.
HPO-1 is a two-story observatory with
an 8” Celestron C-8 telescope.
HPO-1 Photometry
HPO UBV PMT
based photon
counting
photometer.
.
HPO-2
HPO-2 is a
single story
roll-off roof
observatory
with a 12”
LX200 GPS
telescope.
.
HPO Spectroscopy
High resolution
Lhires III
spectrograph
.
Lhires III @ HPO
Lhires III
mounted on
a
12” LX200
GPS.
DSI Pro I
for Guiding
(Black)
.
DSI Pro II
for Imaging
(Blue)
Configuration
.
Introduction
Just a few years ago those who did astronomy
and had a background in physics could only dream
of doing astronomical spectroscopy.
Even with a physics background amateur
astronomers felt spectroscopy was well beyond
reach.
In addition to the extreme cost, it was thought
. large telescopes were needed even on fairly
very
bright stars.
The Start
All that started to change when CCD cameras became
available. It started slow. Cookbook CCD cameras
were among the first for amateurs. Then came web
cams, modified web cams and low cost astronomical
CCD cameras such as the Meade DSI series.
The advantage of these devices was that they were
very sensitive and ideal for astronomical applications.
While most amateur astronomers used these for
.
imaging
pretty pictures, some saw the potential for
more serious work. Photometry was one of the first
uses. Spectroscopy is just now gaining momentum.
Europeans Lead
Just a couple of years ago some serious spectrographs
for amateurs and smaller observatories came on the
market. These are expensive, but not prohibitive for
many amateurs. Coupled with the sensitive CCD
cameras, all at once the field of spectroscopy was
available to the advanced amateur.
User groups started, first in Europe and are just now
expanding
in the USA and other countries. For the first
.
time amateurs astronomers can contribute
professional astronomical spectroscopic data.
Why
Spectroscopy?
While photometry measures the brightness of
stars in definite fixed bandwidths, spectrometry
measures the whole visible spectrum of a star in
fine detail.
.
Photometry
takes the pulse of a star whereas
spectroscopy analyses the soul of the system.
Producing A Spectrum
There are basically two ways to produce a
spectrum.
1.Using a Glass Prism.
.
2.Using a Diffraction Grating
(both transmission and reflective
types)
Prisms
Prisms, triangular pieces of glass, were first used
to produce a spectrum.
.
Refracting Light
A sliver of light enters the prism and is
delayed (refracted) at different wavelengths
causing the light to emerge at different
positions producing a rainbow of colors.
.
Diffraction Grating
A reflective diffraction grating is a reflective surface
with closely spaced groves and will produce a spectrum
similar to that of a prism. High quality gratings can
produce much finer resolution than a prism.
.
Lhires III removable grating.
Physics
Review
Where Do Photons
Come From?
Atoms consist of a nucleus surrounded by electrons.
The electrons are in specific energy states or levels.
If an electron is raised to a higher energy state it
will soon fall back to its lower state and emit a
photon of energy equal to the difference in the two
energy states. This is the most common method of
photon creation
Note: Photons can also be created by other means.
Absorbing Energy
A photon
interacts with
an atom’s orbital
electron and
raises it to a
higher energy
state.
The electron
absorbs
the
.
photon’s energy.
Emitting Energy
After a short time
the electron falls
back to its lower
energy state
emitting a photon
with the energy of
the difference
between the two
.
energy
states
Energy and Frequency
E=h*c/l
or
l= h * c / E
.
Remember
F=1/l
Where:
E = Photon Energy
h = Planck’s Constant
c = Speed of Light
l = Wavelength
F = Frequency
h = 6.62606896 x 10-34 Js
c = 3 X 108 m/s
Element Spectra
Each element has a unique set of precise energy
states that relate to specific spectral Lines.
Like a fingerprint.
By identifying a set of lines, the element that
emitted or absorbed that energy can be
identified.
Elements
Hydrogen Spectrum
Helium Spectrum
Neon Spectrum
.
Sodium Spectrum
Mercury Spectrum
Doppler Shift
In 1842 Christian A. Doppler discovered an effect
that produces a shift in frequency of sound
dependent on the motion of the source.
A train coming toward you has a higher frequency
whistle than when the train is not moving or as the
train passes. The departing train’s whistle is a lower
frequency than when stationary.
.
This is known as the Doppler Effect or Shift and
applies to light as well as sound.
Radial Velocity
It is important to understand that Doppler Shift can
provide a measure of radial velocity. That is motion
directly to or from the observer.
Tangential velocity will not change the wavelength
and there will
be no Doppler Shift.
.
Doppler Equation
Vr = Dl
*
c/l
Dl is the change in wavelength due to radial motion
l is the stationary wavelength
Vr is the relative (radial) velocity
c is the velocity
in the medium
.
8
(speed of light in a vacuum is 3 X 10 m/s)
To get just a 1% change in the frequency of light, a star has to be
moving 1,864 miles per second. For a blue light bulb to look red, it
would have to be flying away from you at 3/4 of the speed of light.
Why is Doppler Shift
Important?
When we know the wavelength shift we
can determine the radial velocity of the
gas that emitted or absorbed the energy
at that wavelength.
This tells us how fast the gas is coming
toward . or going away from us.
Orbital velocities of binary stars can be
determined with this among other things.
Types of Spectra
Continuous Spectrum
Absorption Spectrum
.
Emission Spectrum
Stellar Spectrum
Solar Spectrum
.
Spectrometers
You may here the terms spectrometer,
spectroscope and spectrograph.
What are the differences?
The terms can be used interchangeably,
but for the purist there are slight
.
differences.
Definitions
Spectrometer
Usually is a prim based device.
Spectroscope
Prism or diffraction grating device, used visually.
Spectrograph
.Diffraction grating device used with the a
detector, e.g., CCD or film cameras.
Question!
When used visually the device is called a
spectroscope or telescope.
When a detector is connected to the
spectroscope it becomes a spectrograph.
.
Does that mean when a detector is connected
to a telescope it becomes a telegraph?
Why Hydrogen Alpha?
When there is discussion of spectroscopy of stars
you will here hydrogen alpha line mentioned a lot.
Since stars are mainly hydrogen and since the
hydrogen alpha line is the most prominent line for
the element, it gets lots of study.
For epsilon Aurigae other hydrogen lines are also
of . interest including the beta and gamma lines.
The sodium D lines as well as the Potassium Ki
lines are also of great interest.
Low Resolution
Spectroscopy
.
Star Analyser
Low resolution spectroscopy
For under $200.00
.
Holder
Star Analyser
Star Analyser
Mounted
.
Star Analyser ready for
mounting on a telephoto lens
Star Analyser
Spectroscopy
DSLR Camera
.
Telephoto Lens
Mounted
Star Analyser
Experiment
Before going outside into the night, experiment
inside your house to get familiar with the
equipment and develop an initial technique that
can be refined later.
.
Test Setup
.
Use the above setup to produce a continuous
spectrum for experimentation
Test Spectrum
.
Use the test spectrum to experiment with the
processing software.
Star Analyser
Imaging
Steps:
1.Find a bright star or object and orient the
Star Analyser so that the spectrum is
horizontal. Note the source should be to the
left of the spectrum.
2.Set the camera for maximum aperture, ISO
. 1600 or maximum sensitivity and for a 30
second exposure.
Vega Spectrum
Raw Spectrum
.
Annotated Spectrum
Alpha Lyrae Spectrum
Raw Spectrum
.
Robin Leadbeater
Robin Leadbeater in
Cubria England has
perfected a means
to use the Star
Analyser with a
DSLR camera.
.
Low Resolution
Epsilon Aurigae Spectrum
.
Robin Leadbeater’s Star Analyser spectrum
Hydrogen beta absorption line can be seen at 4,861Å
Note: This is an out-of-eclipse spectrum.
High Resolution
Spectroscopy
.
Lhires III
Spectroscopy
While a Star Analyser can produce a low
resolution spectrum showing the whole
spectrum, to see details of the spectrum
and make precise measurements, a higher
resolution grating is needed.
.
The Lhires III is an excellent spectrograph
that can use gratings of 150, 300, 600,
1,200 and 2,400 (standard) lines/mm.
Lhires III
.
Grating
Assembly
Lhires – Littrow High Resolution Spectrograph
Telescopes & Lhires III
The Lhires III can be used with telescopes F/8
to F/12 and is optimized for F/10.
It works well with telescopes ranging from 8”
to 16”.
For high resolution work on epsilon Aurigae
. with a 12” LX200 telescope, 8 minute
exposures produce excellent Ha spectra.
Lhires III Design
.
Lhires III Light Path
.
Lhires III Specs
Grating – Lhires III (lines/mm)
2400
1200
600
300
150
Dispersion (H ) nm/pix
0.012
0.035
0.074 0.149 0.300
Resolving power
17000
5900
2800
1400 700
Radial Velocity Km/s
5
17
35
75
150
Field of view nm
8.5
25
55
110
230
All visual domain in #images 45
15
7
4
2
Limiting magnitude
5.0
6.8
7.5
8.4
9.2
.
1.0 hour exposure
200mm (8”) f/10 telescope, 30µm slit,
KAF0400 camera, Signal/Noise of 100)
Limiting Magnitude
2400 l/mm Grating
Star type B0V - CCD KAF-0400
Slit Width: 25 µm - Seeing : 4 arcsec.
Resolving Power (R) : 17000
Sampling (KAF-0400) : 0.115 A/pixel
Telescope
D=128 mm (5”) F/D=8
Limiting Magnitude
S/N=50 in 1 hr
6.5
Limiting Magnitude
S/N=100 in 1 hr
5.6
D=200 mm (8”) F/D=10
6.7
5.9
D=280 mm (11”) F/D=10
7.1
6.2
. mm (14”) F/D=11
D=355
7.2
6.3
D=600 mm (24”) F/D=8
8.1
7.2
Note: Lower resolution gratings will allow fainter limiting magnitudes
Spectrograph Slit
To produce a good spectrum, light must pass
through a slit.
While we found the slit in the Lhires III was
fine as received,
if there is any doubt about
.
the slit width or parallelism, it should be
examined and adjusted.
Slit
.
The Lhires III has an easy removable slit. Additional
slits are available in case you need to frequently
change slit width.
Slit Measurement
Measure the distance “x” projected on your screen.
You can then determine the slit width “a” with the
.
formula:
a=D*λ/x*l
l is the laser wavelength:
Red laser is 655 or 671 nm
Green laser is 532 nm.
Slit Numbers
Here are some values for a laser at 650nm
Slit
width
Sample
Distance x
Note that
. the slit adjustment is made manually.
To adjust, loosen the screws, move the half-slit
smoothly, and tighten back. Be careful to keep
the slit parallel. This is easy to control visually,
by looking at light through the slit.
Explore/Experiment
When you first receive your Lhires III it is
suggested you spend several hours during the
day on the bench without a telescope and get
familiar with the unit. Otherwise you are most
likely in for some frustration and wasted time.
. neon calibrator provides a simple
The built-in
means of experimenting.
Confusing Spectrum
The spectrum you will see will be very confusing.
This is particularly true when using a high
resolution grating. Even using the neon spectrum
will be a challenge to positively identify lines.
The spectrum window will be narrow so you see
only a small portion of the visual spectrum.
.
At HPO using a DSI Pro II and 2,400 lines/mm
grating, the window is only about 90 Å wide. The
visible spectrum is over 3,000 Å wide.
Micrometer Calibration
One of the first things you should do is get a rough
calibration of the micrometer. The micrometer is
used to adjust the angle of the diffraction grating
to allow different portions of the spectrum to be
viewed.
.
Note: The micrometer has significant backlash.
Repeatability is approximate and not precise.
Micrometer Setting
Dr. Bob’s Calibration
.
l in Å
Line Identification
As noted earlier even with the neon calibrator,
positively identifying lines is a challenge.
A red laser pointer will produce a single line
around 6,550 Å (some red laser pointers are
6,710 Å) and once found is a good starting point.
A green laser pointer is 5,320 Å.
.
A pickle light is an excellent means of identifying
where the sodium D lines are.
Neon Calibration
The Lhires III has a built-in neon calibrator.
This is convenient for not only calibrating
spectra and the micrometer, but, as mentioned
earlier, for on the bench experimenting and
focusing of the main spectrograph optics.
.
Plan on 1.0 second exposures for the neon
lines, but be sure the peak ADU counts stay
well under 32,000.
Neon Calibrator
.
A built-in
Neon
calibrator
that can be
operated off
2-9V
batteries
(18V) is
provided for
wavelength
calibration.
Neon Spectrum
Ne
6,533 Å
Ne
6,599 Å
Ha
6,563 Å
Region
Neon lines bracketing the Ha area (1 second exposure)
Neon Line Profile
These Ne lines bracket the
. hydrogen alpha wavelength.
Ha
6,562 Å
Region
Neon Lines
.
Laser Pointer
.
Shine the laser pointer at the telescope plastic dust
cover (to reduce the intensity) on the spectrograph.
Laser Pen Line
.
Laser Pen Line is at 6,550 Å
(just one line – 1.0 second exposure)
Laser & Neon Lines
Ne
6,533 Å
Ne
6,599 Å
Laser
6,550 Å
.
Ha
6,563 Å
Region
Neon and laser lines bracketing the Ha area
(1 second exposure)
Pickle Light
.
Sodium D Lines 5,889.95 Å & 5,895.92 Å
Light Leak
.
There is significant light leak in the Lhires
particularly around the grating unit. Half inch
strips of metalized duct tape make a good seal.
Focusing Problem
.
Focus of the Lhires III changes considerable with just a few degrees
change in ambient temperature. Always check the Neon lines for sharp
focusing just before imaging the program star.
Out-of-Focus
.
Histogram peak is 25,781 ADU counts (1.0 Sec)
Focused
.
Histogram peak is 27,764 ADU counts (1.0 Second)
Hydrogen Alpha Spectra
(Line Profiles)
Altair Spectrum
Deneb Spectrum
.
Beta Lyrae Spectrum
P Cygni Spectrum
Taken with a Lhires III and 2,400 l/mm grating
Imaging Techniques
Star above the slit
+
Because the slit cannot be
seen the computer “+”
cursor is placed over the
star & slit for easy reference
+
Star on the slit
Cursor over the slit
Exposure
Once on the star is on the slit with the imaging
camera set for 1.0 second exposure, a faint
spectrum should be seen.
At this time an exposure can be started. Set the
exposure time for 8 minutes and start.
Essential all the techniques for imaging deep
sky objects are used here.
Dark and Flat frames for example.
Image
Processing
Imaging Processing
Software
Once the spectrum has been imaged, the fun
starts. There are two major programs used for
processing the spectral images, Iris and VSpec
(Visual Spec). Also SpcAudace is available, but
I have no experience with it.
These are freeware and very powerful. They
are a challenge to master, however.
The goal is to create a calibrated line profile of
the spectrum so that characteristics can be
measured.
IRIS
Iris has many features, but at HPO, and many
other places, we use it mainly for two
preprocessing steps.
1.Subtracting the sky.
2.Optimizing the spectrum.
Note: Iris uses signed 16 bit .fits images.
This is why it is important to make sure
the peak image pixel ADU counts are
below 32,000.
VSpec
Once the spectrum has been processed in Iris
it is imported into VSpec. Here a line profile is
created and wavelength calibration done.
With a calibrated line profile, Doppler shifts
and Equivalent Widths can be determined
along with other characteristics.
Line Profiles
One of the nice things about CCDs is that they
make creating the line profile very easy.
All that needs to be done is to sum each column
of pixel ADU values.
That means even though a pixel may only have
25,000 ADU counts, when summed with others of
that column the total can easily be in the
hundreds of thousands ADU counts.
Spectrum/Line Profile
Spectrum
Continuum Level
Pixel
ADU
Counts
Absorption Lines
Wavelength
Line Profile
Wavelength Calibration
To be useful the line profile must be calibrated
for the wavelength.
While the Ne lines can be used for a good
calibration, using the multiple atmospheric lines
(if seen) can produce amore accurate calibration.
Atmospheric Template
.
Atmospheric Calibration
.
While the neon
spectral allows
a good
wavelength
calibration
using
atmospheric
lines can
calibrate the
spectrum more
accurately.
What is
Heliocentric
.
Earth’s Orbital Motion
.
Depending where the
star is and where the
Earth is in its orbit
around the Sun, the
radial velocity of the
Earth toward or away
from the star can vary
between + 67,000 MPH
going toward the star to
– 67,000 MPH going
away from the star.
Earth’s Rotation
.
Depending on the latitude of the observer the
Earth’s rotation can contribute up to 1,024 MPH
to the radial velocity.
Heliocentric Calibration
Once the profile
is wavelength
calibrated, a
Heliocentric
correction to
remove the
.
Earth’s motion
must be made.
Some Important
Wavelengths
KI
Na D1
7,699 Å
5,896 Å
Ha
Hb
6,563 Å
4,861 Å
Na D2
5,890 Å
Hg
4,341 Å
OI
7,772 Å
Hd
4,102 Å
He
3,970 Å
.
Red Laser
6,550 Å
6,571 Å
Green Laser 5,320 Å
Wavelength Calibration
Tutorial
A detailed tutorial for calibrating a spectrum
around the hydrogen alpha region using
Vspec can be found on the web site:
http://www.hposoft.com/HaCalibration.pdf
.
Analysis
There are several ways to provide a numerical
analysis of a spectral lines.
1.Doppler shifts (for radial velocities).
2.Equivalent
Widths (EW) of lines.
.
3.V/R (Violet or blue to Red EW) line ratios.
Equivalent Width
Continuum
Using a line .profile's Equivalent Widths allows an
expression of the part's significance or strength. The
area under the curve between the profile part and the
continuum is the EW of that part. The area is equal to
the Intensity (normalized to 1.0 for the continuum)
times EW in angstrom (Å).
Out-of-Eclipse
Spectroscopy
of
Epsilon Aurigae
.
Star System
.
Sodium D Lines
.
Hydrogen Alpha
.
Hydrogen Alpha
Analysis
.
Summary Table
UT
Date
2008
08/11
08/22
09/03
09/05
09/22
09/29
10/12
10/14
10/15
10/19
10/19
10/21
10/21
10/24
10/26
10/28
10/30
11/01
Emissive
Blue Horn
Center l
EW
Å
0.424
6,561.40
0.273
6,561.52
9999.999
0.292
6,561.33
0.342
6,561.51
0.265
6,561.12
0.163
6,560.71
0.225
6,560.52
0.378
6,560.62
0.343
6,561.28
0.256
6,561.66
0.268
6,561.32
0.342
6,561.32
0.262
6,561.50
0.396
6,560.14
0.341
6,561.41
0.359
6,561.30
0.305
6,561.31
0.136
6,561.50
.
Absorption
EW
-1.009
-1.056
Center l
Å
6,563.11
6,563.10
-0.904
-0.887
-0.993
-1.327
-1.003
-1.127
-1.002
-1.088
-1.070
-1.011
-1.015
-0.881
-1.051
-1.025
-0.992
-1.046
6,563.11
6,563.15
6,562.85
6,565.50
6,562.14
6,561.98
6,562.98
6,563.41
6,562.95
6,563.01
6,563.16
6,561.61
6,563.07
6,562.98
6,563.06
6,562.94
Emissive
Red Horn
Center l
EW
Å
0.001
6,564.77
0.000
N/A
-0.023
-0.118
0.059
0.009
0.138
0.108
0.328
0.080
0.220
0.223
0.130
0.275
0.207
0.243
0.256
0.091
6,564.76
6,565.29
6,564.59
6,564.12
6,563.48
6,563.25
6,564.78
6,564.76
6,564.72
6,564.65
6,564.83
6,563.02
6,564.56
6,564.59
6,565.06
6,564.20
VR
424.000
-12.696
-2.898
4.492
18.111
1.630
3.500
1.046
3.200
1.218
1.534
2.023
1.440
1.647
1.477
1.191
1.495
Equivalent Width Plot
.
Horn Dance
Hydrogen Alpha region
of
Epsilon Aurigae
Left horn is the blue
emission line
44 Observations
Center is the main
absorption line
11 August 2008
to
13 April 2009
Right horn is the red
emission line
.
The blue horn and
absorption line remain fairly
stable, but the red horn
dances wildly
?? Questions ??
1. What is the source of the emission lines?
2. What causes the EW of the lines to change?
1. What causes the red horn to vary so much?
2. What will be the effect on these due to the
.
eclipse?
3. What other lines will change and how?
Wavelength
Analysis
Radial Velocities
From the following formula, if the change
in wavelength is known, the Doppler shift
and corresponding radial velocity of the
gas can be determined.
Doppler
Radial Velocity
.
V = Dl
*
c/l
Hydrogen Alpha
Observational Data
Date: 14/15 October 2008
JD: 2,454,755
Line
Observed
Center Ha l
Stationary
Ha l
Ll
V
Ha Blue
6,560.62 Å
6,562.81 Å
-2.19 Å
-100 km/s
Ha Absorption.
6,561.98 Å
6,562.81 Å
-0.83 Å
-38 km/s
Ha Red
6,563.25 Å
6,562.81 Å
0.44 Å
20 km/s
Eclipse Start
Lhires III
.
Robin Leadbeater obtained this spectrum
of epsilon Aurigae in the Ki region showing
evidence of the start of the eclipse.
Potassium Ki line
The eclipsing body has potassium in its outer
edges and as it started to pass in front of the F
star the potassium Ki line (7,699 Å) becomes red
shifted. The amount of red shift relates to a +19
km/s radial velocity.
.
This means the leading edge of the eclipsing
body is rotating away from us at 19 km/s.
Conclusion
For those wishing to do meaningful
astronomical research, spectroscopy, both low
and high resolution offers a means.
While challenging, it is well within the
capability of advanced amateur astronomers.
Perhaps. it will the data from an amateur
astronomer or small observatory that provides
some significant answers to the mystery of
epsilon Aurigae.
THE
END
.