2001
May 8: Transmission grating placed closer to detector - lowering
dispersion to 5nm/pixel for a brighter spectrum of 3C273 - 3m exposure
about minimum for a satisfactory Hb emission
line to gauge a redshift.
2001 April 24: 3C273 spectrum recorded
in short exposures [60s - some co-added] to reveal the redshift Hb
line compared to nearly static reference star of Spica [a
Vir ; exp 0.01s!]. This is the fifth visit to 3C273 and the briefest
exposures yet to show a cosmological redshift which is pretty amazing for
a mag 13 object. This is only possible by comparison to earlier
results shown here. The more difficult Ha
emission line, which tends to merge and be partially masked by the strong
A
absorption
line in far red [right], was off the edge of the CCD at this dispersion.
Fair transparency in Virgo with lvm ~mag4.
2000 June 6: 3C273 revisited in brief exposure which prove sufficient
when viewed together to extract a redshift from two major hydrogen emission
lines using Altair as a fixed comparison. In the 420 days since the
last observations, space between 3C273 and Earth has increased by 1.6 trillion
kilometers =10,000AU = 0.17 lightyears ie 333 times the mean distance [30AU]
of Pluto !
1998
Dec 21: Groundbreaking amateur spectrogram of quasar 3C273 shows
emission lines of hydrogen 'redshifted' i.e. z = 0.16 or 44,000km/s in
recession (see below) due to the expansion of the universe since the Big
Bang. 3C273 (Q1226+023) in Virgo is 2 BLY's away and about a lightyear
across ! To shine at mag 13 at this range requires the collective energy
of a 1000 galaxies crammed into this 'tiny' space and stars and gas falling
into a massive blackhole at the core the host galaxy is the only current
explanation. More quasars are under spectroscopic examination from WPO.
Quasar
3C273 revisited in the spring skies. Repeat spectrograms of quasar 3C273
on 1999 Jan 14 and April 10/11 (shown above)
with exposures of 30 minutes confirm the original results of 1998 Dec 21.
The lower right trace is from a single 5 minute exposure and although
noisy
it clearly shows the redshifted Ha and Hb
emission lines. Note the blue hump or excess which seems characteristic
of quasars of low z.
Three of four quasar spectra captured have been successfully measured
for redshift with z values close to professional results.
Calculating the redshift (z) is straightforward i.e. z = (new
wavelength - wavelength at rest) / wavelength at rest i.e.
[1] z
= (l -lo ) /
lo
= v/c i.e. for low non-relativistic
'speeds'.
Measuring Ha in 3C273 spectrogram, for example,
gives the following in nanometers (755-656)/656 = z0.15 and this repeated
for the three other emission lines giving an average z0.16. However relativistic
effects come into play and at z>1 recessional velocities (v) would exceed
the velocity of light (c) and a modified formula [2 - Universe by
W.J.Kaufmann page 499; ISBN 0-7167-1673-9] is needed to keep recession
below lightspeed .
[2] v/c
= ( z + 1)2 - 1 / (z + 1)2 + 1
i.e. for higher relativistic 'speeds'.
However large z [Astronomy Now website May 2000 quotes z5.8 current
max] is in Formula 2, the numerator is alway 2 less that the denominator
so the result is always less than 1. Applying the 3C273's
z value to the formula thus ((0.16 + 1)2 - 1) / ((0.16
+ 1)2 + 1) = 0.3456 / 2.3456 = 0.147 of light speed
or 44,000km/s in recession so that even for this low z value, 3C273 is
moving away at relativistic speed.
Notes
on spectral gratings........qdisp.doc
words=510
In using a transmission grating to secure quasar spectra some interesting thoughts have surfaced. A star image could be regarded of as an infinite number of spectra of zero dispersion. When an obstruction is placed before a telescope, diffraction patterns are created - particularly noticeable in bright star images as ‘spikes’ emanating for the stellar cores. These spikes are the very low dispersion and overlapping orders of spectra - recordable in colour as adjacent images. When a regular pattern of ‘parallel obstructions’ i.e. a grating is placed into the light beam then a fully dispersed spectrum is created separated from the real star image.
The grating, in theory, forms a complete spectrum covering the entire electromagnetic spectrum from very short wavelength i.e. gamma-rays (abutting the star image) to radio remotely off to one side but obviously only a portion of the spectrum is recorded with a particular detector. A CCD camera only ‘sees’ into visible and near-IR and the receiving ‘dish’ or optics also have a part to play in detection.
As I currently use the grating without a slit (which normally isolates a single star) so all the stars in the field have their spectra recorded together with their real (starlike) image offset to the side. Initially this was considered a disadvantage until is was realized the separation between real image and its spectrum was precise. It could be used to measure any shift in the spectral lines much like a comparison spectrum added to a spectroscope with a slit.
All that is needed is to count the number of pixels (in a sample image like a spectrum of Vega or Altair) from the real image to a particular hydrogen line and do a similar count for the hopefully redshifted quasar. The count difference is the redshift. There is no need to convert to angstroms or nonometers to do this - a pixel count will suffice. Care it obviously needed to ensure that no real images of adjacent field stars overlay the spectrum under observation creating the impression of false emission lines!
The Rainbow Optics grating has proved relatively efficient in this task Some simple photometry indicated ~ 70% of the light goes into the blazed first order spectrum as used; ~ 10% into the zero order spectrum (or real image) and the balance ~20% is lost in other weaker orders of spectra and general transmission loses through the grating itself (its a bit like a very pale grey neutral density filter).
An efficient grating is vital when trying to record faint stellar spectra. A clue to the efficiency of a particular grating may be judged by the brevity of exposure on a bright star and projecting this at x2.512 for each magnitude fainter. Thus to recorded a mag 15 spectrum in an hours exposure the same set-up and spectral resolution must record a mag1 star in 0.01s (1/100 second) ! My tests confirm these results and the x2.5 per mag rule holds up well.
[c] M.Gavin - March’99