3C273- Astronomy and astrophotography CCD camera images

Object name: 3C 273
Object type: Quasar
Magnitude: 12.0
Apparent RA: 12h 29m 23s
Apparent Dec: +2 1' 9"
Constellation: Virgo

This is my first Quasar! 3C 273 is one of a class of objects known as "Quasi-Stellar Objects" - hence Quasar. The they are extremely distant and extremely bright galaxies. 3C273 is one of the brightest of this class and easily within range of our webcams. I have made a number of different observations of this object.

This scores as my most distant object yet imaged - at around 2 billion light years.

The right hand image is a stack of 88x3s frames. In my humble equipment, quasars are not, lets face it, terribly exciting to look at. Its a basically a dot of light in amongst lots of other dots of light. Using the figures published here I was able to estimate its magnitude at 12.5. The teleauto screenshot is here. This figure is only relative, and not terribly accurate as I don't have photometry filters.

Taking inspiration from Robin's experiments, I decided to try and take its spectrum with my 100lines/mm grating.

First of all, I need to calibrate my equipment. I used the hot star eta Virginis which exhibits strong balmer absorbtion lines - see the analysis here. This told me my current focus and grating distance gave me a resolution of 27.8 per pixel. This is a very low resolution, but gives a bright spectrum - vital when imaging a 12th magnitude source.

Next we image the spectrum of the quasar, making sure that the spectrum does not cross any other stars (I use a technique known as "luck" for this) and keeping the spectrum aligned so that RA tracking errors do not effect it.

The result of 100x20s frames can be seen on the right. As you can see, the spectrum is very faint. Processing in vSpec and calibrating with the above resolution figure gives the graph below.

As 3c273 is so far away, its spectrum ought to be red-shifted to to the fact its travelling away from us at great speed.

I have marked the wavelength of what looked like the most promising peaks in the noise. Refering to Robin's result, I'm guessing that the right hand peak is the redshifted H Beta line. Working the maths:

5657 - 4861 = 796
796/4861 = 0.163

The left hand spike could also be an H line - the H Gamma line at 4340.

5045-4340 = 705
705/4340 = 0.162

My major error was forgetting to remove the IR filter from the optical path. This means that the redshifted H Alpha emission line is not visible - it has been redshifted outside of the bandpass of the IR filter.

This result is close enough to the published figure of 0.158 to keep me happy. Due to the low resolution, tiny tweaks to the calibration can change the result by +-0.01. The spectrum is extremely noisey - I need to think of ways of getting better SN ratios in faint spectrums.