SN2014j slight change in spectrum!

[Gasman]

Hi

SN in m82 seems to be changing slightly, a few different lines appearing and SN is still fairly bright and an easy object to find.

Stack of 5x 200secs

Comparison with spectrum from the 11th

Steve

[lunator]

Steve

It looks like the Sodium seems to be declining?

Thanks for sharing

Cheers

Ian

[robin_astro]

Steve

It looks like the Sodium seems to be declining?

Yes that is rather strange. The Na absorption line comes from the interstellar medium (probably mainly in M82) rather than from the supernova material so would be expected to stay constant as a percentage of the supernova emission intensity in that region of the spectrum.  Christian Buil's results for 6th and 21st Feb for example show the NA D lines  ~constant between these two dates.

A background subtraction issue perhaps?

Cheers

Robin

[Gasman]

Hi Robin

So should the background be subtracted or divided from the main spectrum please? Subtraction would imply that its actually subtracted although you recently advised me to divide a daylight spectrum from a Jupiter spectrum to get the spectral signature of Jupiter!

Regards

Steve

[robin_astro]

So should the background be subtracted or divided from the main spectrum please? Subtraction would imply that its actually subtracted although you recently advised me to divide a daylight spectrum from a Jupiter spectrum to get the spectral signature of Jupiter!

Hi Steve,

These are for two different purposes.

Sky background SUBTRACTION should always be done as part of the image processing step on any spectrum (just like dark subtractrion) to remove the spectrum of the night sky, light pollution or in this case mainly contamination from the spectrum of the surrounding galaxy which the SN is embedded in. With a star you typically take the median of zones above and below the spectrum, though other more complex sky background modelling can sometimes be used.  With an extended object which might extend beyond the length of the slit or with fibre fed spectrographs you need to take a separate spectrum of an adjacent bit of sky and subtract this.  You can see an example of the selection of zones in a tricky situation using the Star Analyser and ISIS here.  

http://www.threehillsobservatory.co.uk/astro/T_Tauri_campaign/ISIS_binnning_background_zones.png

With a slit spectrograph it is much easier, though even then with cases like the SN in the Galaxy it can be tricky to chose representative background zones. (The objective is to get the best approximation to the background at the SN location)

To get the true reflectivity of an object illuminated by a light source however you need to DIVIDE the spectrum of the object by the spectrum of the light source (in this case you divide the Jupiter spectrum by the solar spectrum) to remove the effect of the light source spectrum.  (sort of similar to the effect a flat has on an image)

(Like any other spectrum, you would  have already SUBTRACTED the sky backgound during the processing of  the  Jupiter spectrum)

Cheers

Robin

[robin_astro]

Just to clarify further,  The background SUBTRACTION only removes the light contaminating the spectrum. It will not remove the effects of absorption between us and the object, like the sodium interstellar line  (which we know is not from the supernova because it is very narrow and not Doppler broadened) or for example Telluric lines from our atmosphere. To remove these effects we need to DIVIDE  by an estimate of the absorption spectrum.

To sumarise:-

If the contamination is from light which gets added into our spectrum from adjacent regions then we SUBTRACT this contamination

If the contamination is from absorption between us and the object (or is due to the illuminating light source for reflecting objects like planets) which modifies our spectrum, then we DIVIDE the spectrum by the contaminating spectrum.  

This absorption effect  is relatively easy to measure and correct  for if it comes from our atmosphere but is tougher to quantify if it is from interstellar material.  

Effects such as the strength of the interstellar lines (like Na D here) and the amount of redening of an object can be used to estimate the effect though, which is large for this object  (The visual brightness only reached Vmag 10.5 compared with 8.5 if there had not been any absorption which means that something like  85% of the visible light from this object never reached us and was absorbed en route)

Cheers

Robin

[Gasman]

Ah! I see. That's exactly what I did with this spectrum.I took a spectrum of the SN and then took one a short distance away for the pollution and subtracted this from the first. I also calibrated them both first before the subtraction just to make sure they were in line, not sure that's necessary? Intuition would dictate that if something is there that you don't want then you would subtract it. (At least my intuition does ;-))

Cheers

Steve

[nytecam]

Good tutorial Robin on this complex object