supernova ASASSN-15fa at ~mag 17.5 with the ALPY 200

[Robin Leadbeater]

I have the ALPY configured in the low resolution mode again with a 200 l/mm grism to test how faint I can get. This is the first 20 min exposure of fading supernova in NGC6319 which was reported at mag 17.1 on 20150409 so is probably around mag 17.5 now which is a new record for me. (Has anyone else gone fainter?) The over exposed bright star at the bottom is the edge of a mag 11 star caught in the slit. I could not see the target so to place it on the slit, I had to calculate the offset relative to the guide star. A quick reduction even of this single exposure shows the typical features of a type 1a though it is rather noisy. Hopefully I can get a 2 hour run on it tonight.

Robin

[HughAllen]

Hi Robin, could you explain in more detail the set-up to obtain this spectrum. It looks like you have the 'zero order' slit in your image, is that correct? And what do you mean by the Alpy 600 being in the low resolution mode, I thought it had a fixed 600l/mm grating? I guess these two questions are linked, I am intrigued
Cheers
Hugh

[Robin Leadbeater]

Hi Hugh,

In my quest for going fainter eg for supernova identification, I can change the standard 600 l/mm grism in my ALPY for 200 l/mm one which I got Paton Hawksley to make for me. This is more efficient and also gives lower dispersion, hence the zero order appearing in the field. There are more details in the poster paper here
http://www.threehillsobservatory.co.uk/ ... opy_20.htm

Cheers
Robin

[HughAllen]

Thanks Robin, that's a very interesting poster. Could I ask you one more thing, to explain the calculation behind the resolution of the two Alpy set-ups (11 v 45Å)?
Much appreciated
Hugh

[Robin Leadbeater]

Hi Hugh,

The resolution here was measured as the typical width (FWHM) of the lines in the calibration lamp spectrum but the 4x reduction in resolution is as expected given the 4x reduction in dispersion between the 600 and 200 grisms. You can also calculate the resolution as it is effectively the dispersion (eg in A/pixel) * width of the image of the slit (in pixels)

Actually in the observation I made last night, I 2x binned the image to get a bit more signal so was slightly undersampled with just under 2 pixels/slit image width so the resolution would have been slightly lower, around 55-60A. still good enough for a preliminary supernova identification though (The GELATO supernova identification system for example applies a 50A boxcar filter to submitted spectra by default in any case)

Robin