C2020 F3 (NEOWISE) spectrum

[Hubert Boussier]

Thank you Robin,

I understand the principle of the method, but at the end of the process some details are non so clear, so I summarize what I would do if I intended reproduce the process, please tell me if I am OK or not.

"The spectra were first calibrated in relative flux using a nearby reference star." ==> OK clear

"The sky background was removed using a region at the edge of the slit away from the comet" ==> OK clear

" The dust component was then subtracted using a template derived from the Pickles G2v spectrum" ==> I understand that you divide the comet profile by a profile derived from the G2V spectrum.
Now the point is what will I do to obtain this template?

1/ using ISIS tool "continuum" I extract the continuum profile of the GV2 star profile and save it
2/ Then I open the comet's profile
3/I divide the comet's profile by the GV2 profile continuum

I obtained a comet spectrum in which the relative intensities between the different emission lines are not modified, with a quite flat continuum. The corrected profile is not drastically different from the uncorrected comet profile, but the difference is visible.


Hubert

[Robin Leadbeater]

Hi Hubert,

" The dust component was then subtracted using a template derived from the Pickles G2v spectrum" ==> I understand that you divide the comet profile by a profile derived from the G2V spectrum.
Now the point is what will I do to obtain this template?

I subtracted the modified G2v spectrum, not divided by it. (Dividing by it will distort the relative strengths of the emission lines.) The total spectrum is emission + dust reflection. I wanted just the emission spectrum with the continuum = 0 where there is no emission.

I first subtracted the Pickles G2v spectrum from the comet spectrum (both scaled to 1 at ~5500A)
If the dust reflectance was the same at all wavelengths this should leave just the emission but because the dust reflectance curve is not flat there was some remaining continuum. I extracted this (smooth) continuum and corrected the G2v spectrum by it to produce an improved template. (The G2v spectrum multiplied by the dust reflectance curve) I then subtracted this from the original comet spectrum to leave just the emission with the continuum zero everywhere.

Cheers
Robin

[Daniel Dejean]

Hi Robin,
Many thanks for the data you are sharing.
It shows a significant difference between the results computed with a 10pxl binning and 40pxl binning. Based on rough measurement on your graph, the impact is about 60% difference in the computed ratio. Considering this windows as the minimum and maximum value for other measurements, a +/-30% incertitude interval has been added to the graph. Your result (min/max window in green) for july13 looks coherent with others data.



Obviously such rough analysis of the coma centre could be greatly improved by a better knowledge of the binning used for others measurements. If interest it could be also valuable to all use the same binning value, for example 10pxl (45microns with your Atik428) to standardize these results.
Bon ciel,
Daniel

[Robin Leadbeater]

Hi Daniel,

Obviously such rough analysis of the coma centre could be greatly improved by a better knowledge of the binning used for others measurements. If interest it could be also valuable to all use the same binning value, for example 10pxl (45microns with your Atik428) to standardize these results.
l

It is a start but image scale also has an effect so better would be a standard height for the binning zone in arcsec. Also we are measuring the flux in a rectangular area of sky (slit width x binning zone ) so the width of the slit also has an effect. The distance to the comet also changes so we should perhaps measure over a standard area in km? There are lots of possible systematic differences between observations to consider for an accurate result. The trend is clear though

Cheers
Robin