RIQ MET-2112 MvdA : Background subtraction for IFU data

[astronomyk]

https://jira.eso.org/browse/MET-2112

The background subtraction schema described here is perfectly appropriate and corresponds to the best available approaches for imaging, but appears to imply that a similar schema will be followed for the IFU. One of the basic issues with background subtraction in IFU data relates to the presence of sharp emission lines in the night sky background. Variations in the spectral response over the field of view will cause these to be only visible in some parts of the field of view, leading to strong background variations, the effects of which can be mitigated using the spectral information in addition to the spatial information for the background subtraction (as is e.g. implemented in the SINFONI and MUSE pipelines). Is such an approach foreseen for METIS IFU data?

@sesquideus @hugobuddel @ivh - are any of you aware how strong these background variations could be? Or how this is implemented in MUSE for example?

[ivh]

I don't quite understand what he means by "Variations in the spectral response over the field of view will cause emission lines to be only visible in some parts of the field of view".
The spectral response is more or less the same over the IFU FOV, no? @Rumpelstil
We should confirm with Wolfgang & Roy that they deem the observing strategy (nodding, sky pointings) to be sufficient for background subtraction.

[hugobuddel]

I'm not sure what to answer here. This indeed goes beyond the scope of the DRLD and is more a question of the Calibration Plan, so it would be appreciated if @rvboekel / @Rumpelstil could answer.

In 6.6.3 "metis_ifu_reduce: IFU basic data reduction" we state

The sky and thermal background, as well as residual straylight, is estimated from blank sky observations if these are available in the sequence of input frames or by combining (dithered) science frames.

If we don't come up with something better, I suggest a deflecting reply along the lines of

We will investigate whether the background subtraction as implemented in the SINFONI or MUSE pipelines would be an improvement of our current IFU background subtraction algorithm.

[hugobuddel]

After discussion with @rvboekel , answered with

We will do spatial modulation for the IFU mode. That is, there will be a separate observation with just the sky that is subtracted. This process should remove the sharp emission lines.

SINFONI and MUSE don't do spatial modulation and therefor have to apply a more complex background subtraction scheme that is not necessary for METIS.

post scriptum from RvB: as discussed yesterday I don't know for a fact that SINFONI and MUSE do not use spatial modulation, but assuming this is the case the RIX makes sense whereas otherwise I think it does not. Did you check the SINFONI/MUSE approach? (if not, then still no need to spend effort on this now, we'll see soon enough in the RIX answer feedback from ESO)