SPIRE spectroscopy data reduction #2

Tuesday October 11th 2011


Ed Polehampton, Ivan Valtchanov, Catherine Espaillat, Maximilien Schirm,

Ros Hopwood, Peter Imhof, Trevor Fulton, Emilie Habart, Anne-Laure Mealier,

Dominique Benielli, Kate Isaak, Jing-Hua Yuan

  • Round table with observers

Catherine Espaillat:

Not that many observations yet - some just being observed. All Low Resolution.

No detections so far due to high background and weak sources.

Have tried to look at QC data.

Max Schirm:

Looking at mapping observations from SAG2.

Concentrating on point source conversion for SSW detectors.

Emilie Habart:

Working on Orion Bar and NGC 7023.

Comparison between Spectrometer and Photometer maps

Problem with discrepancy and currently scaling the spectra to match the photometer.

  • Ed presenting what's coming in v8

(see SpireSpec_v8_update_11Oct2011.pdf)

- Some questions on the line fitter useful script

- Question from Ivan: does the output contains the fitted parameter errors?

Line flux error is currently given in the script. There is also a

Chi squared object. Ivan and Peter to check this.

- Question from Catherine: where does the line list come from?

At the moment the CO lines are hardcoded into the script, but you

can also read in an ASCII file (example lines to do this are there but

commented out)

- Question from Emilie: can we use the script for cubes?

Peter: the example script itself is supposed to work on sparse mode only.

Ed: it should be possible to fit data in a cube.

PACS are also working on this - i.e. to make a cube of the fitted parameters

to visualise the map of the fit. We need to check how far this development

has got with PACS.

In summary, should you reprocess or not? If you have a good result that you are happy with

from v7 then there is probably no need to reprocess with v8.

  • Ros presenting the latest recommendations on background subtraction:

(see SpireSpec_dark_off_subtraction_11Oct2011.pdf)

The examples are point sources, and so the step between bands cannot be due to the beam

size change.

- Question from Kate: is there a script to do this?

Ros: we are working on an example script that could be included in v8 as a Useful Script.

- Question from Kate: in the final best spectrum, how does this match with the photometer data?

What causes these features and steps?

Ed, Trevor: residual background is to blame.

The first subtraction is done with the telescope model, and what remains is a residual

that changes day by day. There is an additive uncertainty of ~1.5 Jy. The subtraction of the

daily dark or off-axis detectors aims to correct for this additive uncertainty.

Also, the residual background is extended, whereas we are looking at the source after

point source calibration. So there may be beam-size effects in the residual causing wiggles.

- Question from Kate: what is the confidence that a feature seen in the spectrum is real or artefact

and due to the processing? Can a limit be set on the size of wiggles expected?

Ros: it is difficult to set a limit on the wiggles as the residual changes on a daily

basis. Average uncertainty is ~1.5 Jy.

Ivan: one can look at level-1 data, the off-axis detectors and check if there are wiggles

and steps seen in every detector (i.e. the off-axis subtraction method).

- Kate is going to try the subtraction script for improved processing by Ros.

  • Any other issues?

Peter: photometry and spectroscopy comparison.

- Kate did a comparison a while back but it didn't match up - the subtraction of background may help.

- to include in the manual how to do the photometer-spectrometer comparison.

- part of the continuum offset with respect to the photometry may be due to the background subtraction

but also to a pointing offset. The first will not affect the line fluxes while the second will.

There is no way currently to quantify the pointing offset.

Max: point source calibration of the off-axis detectors - is there any chance of getting more than the

co-aligned detectors?

- Trevor: yes, we now also have the first ring of SSW detectors

- Max will try this out.

- We also need to check whether this works OK - the point source calibration should just apply to true

point sources centred on the detector. If there is some extension a further correction may be needed.

blog comments powered by Disqus