Discussion items

General UpdatesNikole Lewis
 TSO Branch of Pipeline OverviewNikole Lewis 
 'Optimal' CalTSO3 Photometry Procedure ProposalEverett Schlawin, Jonathan Fraine
Jonathan Fraine
Multiple methods of background subtraction available to users in the pipeline.
Four methods of centering. For out of focus data, flux-weighted centering misses the target by a pixel or two. Gaussian centering is precise but very temperamental. 
Take a median image and MAD image and compare to flag cosmic rays through the data cube.
Background calculation
  1. Median of the whole frame (not for weak lens case)
  2. Median of an annulus around the source (avoids flat fielding problems) Might be the best method.
  3. Median of whole frame with source masked
 'Optimal' CalTSO3 Spectral Extraction ProposalKevin Stevenson
Kevin Stevenson

Compared results from spectral extraction methods in CALTSO3 pipeline
Added hot pixels and cosmic rays to test
Basic: Sum along column
Standard: Sum fluxes within 16 pixels of trace
Optimal: Weighted spectral extraction (Horne 1986)
Make a stack from all images and make a median image from which to get a trace mask.
Made a mask for the traces and used the rest for background subtraction
The choice of spectral extraction dilutes the white light transit depth predicted by the model
Standard and optimal methods are equivalent if bad pixels are removed. Optimal is necessary and safest way if cosmic rays are present.
 Other Proposals?all
Stephan Birkmann

“Columns of Death” are a symptom of a moving bias across the frames.
2 column wide defect that moves one column each integration.
Weaker for NIRISS, pronounced in NIRSpec
Tracked by the reference pixels so could be subtracted. At what step in the pipeline though?
Loic says a rolling reset can be used on NIRISS, resetting 1/4 ramps before each integration.
 General Discussionall 

Spectral Extraction Code

Location of NIRISS/SOSS data:

Action items