Fix setting of boresight rotation angle for imsim data

XMLWordPrintable

Details

• Type: Story
• Status: Done
• Resolution: Done
• Fix Version/s: None
• Component/s:
• Labels:
None
• Story Points:
4
• Sprint:
DRP F19-2, DRP F19-4
• Team:
Data Release Production

Description

For individual obs packages, we are using "translators" to convert metadata from raw files into internally consistent conventions (as many telescopes have different sets of conventions and header naming).  Our convention for boresight rotation angle is:

 At boresightRotAngle = 0, the focal plane pixels are aligned: +x: E->W (-ve RA), +y: S->N (+ve Dec) 

(and see the documentation).

Accordingly, the imsim translator in obs_lsst should reflect that:

 boresightRotAngle = 90 deg - ROTANGLE 

(where the ROTANGLE is the FITS card in the raw data).

The reason data is currently being processed properly is due to a compensation (prior to the translator functionality) for this 90 deg shift in the lsstCamMapper here. This latter "correction" should be removed from there and accommodated in the imsim translator.

This change will be validated by running a visit of DC2 data before and after and checking that we get consistent results.

Activity

Hide
Lauren MacArthur added a comment -

Sounds good to me.  Thanks for the speedy review...merged to master.

Show
Lauren MacArthur added a comment - Sounds good to me.  Thanks for the speedy review...merged to master.
Hide
Simon Krughoff added a comment -

If the question is ROTANGLE vs. ROTANGZ, I don't believe phosim images ever have both. At some point the header card changed. I think they both needs to be treated the same way.

Show
Simon Krughoff added a comment - If the question is ROTANGLE vs. ROTANGZ, I don't believe phosim images ever have both. At some point the header card changed. I think they both needs to be treated the same way.
Hide
Lauren MacArthur added a comment -

Great, that answers one part of the question.  The other is “how should they be treated?”  I.e. does the angle that exists need the same

 boresightRotAngle = 90 deg - ROTANGLE(or ROTANGZ) 

conversion in its translator?

Show
Lauren MacArthur added a comment - Great, that answers one part of the question.  The other is “how should they be treated?”  I.e. does the angle that exists need the same boresightRotAngle = 90 deg - ROTANGLE( or ROTANGZ) conversion in its translator?
Hide
Simon Krughoff added a comment -

I believe that same correction needs to be made to both imsim and phosim translators. The DC1 dataset had both imsim and phosim data and were both using the same mapper, so I believe that correction should continue to be applied to both.

Show
Simon Krughoff added a comment - I believe that same correction needs to be made to both imsim and phosim translators. The DC1 dataset had both imsim and phosim data and were both using the same mapper, so I believe that correction should continue to be applied to both.
Hide
James Chiang added a comment -

I think this correct: both phosim and imsim get that angle from the rotSkyPos parameter in the instance catalogs that both currently use for simulating specific visits.

Show
James Chiang added a comment - I think this correct: both phosim and imsim get that angle from the rotSkyPos parameter in the instance catalogs that both currently use for simulating specific visits.

People

Assignee:
Lauren MacArthur
Reporter:
Lauren MacArthur
Reviewers:
Tim Jenness
Watchers:
Eli Rykoff, James Chiang, John Parejko, Lauren MacArthur, Robert Lupton, Simon Krughoff, Tim Jenness, Yusra AlSayyad