NIRISS SUBSTRIP96 location is incorrect

[tapastro]

A refactor of NIRISS SOSS code in the jwst repository found a discrepancy with the subarray definition for SUBSTRIP96:
https://github.com/spacetelescope/jwst/pull/9000/files#r1909498841

@hover2pi Can you assist in correcting this entry?

pysiaf/pysiaf/prd_data/JWST/PRDOPSSOC-068/SIAFXML/SIAFXML/NIRISS_SIAF.xml

Lines 3725 to 3757 in 0834e98

	<SiafEntry>
	<InstrName>NIRISS</InstrName>
	<AperName>NIS_SUBSTRIP96</AperName>
	<DDCName>NIS_CNTR</DDCName>
	<AperType>SUBARRAY</AperType>
	<AperShape>QUAD</AperShape>
	<XDetSize>2048</XDetSize>
	<YDetSize>2048</YDetSize>
	<XDetRef>94.00</XDetRef>
	<YDetRef>850.00</YDetRef>
	<XSciSize>2098</XSciSize>
	<YSciSize>317</YSciSize>
	<XSciRef>1980.00</XSciRef>
	<YSciRef>258.00</YSciRef>
	<XSciScale>0.06525472</XSciScale>
	<YSciScale>0.06577458</YSciScale>
	<V2Ref>-351.786079</V2Ref>
	<V3Ref>-686.173172</V3Ref>
	<V3IdlYAngle>0.56126717</V3IdlYAngle>
	<VIdlParity>-1</VIdlParity>
	<DetSciYAngle>180</DetSciYAngle>
	<DetSciParity>1</DetSciParity>
	<V3SciXAngle>-89.70138693</V3SciXAngle>
	<V3SciYAngle>0.56126717</V3SciYAngle>
	<XIdlVert1>-129.2904</XIdlVert1>
	<XIdlVert2>7.7307</XIdlVert2>
	<XIdlVert3>7.7314</XIdlVert3>
	<XIdlVert4>-129.2957</XIdlVert4>
	<YIdlVert1>-16.3352</YIdlVert1>
	<YIdlVert2>-16.9581</YIdlVert2>
	<YIdlVert3>3.8884</YIdlVert3>
	<YIdlVert4>4.5345</YIdlVert4>
	<UseAfterDate>2014-01-01</UseAfterDate>

[Witchblade101]

@tonysohn Can you double check this, when you have a chance? The highlighted entry above matches what I see from the last NIRISS update I can find (JSOCOPS-200).

[tonysohn]

@tonysohn Can you double check this, when you have a chance? The highlighted entry above matches what I see from the last NIRISS update I can find (JSOCOPS-200).

Sure, I'll check this and get back to you here.

[tonysohn]

I've consulted with @goudfroo who worked on the SIAF parameters and got answers as well as some useful info.

It turns out that the current NIRISS PRD SIAF is not the issue here. The SIAF parameters for the SOSS subarrays, including SUBSTRIP96 and SUBSTRIP256, are not so straightforward to interpret since they refer to the part of the detector for which the spectral traces on the sky are projected by the GR700XD grism (which has significant optical power as you know) onto the subarrays. This is also explained in the following j-docs page (see paragraph starting with "Note that..."):

https://jwst-docs.stsci.edu/jwst-near-infrared-imager-and-slitless-spectrograph/niriss-operations/niriss-apertures

I think the proper way to handle this issue is to not hard-code the 10 pixels Y-offset in the code, but to read in the parameters from the part of the PRD Subarrays file. These can be found in e.g., PRDOPSSOC-068/subarraysXML/subarraysXML/NIRISS_Subarrays.xml. I've copied and pasted the relevant parts for SUBSTRIP96 and 256 below:

SUBSTRIP96:
        <SubarrayCorners>
            <RowCorner>1</RowCorner>
            <ColumnCorner>151</ColumnCorner>
            <SciRowCorner>1803</SciRowCorner>
            <SciColumnCorner>1</SciColumnCorner>
        </SubarrayCorners>

SUBSTRIP256:
        <SubarrayCorners>
            <RowCorner>1</RowCorner>
            <ColumnCorner>1</ColumnCorner>
            <SciRowCorner>1793</SciRowCorner>
            <SciColumnCorner>1</SciColumnCorner>
        </SubarrayCorners>

Note that the subarray Y-corners are correctly defined with offsets of +10 for SUBSTRIP96 vs SUBSTRIP256 as @hover2pi mentioned.

Changing the code as suggested above will probably require working in the full-frame coordinates frame, and then converting back into the subarray coordinates, but this way, it wouldn't rely on the cryptic SIAF parameters.

[Witchblade101]

@tapastro Does Tony's proposed fix work for you?