update NRC coron dispersion coeffs

[JarronL]

Update NIRCam coronagraphic wavelength-dependent tilt coefficients. These were miscalculated prior to launch and have since been producing incorrect PSF morphologiers. New values represent actual dispersion from CODEV and Zemax models.

LW example for F444W:

SW example for F200W:

[mperrin]

I confirm this is a necessary fix. I had noticed a while ago the dispersion issue in particular with wide filter coronagraphic PSFs, and brought this to Jarron's attention. Realizing now I had not made a GitHub issue for any of that.

Briefly, for the record:

• Comparing SW PSFs in F200W, or LW in F356W, F444W showed clearly that the sim speckles did not match well, in particular for the angle or dispersion of the speckles at left and right sides.
  
• Multiplying by an ad-hoc scale factor of -10 resulted in a significantly better fit.
  
• For LW I also found discrepant results. Again the sign seems flipped, but the amplitude seems too big instead of too small. Best fit scale factor is like -0.4 for F356W, -0.6 for F444W (very coarsely)

[mperrin]

This PR will require updating the test function tests/test_nircam.py/test_nircam_coron_wfe_offset()

[JarronL]

Done!

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Files	Coverage Δ
webbpsf/optics.py	60.66% <100.00%> (ø)

📢 Thoughts on this report? Let us know!

[mperrin]

Looks good to me. Nice improvement in PSFs for wide filters with coronagraphy.

[mperrin]

@obi-wan76 FYI this is ready for merging. If you'd like I can go ahead and do so myself, just let me know, but by default I'm assuming you and/or @BradleySappington are handling all the PR merges now.