Goals

• Retire aXe
• Retire aXeSIM
• Leverage the JWST infrastructure
  • just crds, gwcs and asdf?
• Enable common "Grismconf" infrastructure for:
  • pyLINEAR
  • grizli
  • EM2D, and Nor's other modules
  • JWST pipeline
  • (WFIRST)
• Ensure that users can do all the flavors of geometric transformations (and can understand them)
• Make sure calibration (files) are consistent with the approach to geometry
• Support the following HST modes:
  • WFC3/IR
  • ACS
  • WFC3/UVIS?
• Compatible with JWST pipeline outputs of WFSS modes (NIRISS and NIRCam)

Low-Hanging Fruit?

It would be good to start development work right away during the sprint, if at all possible. Even if we aren't completely certain of the end goals. This will start to get people familiar with the code bases and documentation, and make shake out some subtleties that are difficult to find without looking at the code. So what are some low-hanging fruit coding tasks?

1. Refactor grizli | pyLINEAR | Nor's code to use JWST version of grismconf?
2. Assess performance issues of grismconf vs. JWST pipeline version?
3. Jupyter notebook illustrating how to do the equivalent aXedrizzle without using aXe?
   • E.g. reproduce cell 7 of the WFC3 aXe cookbook without using Astrodrizzle
     • Possibly first converting WCSs to gWCS to include distortion info?
   • Using reproject package (doesn't yet have drizzle algorithm though)
   • Using JWST pipeline version of drizzle

Use-Cases / Workflows

1D spectral extraction for perfectly registered & perfectly calibrated data

Story: As an impatient astronomer, I just want to extract a 1D spectrum for a single source given perfectly registered images and assuming calibrations are correct. I want it fast and I don't want to think about forward modeling.

What do I need as inputs?

• Direct and slitless images with perfect WCSs
• catalog information (coordinates, magnitudes ...)
• Exposure times?
• Noise model? Or components of the noise model (gain, exptime, read noise, subtracted background)?
  • Or is this already in an uncertainty array?
• Detector signatures already removed (e.g. CALWEBB_DETECTOR1)
• Identification of the pixels in the direct image corresponding to "the source" (e.g. from a segmentation map)
• Identification of any pixels in the direct image corresponding to contaminating sources

What does this step do for me?

• Maps pixels in the direct image to x,y, wavelength
• Accumulates the flux at each wavelength (onto some wavelength grid)
  • Perhaps weighted by the flux in the direct image at this cross-dispersion location
• Accumulates uncertainties at each wavelength
• Accumulates some kind of contamination flag at each wavelength

Why would I not want to do this?

• This doesn't do much to address contamination
• This mixes together wavelength and position more than a forward-modeling approach. 
• Uncertainties are likely to be quite misleading

Simulation/Model based extraction of 1D spectrum for perfectly registered and calibrated data taken at several different orientations

Story: As a careful observer, I have taken my data at several different orientations with direct images as well. I would like to extract 1D spectra taking advantage of the observed morphology of the source in the direct image, and a first guess at its spectrum and that of the sources that may be overlapping.

What do I need as inputs?

• Direct and slitless images with perfect WCSs
• catalog information (coordinates, magnitudes ...)
• Exposure times?
• Noise model? Or components of the noise model (gain, exptime, read noise, subtracted background)?
  • Or is this already in an uncertainty array?
• Detector signatures already removed (e.g. CALWEBB_DETECTOR1)
• Identification of the pixels in the direct image corresponding to "the source" (e.g. from a segmentation map)
• Identification of any pixels in the direct image corresponding to contaminating sources

What does this step do for me?

• Maps pixels in the direct image to x,y, wavelength 
• Folds the model spectra through the instrument response 
  • for the target of interest and the contaminants
  • for each orientation
• Fits the models to the data
  • Somehow? 
    • How do we specify parameters to vary?
    • What do we use as the merit-function for the fit? 
    • How do we weight the data?
    • What do we use as the optimizer?
    • How do we calculate uncertainties?
• Accumulates some kind of contamination flag at each wavelength

Why would I not want to do this?

• I might be impatient (see first story), or only interested in a specific feature.
• I might instead be searching for isolated emission lines

Epics

Identify and match data sets

• Should this be part of Astrogrism, or outside the package (e.g. in a notebook, as a tutorial)?
• Archive queries via API for both direct and slitless observations
• Build associations
  • Can/should we use the JWST pipeline machinery for such associations?

Geometric transformations

• Direct image: x,y ↔ RA, DEC

• Slitless image: 

  • ra, dec, wavelength → x, y

  • x,y | RA, dec →  wavelength

  • x,y | wavelength → ra, dec

Astrometric registration

Simulations 

aXe and aXeSIM information at - https://www.stsci.edu/scientific-community/software/axe

aXeSIM relevant ISR https://www.spacetelescope.org/static/archives/stecfnewsletters/pdf/hst_stecf_0043.pdf#page=8

Contamination

Background Subtraction

HST ACS sky background

ACS uses the original aXe implementation of 'Master sky image' (global background) and local background subtraction approach.
Basic method is outlined in aXe paper at
https://iopscience.iop.org/article/10.1086/596715/pdf (section 2.6)

HST WFC3 sky background

Master sky images have been provided for both G102 and G141, however the single images used by aXe do not take into account the full complexity of the sky background of grism observations. A more accurate background subtraction can be achieved by using separate images for each of the background components: zodiacal light, He I emission and scattered light (for G141).

More info at
https://www.stsci.edu/hst/instrumentation/wfc3/documentation/grism-resources/ir-grism-master-sky-images

Relevant ISRs
https://www.stsci.edu/files/live/sites/www/files/home/hst/instrumentation/wfc3/documentation/instrument-science-reports-isrs/_documents/2015/WFC3-2015-17.pdf

https://www.stsci.edu/files/live/sites/www/files/home/hst/instrumentation/wfc3/documentation/instrument-science-reports-isrs/_documents/2017/WFC3-2017-05.pdf

Relevant paper
https://iopscience.iop.org/article/10.3847/1538-4357/aa81cc/pdf (section 3.2.6)

Most recent 'WFC3_Back' (Nor's slides)

Flatfielding

1D Spectral Extraction

1-D extraction
non-weighted extraction (section 3.3.2, figure 11)
https://iopscience.iop.org/article/10.3847/1538-4357/aa81cc/pdf
and
optimal extraction (section 3.3.2, figure 12)
https://iopscience.iop.org/article/10.3847/1538-4357/aa81cc/pdf
- original concept from Horne 1986
https://ui.adsabs.harvard.edu/abs/1986PASP...98..609H/abstract

2D Spectral Extraction

2-D extraction
Simulation based extraction (SBE) details in section 3.2.4 (figure 10)
https://iopscience.iop.org/article/10.3847/1538-4357/aa81cc/pdf

1D Forward modeling

2D Forward modeling

Combine spectra single PA

Combine spectra multiple orients

Documentation

•  Getting Started (landing page)
•  Installation
•  Overview
•  Input/Output
  • images
  • spectra
  • catalogs and tables
  • reference files
•  Matching direct and slitless observations
•  Astrometric registration
•  Background subtraction
•  Spectral extraction
•  Contamination
•  Combining spectra
•  Simulations and forward modeling
•  Using Astrogrism with other packages
  • specutils
  • grizli
  • pylinear
  • EM2D
  • MIRAGE?
• Terminology:
  • Use "slitless" instead of grism?