SIR questions (key reference: SIR Q1 paper):

1. How do I find the 2D grism images corresponding to the 1D spectrum?
There are no tools available for doing this at present. There is no easy way to do this based on information currently provided by ESA.

2. How do I extract the spectrum myself from the 2D spectral exposure?
There are no tools available for doing this at present. There is no easy way to do this based on information currently provided by ESA.

3. Is it possible to have the location table of the spectra in the 2D so that I can extract the spectra myself?
ESA is not currently providing a location table.

4. How are spectra of extended sources non circularly symmetric combined together?
Single-dither spectra are first extracted using per-dither virtual slit, and then combined. It therefore relies on the spectro-spatial separability assumption (see SIR Q1 paper), and further neglects internal kinematics.

5. Why does the continuum jump at X wavelength?
Probably because of issues such as  background subtraction, decontamination, flux calibration.

6. How do I get the line spread function for an extracted spectrum? 
The SIR product DpdSirScienceFrame provides the standard deviation of the effective line-spread function (LSF). The LSF is an intricate mixing of instrumental PSF and intrinsic source extent. The object thumbnail (after convolution with the NIR/SIR differential PSF) is rotated and sheared to bring dispersion direction horizontal, and virtual slit vertical. The LSF standard deviation is estimated from a 1D Gaussian fit to the resulting thumbnail marginalised over the cross-dispersion direction.

7. How is persistence masked in the SIR 2D spectral images?
There's currently no persistence masking in dispersed images.

SPE questions (key reference: SPE Q1 paper):

1. What are the flags that I should use to select good quality spectra? How do I select valid wavelength ranges in the spectra? 
To select good quality spectra, choose spe_rank = 0 (best solution), and set these thresholds: f(Hα) ≥ 2×10⁻¹⁶ erg s⁻¹ cm⁻², Hα line S/N ≥ 5, and SPE redshift probability threshold spe_z_prob ≥ 0.99. In addition, check the line catalog for Hα detections. The valid wavelength range is 1.25 to 1.85 microns, corresponding to the H-alpha redshift range of 0.9 < z < 1.8. 

2. How do I find quantitative information on data quality? Is there a summary data quality value for each spectrum?
The spectroscopic data quality is indicated by SPE redshift probability threshold, gal_spe_z_prob, with higher values indicating more reliable redshift measurements.

3. What is the meaning of the redshift probability flag? How do I know the quality of a redshift measurement?
Each object in the spectroscopic catalog has a probability distribution function (PDF) of the redshift measurement. The spe_z_prob parameter gives the value of the integral of the PDF under the corresponding peak, over a +/- 3σ interval around the central value, where σ is the width of the Gaussian fit to the peak. This provides an indicator of the reliability of the redshift estimate: a spectrum with two or more strong coherent emission lines will typically yield a PDF with a single peak whose integral is close to unity. A spe_z_prob value very close to 1 indicates good quality redshift measurement.

4. How are the spectroscopic redshift measurements made?
Redshifts are measured using a modified version of the Algorithm for Massive Automated Z Evaluation and Determination (AMAZED; Schmitt et al. 2019), tailored for Euclid spectroscopic properties, to calculate redshifts from spectra. This provides a broad classification of spectra into three categories: galaxies, stars, and quasars. For each category, dedicated models are fitted using a least-squares-fitting algorithm, yielding a probability distribution function (PDF) and a Bayesian evidence estimate for each category. For each category, the best-redshift solutions are identified by the most prominent peaks in the PDF. Up to five solutions are provided for each category, and are listed in three catalogues named spectro_zcatalog_spe_TYPE_candidates, where TYPE can be any of galaxy, star, or qso. Each object, identified by its object_id as assigned by the PF-MER pipeline, appears as many times as the number of redshift solutions, each identified by the spe_rank parameter, ranging from 0 (best solution) to at most 4. These solutions are provided for all categories, independent of the actual classification result.

PHZ questions:

1. How to select reliable photometric redshifts?
The quantities phz_90_int2/1 can be used to select sources with 90% of the redshift probability within an interval of a given size. To select photo-z's expected to be good to 10%, for example, one can apply the selection (phz.phz_90_int2 - phz.phz_90_int1) / (1 + phz.phz_median)) < 0.1.

2. What error bars should I use on photo z?
The 70% probability upper and lower values from phz.phz_70_int2/1 can be interpreted as roughly the 1-sigma error values. I.e., upper error bar is size (phz.phz_70_int2 - phz.phz_median), lower error bar is size (phz.phz_median - phz.phz_70_int1).

MER question:

How do I convert MER fluxes to magnitudes?
See MER Photometry Cookbook at http://st-dm.pages.euclid-sgs.uk/data-product-doc/dmq1/merdpd/merphotometrycookbook.html:
Fluxes are in units of micro Jy, mag = -2.5*log10(flux) +23.9