Roman#

Roman is supported by StreamObs.

Available releases#

release	bands	footprint	reference band	median reference-band depth
`roman/dc2`	F106, F129, F158	~21 deg² (Roman–Rubin DC2, RA 51–56, Dec −42 to −38)	F158	26.375 AB
`roman/hlwas_wide`	F158	3,513 deg²	F158	26.284 AB
`roman/hlwas_medium`	F158	2,979 deg²	F158	26.289 AB
`roman/hlwas_all`	F106, F158	6,033 deg² (F158); F106 covers only 2,979 deg² within this footprint (see Depth and bands)	F158	26.289 AB

roman/dc2 is the calibration reference (Troxel et al. 2023) for all completeness and photometric-error products. The HLWAS tier maps reuse the DC2-derived tables and apply them via the shared \(\Delta m = m - m_{\rm lim}\) convention; see Roman HLWAS Survey Files for how the tiers are defined and how the depth maps are built, and Roman DC2 Survey Files for the DC2 reference data sheet.

Products#

File(s)	Contents	Drives
`roman_dc2_maglim_{f106,f129,f158,f184}_nside128.fits.gz` (`dc2`); `roman_hlwas_wide_maglim_f158_nside128.fits.gz`; `roman_hlwas_medium_maglim_f158_nside128.fits.gz`; `roman_hlwas_all_maglim_{f106,f158}_nside128.fits.gz`	HEALPix 5σ point-source depth map, one file per band	`Survey.get_maglim(band, pixel)`
`roman_stellar_efficiency_cutf158.csv`	`mag_f158, delta_mag, detection_eff, classification_eff, classification_detection_eff`	`Survey.get_completeness(band, mag, maglim)` — detection + star-classification probability
`roman_photoerror_f158.csv` (sample) / `roman_photoerror_f158_catalog.csv` (catalog)	`delta_mag, log_mag_err`	reference-band (F158) magnitude noise draw / reported-error S/N cut
`roman_photoerror_f158_nocut.csv` / `roman_photoerror_f158_catalog_nocut.csv`	same columns, no S/N selection applied	every other band (forced photometry)
`roman_galaxy_misclass_cutf158.csv`	`delta_mag, missclassification_eff`	compact-galaxy contamination rate (informational; not yet consumed by the injector)
`ebv_sfd98_lowres_nside_512_ring_equatorial.fits`	E(B−V) HEALPix map, shared across releases	`Survey.get_extinction(band, pixel)`

The completeness, misclassification, and photo-error tables above are shared across all four releases; only the depth map differs per release/tier. See Creation for how that sharing is implemented.

Depth and bands#

All maps are nside 128 (RING) HEALPix; off-footprint pixels are set to hp.UNSEEN.

roman/dc2 — per-band maglim maps built via the desqr recipe and truth-anchored (the median is shifted to the S/N = 5 magnitude of the truth-based scatter). Medians: F106 = 26.276, F129 = 26.372, F158 = 26.372, F184 = 25.343 AB, all over the ~21 deg² DC2 footprint as shipped at nside 128 (26.279 / 26.375 / 26.375 / 25.347 over 16.4 deg² before degrading). F184 has a map but is not one of the configured selection-function bands (see Caveats). The DC2 maps serve as the calibration anchor for all HLWAS tier depth maps.

Magnitude-limit maps per band, shown at their nside-1024 build resolution

Truth-anchored S/N = 5 maglim maps over the DC2 calibration footprint in F106, F129, and F158, shown at the nside-1024 build resolution (the maps ship degraded to nside 128). F184 is excluded from the selection-function products (see Caveats).

HLWAS tiers (hlwas_wide, hlwas_medium, hlwas_all) — exposure-time-scaled quasi-depth maps anchored to the DC2 F158 truth-anchored reference depth, so all tiers share the \(\Delta m\) convention of the DC2 tables. F158 medians: wide 26.289 over 3,513 deg², medium 26.290 over 2,979 deg², all 26.291 over 6,033 deg². The hlwas_all release additionally ships an F106 map, but it only covers 2,979 deg² — the medium-tier footprint, not the full 6,033 deg² F158 footprint — so F106 must not be queried outside that region for this release; its median there is 26.194.

HLWAS all-tier all-sky depth maps in F158 and F106

All-sky Mollweide view of the HLWAS “all” tier (wide + medium + deep + ultra-deep) exposure-time-scaled depth maps in F158 (left, median 26.289 AB) and F106 (right, median 26.194 AB), both anchored to the DC2 truth-anchored S/N = 5 reference (nside 128). Off-footprint pixels are shown in white.

Band coverage. The reference band is F158 in every release. roman/dc2 additionally supports F106 and F129 (bands configured: [F106, F129, F158]). The HLWAS tiers support only F158, except hlwas_all, which also ships the restricted-footprint F106 map above. F184 is excluded from every release’s configured bands (see Caveats).

Photometric errors#

The reference band is F158 (completeness_band: F158). Survey.get_photo_error carries two pairs of delta_mag, log_mag_err curves and picks between them automatically:

For the reference band (F158): the sample curve (roman_photoerror_f158.csv, truth-based scatter of obs − true — drives the per-source noise draw) and the catalog curve (roman_photoerror_f158_catalog.csv, median reported SExtractor magerr_auto — drives the S/N cut). Both are measured on the S/N > 5 detected population, because that is the population the reference-band curve is applied to.
For every other band (F106, F129 — forced photometry from the F158 detection): the _nocut pair (roman_photoerror_f158_nocut.csv, roman_photoerror_f158_catalog_nocut.csv), measured with no S/N selection, because a non-reference band’s photometry is not conditioned on detection in that band. Using the detected-population curve there would understate the errors.

Call survey.get_photo_error(band, mag, maglim, kind="sample"|"catalog"); StreamObs resolves the correct pair from band and raises ValueError if the _nocut curve a non-reference band needs is not loaded, rather than silently falling back to the reference-band curve.

Photometry is on the AB system. get_photo_error returns NaN for magnitudes brighter than the configured saturation limit (17.0 mag) — the curves do not model the saturation regime. See Photometric-error derivation for how the curves were built and why the two-curve model is needed at all.

Extinction coefficients#

Dust extinction is modeled using the Schlegel et al. (1998) reddening maps.

The adopted extinction coefficients are the official STScI values from Roman-STScI-000825 (Sharma, Table 3), bandpass-integrated with synphot for a solar-parameter Phoenix spectrum:

Filter	\(A_{\rm band}\ /\ E(B-V)\)
F106	1.1495
F129	0.8497
F158	0.6140

These values are used to compute

\[ A_j = R_j\, E(B-V), \]

and are applied consistently when generating observed magnitudes. The band ratios have been independently validated to ~1% against the DC2 truth dereddening corrections.

Using it in streamobs#

Configured by config/surveys/roman_{release}.yaml, data in data/surveys/roman_{release}/:

from streamobs.surveys import SurveyFactory

survey = SurveyFactory.create_survey(
    "roman",
    release="dc2"
)

maglim = survey.get_maglim("F158", pixel=pix)

completeness = survey.get_completeness(
    "F158",
    mag,
    maglim
)

photo_error = survey.get_photo_error(
    "F158",
    mag,
    maglim
)

For the HLWAS tiers, substitute release="hlwas_wide", release="hlwas_medium", or release="hlwas_all". The completeness and photo-error tables are shared across all releases; only the depth map differs per tier.

Caveats#

Selection functions are derived from the ~20 deg² DC2 calibration region and extrapolated across the full footprint — including the HLWAS (~3,400–5,900 deg²) — through the local magnitude-limit parameterisation \(\Delta m = m - m_{\rm lim}\).
The size-envelope classifier is single-band F158; morphological performance in other bands is not separately characterised.
F184 is excluded from the selection-function products: it is deep-tier-only in the community-defined HLWAS and has an unresolved chromatic calibration issue in the DC2 mock.
DC2 simulations correspond to the HLIS reference design depth (~26.9 AB in F158); HLWAS-wide behaviour (~26.3 AB) is obtained by translation in \(\Delta m\), relying on the universality of the \(\Delta m\) parameterisation.
The hlwas_all map includes deep/ultra-deep pixels where the DC2-derived selection function may be unreliable (see Roman HLWAS Survey Files).
get_photo_error returns NaN for magnitudes brighter than the configured saturation limit (17.0 mag) — the curves do not model the saturation regime.

Creation#

The methodology for deriving all selection-function products is documented in Selection-Function Derivation Methodology.

How the survey was simulated#

All selection-function quantities are measured from the Roman–Rubin DC2 synthetic survey of Troxel et al. (2023): ~20 deg² of image-level simulations of the Roman High Latitude Imaging Survey reference design at full depth, reaching a 5σ point-source depth of ~26.9 AB in the F106/F129/F158 bands and 26.2 AB in F184. Stars are drawn from a Galfast Milky Way model and galaxies from the cosmoDC2 extragalactic catalog. Object detection and photometry are performed with SExtractor on a median coadd detection image, with forced photometry per band over 1039 coadd tiles; the recommended catalog-level selection of S/N > 5 in the detection image is applied on top.

The matched detection→truth catalog links every detected source to its truth counterpart (1″ positional match, flags == 0 + true-S/N > 5 in F158 selection), enabling direct characterisation of photometric uncertainties, detection efficiencies, and stellar classification performance. For detailed construction steps see Roman DC2 Survey Files.

The HLWAS tiers (hlwas_wide, hlwas_medium, hlwas_all) reuse the DC2-derived completeness, misclassification, and photo-error tables unchanged and pair them with their own exposure-time-scaled depth maps (scripts/roman/build_hlwas_maglim_maps.py); only roman/dc2 itself is independently derived end-to-end (scripts/roman/create_streamobs_files_hlwas.py).

Depth-map derivation#

Depth maps describe the spatial variation of the Roman 5σ limiting magnitude across the survey footprint.

DC2: built via the desqr recipe and then truth-anchored — the median is shifted to the S/N = 5 magnitude of the truth-based scatter of (obs − true). Truth-anchored DC2 F158 median: 26.375 AB.

HLWAS tiers (Option B — DC2 truth-anchored, exposure-ratio scaling):

\[ {\rm depth(pix)} = {\rm DC2\_REF\_DEPTH} + 1.25 \cdot \log_{10}\!\left(\frac{t({\rm pix})}{{\rm DC2\_REF\_EXPTIME}}\right) \]

with DC2_REF_DEPTH = 26.375 AB (F158) / 26.279 AB (F106) — the median of the corresponding DC2 truth-anchored maglim map — and DC2_REF_EXPTIME = 770.0 s (the DC2 HLIS reference per-pixel exposure = 5.5 dithers × 140 s/exposure, Troxel et al. 2023 Sec. 3.1); t(pix) is the per-pixel exposure time from the tier’s .hsp exposure-time map, used directly. F158 wide ≈ medium (single-band) because medium’s extra HLWAS depth vs. wide is in additional bands, not F158 alone; both tiers have the same F158-only median exposure time (~645 s) and the same DC2 reference anchor. Generated by scripts/roman/build_hlwas_maglim_maps.py.

Star/galaxy classification#

The stellar selection function uses a single-band F158 size-envelope classifier that separates point sources from resolved objects in the F158 half-light-radius vs. magnitude plane. Stars that pass the S/N > 5 gate and fall within the size envelope are counted as classified detections. Full details of the size-envelope boundaries and the freeze magnitude are given in Selection-Function Derivation Methodology.

The data file roman_stellar_efficiency_cutf158.csv provides, as a function of \(\Delta m_j = m_j - m_{{\rm lim},j}\):

detection_eff — fraction of true stars with a clean true-S/N > 5 detection, against the full truth-star denominator (the S/N > 5 cut is baked in once here);
classification_eff — fraction of those detected stars classified as point sources by the F158 size envelope;
classification_detection_eff — their product, the combined efficiency used by StreamObs to probabilistically determine whether an injected star is observed.

The combined efficiency is the single curve that StreamObs applies; it encodes both the detection step and the morphological classification step in the shared \(\Delta m\) convention, so the S/N > 5 cut is never re-applied independently.

Compact galaxies (true \({\rm size\_true} < 0.3\ {\rm arcsec}\)) can be misclassified as point sources by the size-envelope classifier, producing a contaminant population for stellar-stream analyses. The galaxy contamination curve is stored in roman_galaxy_misclass_cutf158.csv as missclassification_eff vs \(\Delta m\). The compact-galaxy size used to select the sample is an interim measured-Roman proxy (SIZE_SOURCE = measured_roman; the cosmoDC2 true-size upgrade is pending).

Detection efficiency, stellar classification efficiency, combined stellar efficiency, and galaxy contamination efficiency as a function of distance to the local F158 magnitude limit \(\Delta m = m - m_{\rm lim}\). The bright cut at \(\Delta m = -8.7\) (saturation) removes rows where flux non-linearity affects the size measurement.

Photometric-error derivation#

The reported SExtractor magerr_auto underestimates the truth-based scatter of \((m_{\rm obs} - m_{\rm true})\) by a flat factor of approximately 2 in all bands (underestimated correlated coadd noise). StreamObs therefore uses the two-curve model described in Photometric errors above: the sample curve is the truth-based scatter, the catalog curve is the median reported magerr_auto. Both curves are stored as log_mag_err (base-10 logarithm of magnitude uncertainty) vs delta_mag. A bright-end saturation cut removes rows with \(\Delta m < -8.7\); for brighter sources the scatter increases discontinuously due to pixel saturation.

Roman F158 photometric error model

Photometric uncertainty \(\sigma_m\) as a function of distance to the local F158 magnitude limit for the sample (truth-based scatter, drives the noise draw) and catalog (reported magerr, drives the S/N cut) curves. The factor-of-~2 separation between the curves reflects the underestimated correlated coadd noise in the DC2 mock.

Derivation-level limitations#

The galaxy misclassification curve uses an interim measured-Roman proxy for compact-galaxy size (SIZE_SOURCE = measured_roman); the cosmoDC2 true-size upgrade is pending.

Roman

Contents