IAC Stripe82 Legacy Project

• IAC Stripe82 Legacy Project is a re-reduction of SDSS Stripe 82 data optimized to preserve faint, large-scale diffuse emissions.
• It employs non-aggressive sky subtraction and precise sky rectification techniques to maintain the integrity of low-surface brightness features.
• The project supports deep investigations of stellar haloes, tidal streams, ultra-diffuse galaxies, and Galactic cirrus with enhanced PSF characterization.

Searching arXiv for recent and foundational papers on the IAC Stripe 82 Legacy Project and related Stripe 82 data products. arXiv search query: "IAC Stripe 82 Legacy Project Stripe 82 low surface brightness astronomy" The IAC Stripe 82 Legacy Project is a public re-reduction of the Sloan Digital Sky Survey Stripe 82 data set, optimized to preserve extremely faint, large-scale emission that is commonly suppressed in standard processing. It covers a $2.5^\circ$-wide stripe along the Celestial Equator in the Southern Galactic Cap, with $-50^\circ < {\rm R.A.} < 60^\circ$ and $-1.25^\circ < {\rm Dec.} < 1.25^\circ$, for a total of $275\,{\rm deg}^2$, in the five SDSS filters $u,g,r,i,z$, together with an additional deeper combination band $r_{\rm deep}\equiv g+r+i$ (Fliri et al., 2016, Román et al., 2018). Its defining methodological feature is a non-aggressive sky treatment intended to preserve the characteristics of the background (sky + diffuse light), which makes the survey particularly suitable for low-surface-brightness astronomy.

1. Survey domain and relation to Stripe 82

Stripe 82 is the equatorial SDSS stripe covering $-50^\circ \le {\rm RA} \le +60^\circ$ and $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$. The underlying SDSS region had already been used for deep coaddition, with repeated scanning in five bands over $275\,{\rm deg}^2$, reaching approximately two magnitudes deeper than single-pass SDSS and supporting applications such as photometric redshift estimation, cluster finding, weak lensing, and cross-wavelength studies (Annis et al., 2011). The IAC project differs from that earlier coadd in its explicit optimization for faint surface brightness structures rather than standard pipeline photometry.

In the IAC reduction, each position was imaged about $80$ times over roughly a decade, from 1998 to 2007; of the 303 drift-scan runs, approximately two thirds pass quality cuts on seeing, sky brightness, and transparency and are co-added (Fliri et al., 2016). The resulting data set is between $-50^\circ < {\rm R.A.} < 60^\circ$0 and $-50^\circ < {\rm R.A.} < 60^\circ$1 mag deeper than the single-epoch SDSS releases, while retaining a wide-area footprint and multiband coverage. This combination of depth, area, and sky fidelity is the central reason the project is used for the study of stellar haloes, disc outskirts, tidal debris, intra-cluster light, ultra-diffuse galaxies, and Galactic cirrus.

2. Reduction philosophy and co-addition procedure

The reduction is organized around preserving diffuse emission on all scales. Each SDSS fpC frame is flux-scaled to a common zero-point $-50^\circ < {\rm R.A.} < 60^\circ$2 mag via

$-50^\circ < {\rm R.A.} < 60^\circ$3

so that $-50^\circ < {\rm R.A.} < 60^\circ$4 and

$-50^\circ < {\rm R.A.} < 60^\circ$5

Object masks are generated with SExtractor and conservatively dilated by $-50^\circ < {\rm R.A.} < 60^\circ$6 pixels. The global sky level $-50^\circ < {\rm R.A.} < 60^\circ$7 and its dispersion $-50^\circ < {\rm R.A.} < 60^\circ$8 are then measured by placing $-50^\circ < {\rm R.A.} < 60^\circ$9 random $-1.25^\circ < {\rm Dec.} < 1.25^\circ$0-pixel boxes and using iterated $-1.25^\circ < {\rm Dec.} < 1.25^\circ$1-$-1.25^\circ < {\rm Dec.} < 1.25^\circ$2 clipping. The key design choice is a non-aggressive, single-value sky subtraction: only the global $-1.25^\circ < {\rm Dec.} < 1.25^\circ$3 is removed from each frame, with no 1D or 2D sky model, specifically to avoid over-subtracting diffuse low-surface-brightness features (Fliri et al., 2016).

Quality cuts reject images if

$-1.25^\circ < {\rm Dec.} < 1.25^\circ$4

with

$-1.25^\circ < {\rm Dec.} < 1.25^\circ$5

for $-1.25^\circ < {\rm Dec.} < 1.25^\circ$6, removing about one third of the data. The accepted images are reprojected with SWarp onto a common tangent grid with $-1.25^\circ < {\rm Dec.} < 1.25^\circ$7 pixels and combined into co-adds. The 2016 survey description specifies an unweighted median stack plus a weight map, with Lanczos3 interpolation and no further sky removal in SWarp, whereas the later survey overview summarizes the combination as median-combined with inverse-variance weighting after sky subtraction, photometric calibration, and astrometric alignment (Fliri et al., 2016, Román et al., 2018). This difference reflects distinct descriptions of the co-addition stage rather than a change in the survey’s basic low-surface-brightness objective.

3. Sky rectification and surface-brightness performance

The survey’s background treatment was refined further in the 2018 release of improved sky-rectified images. The rectification begins by constructing masks on the deepest $-1.25^\circ < {\rm Dec.} < 1.25^\circ$8 co-add using SExtractor. An initial segmentation mask for stars and galaxies is dilated with a Gaussian kernel of $-1.25^\circ < {\rm Dec.} < 1.25^\circ$9 pixels to include low-level wings, and a second “diffuse-light” mask is generated via SExtractor in background mode, masking all pixels above threshold. The final $275\,{\rm deg}^2$0 mask is transferred to each of the $275\,{\rm deg}^2$1 frames. On each masked image, a single global sky value $275\,{\rm deg}^2$2 is computed as the clipped mean of all unmasked pixels; for each row $275\,{\rm deg}^2$3, the row sky $275\,{\rm deg}^2$4 is computed by averaging the unmasked pixels in that row; the row correction is then

$275\,{\rm deg}^2$5

Applying $275\,{\rm deg}^2$6 to every pixel in row $275\,{\rm deg}^2$7 produces a sky-rectified image in which residual striping is removed while large-scale background and diffuse emission are preserved (Román et al., 2018).

The survey reports the $275\,{\rm deg}^2$8 surface-brightness limit over an aperture $275\,{\rm deg}^2$9 as

$u,g,r,i,z$0

where $u,g,r,i,z$1 is the pixel-to-pixel RMS of the sky in flux units and $u,g,r,i,z$2 is the photometric zero point. The mean limits are $u,g,r,i,z$3, $u,g,r,i,z$4, $u,g,r,i,z$5, $u,g,r,i,z$6, and $u,g,r,i,z$7 mag arcsec$u,g,r,i,z$8 in $u,g,r,i,z$9, $r_{\rm deep}\equiv g+r+i$0, $r_{\rm deep}\equiv g+r+i$1, $r_{\rm deep}\equiv g+r+i$2, and $r_{\rm deep}\equiv g+r+i$3, respectively (Román et al., 2018). The earlier survey paper quotes an $r_{\rm deep}\equiv g+r+i$4-band surface-brightness limit of about $r_{\rm deep}\equiv g+r+i$5 mag arcsec$r_{\rm deep}\equiv g+r+i$6 and an effective surface-brightness limit, defined as $r_{\rm deep}\equiv g+r+i$7 completeness for an exponential light distribution, of $r_{\rm deep}\equiv g+r+i$8 mag arcsec$r_{\rm deep}\equiv g+r+i$9, with

$-50^\circ \le {\rm RA} \le +60^\circ$0

for total magnitude $-50^\circ \le {\rm RA} \le +60^\circ$1 and effective radius $-50^\circ \le {\rm RA} \le +60^\circ$2 in arcsec (Fliri et al., 2016).

4. Point-spread function, calibration products, and public release

The final co-adds have characteristic seeing of order $-50^\circ \le {\rm RA} \le +60^\circ$3. Median PSF FWHM values in the 2016 release are $-50^\circ \le {\rm RA} \le +60^\circ$4, $-50^\circ \le {\rm RA} \le +60^\circ$5, $-50^\circ \le {\rm RA} \le +60^\circ$6, $-50^\circ \le {\rm RA} \le +60^\circ$7, and $-50^\circ \le {\rm RA} \le +60^\circ$8 in $-50^\circ \le {\rm RA} \le +60^\circ$9, $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$0, $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$1, $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$2, and $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$3, with $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$4 in the combined $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$5 image; the 2018 summary gives an average seeing around $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$6 for the Stripe 82 data set (Fliri et al., 2016, Román et al., 2018). The release includes PSF “stamps” produced with PSFEx, and an ultra-deep PSF extending to radii of $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$7 with a dynamical range greater than $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$8 mag, built from PSFEx cores combined with the stacked halos of saturated bright stars across Stripe 82. This extended PSF characterization is essential because faint PSF wings can redistribute central light into galaxy outskirts (Fliri et al., 2016, Martínez-Lombilla et al., 2019).

The delivered products include co-added FITS images in $-1.25^\circ \le {\rm Dec} \le +1.25^\circ$9 and $275\,{\rm deg}^2$0, exposure-time maps, weight maps, sky-rectified co-adds, PSF products, and object catalogues. The images are provided in $275\,{\rm deg}^2$1 tiles, with identifiers of the form fxxxy. For co-added images in DN units, the photometric convention is

$275\,{\rm deg}^2$2

The catalogues provide separate star and galaxy lists down to $275\,{\rm deg}^2$3 mag. Detection requires $275\,{\rm deg}^2$4 in $275\,{\rm deg}^2$5, $275\,{\rm deg}^2$6, and $275\,{\rm deg}^2$7 with at least three connected pixels at $275\,{\rm deg}^2$8; Kron magnitudes, fixed-aperture magnitudes, effective radii, and moments are included. Star-galaxy separation uses DAOPHOT SHARP, with $275\,{\rm deg}^2$9, and $80$0 on $80$1 (Fliri et al., 2016). The data are publicly available through the survey webpage at http://www.iac.es/proyecto/stripe82/, and the 2016 release is also distributed via SDSS DAS/CASjobs.

5. Scientific scope in low-surface-brightness astronomy

The survey was designed for the low-surface-brightness Universe, and its principal science cases are explicit. These include mapping stellar haloes around nearby galaxies, studies of disc truncations, discovery and characterization of ultra-diffuse galaxies and tidal dwarfs, quantitative analysis of intra-cluster light, and detailed imaging of Galactic cirri and diffuse interstellar dust (Fliri et al., 2016, Román et al., 2018). The 2018 release summary states that the data permit detection and quantitative photometry of features fainter than $80$2 mag arcsec$80$3, while the listed applications include mapping stellar haloes around nearby galaxies out to $80$4 mag arcsec$80$5, quantitative study of intra-cluster light in groups and clusters at $80$6, and detection of faint tidal streams and shells as tracers of recent accretion events (Román et al., 2018).

Specific demonstrators were already presented in the 2016 survey paper. Around NGC 0936, a loop with $80$7–$80$8 mag arcsec$80$9 extends about $-50^\circ < {\rm R.A.} < 60^\circ$00 kpc from the center and is invisible in single-epoch SDSS. At the loop’s tip, a diffuse low-surface-brightness dwarf with $-50^\circ < {\rm R.A.} < 60^\circ$01 mag arcsec$-50^\circ < {\rm R.A.} < 60^\circ$02 in the core, $-50^\circ < {\rm R.A.} < 60^\circ$03 at roughly $-50^\circ < {\rm R.A.} < 60^\circ$04 kpc from the center, and $-50^\circ < {\rm R.A.} < 60^\circ$05 may be the progenitor. Around NGC 0426, the survey reveals ongoing disruption of a dwarf, a $-50^\circ < {\rm R.A.} < 60^\circ$06 kpc-long arc at $-50^\circ < {\rm R.A.} < 60^\circ$07 mag arcsec$-50^\circ < {\rm R.A.} < 60^\circ$08, and an asymmetric stellar halo traced to about $-50^\circ < {\rm R.A.} < 60^\circ$09 kpc; ELLIPSE profiles show an isophotal break at $-50^\circ < {\rm R.A.} < 60^\circ$10 kpc, beyond which ellipticity declines from $-50^\circ < {\rm R.A.} < 60^\circ$11 to approximately $-50^\circ < {\rm R.A.} < 60^\circ$12 (Fliri et al., 2016).

The same data set has also been used to characterize diffuse Galactic light, or optical cirrus, which overlaps in surface brightness with extragalactic low-$-50^\circ < {\rm R.A.} < 60^\circ$13 features. Stripe 82’s uniform depth and area, spanning Galactic latitudes of approximately $-50^\circ < {\rm R.A.} < 60^\circ$14 to $-50^\circ < {\rm R.A.} < 60^\circ$15, allow filamentary structures to be mapped down to $-50^\circ < {\rm R.A.} < 60^\circ$16 mag arcsec$-50^\circ < {\rm R.A.} < 60^\circ$17 (Fliri et al., 2016). A plausible implication is that the survey is useful not only for detecting extragalactic structure but also for disentangling Galactic foregrounds that can mimic it.

6. Methodological consequences and later analyses

Subsequent work has shown that the IAC Stripe 82 Legacy Project is not only a survey resource but also a methodological testbed for ultra-deep imaging. In the thick-disc analysis of five edge-on galaxies, the $-50^\circ < {\rm R.A.} < 60^\circ$18 images were modeled with bulge, bar, thin-disc, and thick-disc components convolved with the measured Stripe 82 PSF using imfit, and the authors constructed “PSF-cleaned” images by adding PSF-convolved residuals back onto the deconvolved model (Martínez-Lombilla et al., 2019). Vertical luminosity profiles were then fitted with a model of two gravitationally coupled, vertically isothermal stellar fluids in hydrostatic equilibrium,

$-50^\circ < {\rm R.A.} < 60^\circ$19

with the ansatz

$-50^\circ < {\rm R.A.} < 60^\circ$20

Mass-to-light ratios were derived from local colour using

$-50^\circ < {\rm R.A.} < 60^\circ$21

That study found that PSF effects are significant when very low surface brightness is reached, especially in vertical profiles. In the radial direction, PSF wings become significant below $-50^\circ < {\rm R.A.} < 60^\circ$22 mag arcsec$-50^\circ < {\rm R.A.} < 60^\circ$23, adding up to about $-50^\circ < {\rm R.A.} < 60^\circ$24–$-50^\circ < {\rm R.A.} < 60^\circ$25 the intrinsic disc light if uncorrected. Vertically, for intermediate-mass systems the PSF affects profiles already at $-50^\circ < {\rm R.A.} < 60^\circ$26 mag arcsec$-50^\circ < {\rm R.A.} < 60^\circ$27 and can boost apparent outskirts mass by factors of about $-50^\circ < {\rm R.A.} < 60^\circ$28–$-50^\circ < {\rm R.A.} < 60^\circ$29; for two low-mass diffuse galaxies the breakpoint is $-50^\circ < {\rm R.A.} < 60^\circ$30 mag arcsec$-50^\circ < {\rm R.A.} < 60^\circ$31, with mass overestimates of about $-50^\circ < {\rm R.A.} < 60^\circ$32–$-50^\circ < {\rm R.A.} < 60^\circ$33 if PSF wings are ignored. The general conclusion is that neglecting PSF deconvolution can produce spuriously large thick-disc light and mass, by up to a factor of approximately $-50^\circ < {\rm R.A.} < 60^\circ$34 in the worst cases (Martínez-Lombilla et al., 2019).

This result clarifies an issue already noted in the project’s science discussion: simulations of stellar haloes and disc outskirts show that PSF scattering can produce halo-like features, although it cannot fully account for the observed light in the Stripe 82 sample (Fliri et al., 2016). The IAC Stripe 82 Legacy Project is therefore best understood not simply as a deep imaging release, but as a survey in which sky subtraction, sky rectification, and extended PSF characterization are inseparable from the astrophysical interpretation of faint structures.