AI-generated illustration for Galileo Whispers. Diffuse-galaxy concept with illustrative brightness and colors; not a telescope observation, calibrated surface-brightness map or survey detection.
Redshift surveys can miss some low-surface-brightness galaxies because an object must first be detected and selected before a useful spectrum can place it in a three-dimensional catalog. Spreading a galaxy’s light across a large angular area can make it difficult to separate from the sky, even when its integrated brightness would appear sufficient. Selection rules introduce additional boundaries.
This does not mean a redshift survey is a failed census. It means its scientific sample has a definition. Understanding that definition is essential before using the catalog to infer which kinds of galaxies exist or how common they are.
Getting into an image is only the first step
A redshift is inferred from spectral features. Many surveys choose objects for spectroscopy using an earlier imaging catalog. A galaxy absent from that input catalog cannot normally become a spectroscopic target through that selection pathway.
The stages answer different questions:
| Stage | What can prevent inclusion? |
|---|---|
| Detection in imaging | Diffuse light may not stand out sufficiently from the estimated sky background. |
| Selection as a target | Measured brightness, surface brightness, classification, or data-quality flags may fail the sample’s rules. |
| Assignment for spectroscopy | A selected target may not receive an available fiber or other observing opportunity. |
| Successful redshift measurement | A spectrum may lack enough usable information for a reliable redshift. |
The SDSS DR4 target-selection documentation explicitly distinguishes selecting an object from actually observing it with the spectrograph. Treating every missing redshift as a detection failure would confuse separate parts of the process.
Equal integrated light can have different detectability
Integrated magnitude adds the light within a stated measurement aperture. Surface brightness concerns light per angular area. Two objects can supply comparable total light while presenting very different contrasts against the background.
Imagine the same total galaxy light distributed over a compact profile and a more extended, diffuse one. The extended case asks the image-processing system to distinguish a weaker excess over the sky across more area. Background estimation, the brightness profile, and the detection method then matter alongside the total flux.
This is not a universal rule that every large galaxy is missed. A bright central region can aid detection or provide a useful spectrum. Different definitions, such as central surface brightness and mean brightness within a half-light radius, describe different properties. A threshold written for one cannot be applied automatically to another.
A documented historical example: the SDSS main sample
Michael Strauss and colleagues’ 2002 paper on SDSS spectroscopic target selection describes a main galaxy sample using an r-band Petrosian magnitude limit and a mean surface-brightness limit within the half-light radius. “Petrosian” refers to a defined way of choosing a photometric aperture; it is not a claim to capture every photon from every galaxy.
The DR4 documentation describes the magnitude boundary at r = 17.77 and surface-brightness selection extending to 24.5 mag/arcsec², with additional criteria applied in the fainter-surface-brightness range. The band, aperture definition, and extra rules belong with those numbers. They are historical SDSS specifications, not settings for every redshift survey or a current universal detection limit.
Those rules also help reject spurious detections. A suspicious diffuse patch may be a real galaxy, an image defect, or an effect of neighboring bright objects and the background treatment. Target selection balances the scientific sample definition with the practical need to obtain useful spectra.
What does “highly complete” actually include?
Completeness is meaningful only relative to a stated population and selection. A sample can be highly complete for galaxies that satisfy its criteria without including all galaxies outside those criteria.
The Strauss paper reports tests of its selection and explicitly discusses surface-brightness effects. That is a statement about a defined survey sample, not a guarantee that no diffuse population can lie beyond its boundary. Reading a completeness percentage without its denominator can turn a careful technical result into an unsupported claim about the universe.
The SDSS discussion of target-selection quality adds another distinction: targeting and fiber assignment are separate. A close pair can present an observing-allocation issue even when both galaxies were detected. Such an omission should not be counted as evidence that the image could not reveal a low-surface-brightness object.
A missing catalog entry does not identify its own cause
To investigate a missing object, start by asking whether it appears in the imaging catalog. Then check the relevant sample flags, photometric measurements, observing coverage, and redshift-quality information. The appropriate questions depend on the survey; the SDSS example supplies a method of thinking, not a universal checklist of identical flag names.
This matters when interpreting apparently empty regions. Our explanation of cosmic voids notes that a map traces selected objects rather than directly photographing all matter. A galaxy’s absence from one sample does not establish that its location is physically empty.
Deeper imaging, different selection criteria, or another measurement strategy can reveal objects excluded by an earlier design. Comparing the resulting catalogs requires accounting for those differences before attributing every change to the galaxies themselves. The selection function is part of the evidence, not an inconvenience to be forgotten once the map looks finished.
