Flag definitions and categories#
Flag inventory across DP1 tables#
DP1 contains 660 flag columns distributed across five catalog tables, as listed in the table below.
Table |
Flag columns |
Flag types |
|---|---|---|
512 |
Pixel quality, measurement failure, extendedness, deblending, and calibration flags across 5 photometric bands (u, g, r, i, z). |
|
82 |
Centroiding, photometry, deblending, and calibration flags. |
|
15 |
Pixel quality and forced photometry flags. |
|
36 |
Pixel quality, real/bogus classification, and dipole flags. |
|
15 |
Pixel quality and forced difference-image photometry flags. |
Flag categories#
To help users interpret flag meanings, the following sections organize flags into categories based on what the flag indicates.
Pixel quality flags#
Pattern: {band}_pixelFlags_*
Purpose: Report on issues with individual pixels in the source footprint, derived from image mask planes.
The most commonly used pixel quality flags are listed in the table below.
Flag name |
Tables |
Meaning when set to 1 |
|---|---|---|
|
Object, Source, DiaSource |
Saturated pixels in footprint; photometry unreliable. |
|
Object, Source, DiaSource |
Saturated pixel in central 3x3 footprint; critical quality issue. |
|
Object, Source, DiaSource |
Cosmic ray detected and interpolated in footprint. |
|
Object, Source, DiaSource |
Cosmic ray at center. |
|
Object, Source, DiaSource |
Interpolated pixels in footprint (from CRs, defects, saturation). |
|
Object, Source, DiaSource |
Interpolated pixel at center; affects core photometry and shapes. |
|
Source, DiaSource |
Source on CCD edge (deprecated for Object coadds; see |
|
Object |
Detector boundary crossed footprint. |
|
Object |
Detector edge near center; important for coadds. |
|
Object, Source, DiaSource |
Known bad pixels (detector defects) in footprint. |
|
Source, DiaSource |
Suspect pixels (near saturation, non-linear response). |
|
Source, DiaSource |
Suspect pixel at center. |
|
Object |
Artifact rejection during coaddition excluded input pixels. |
|
Object |
Clipping occurred at center. |
|
Object, Source, DiaSource |
No pixel data available (off coverage area). |
|
Object |
PSF model discontinuous at center (OR of CLIPPED, REJECTED, SENSOR_EDGE). |
Key points:
Center variants: Flags with
Centersuffix indicate the issue affects the object’s central footprint (typically a 3x3 pixel box), which is more critical for photometry and shapes than flags affecting only the outer footprint.Coadd-specific flags: On Object table coadds,
pixelFlags_edgeis deprecated. UsepixelFlags_sensor_edgeandpixelFlags_sensor_edgeCenterinstead, which indicate where detector boundaries from input visits crossed the object.
Measurement failure flags#
Pattern: *_flag (algorithm-specific)
Purpose: Indicate that a particular measurement algorithm failed or produced unreliable results.
The general measurement failure flags are listed in the table below.
Flag name |
Tables |
Meaning when set to 1 |
|---|---|---|
|
Object, Source, ForcedSource |
PSF flux measurement failed; do not use PSF flux. |
|
Object, Source |
Galaxy model (CModel) fit failed; do not use model fluxes. |
|
Object |
Kron aperture flux failed (bad radius, near edge). |
|
Object, Source |
Aperture flux in NN-pixel aperture failed. |
|
Object |
GAaP (Gaussian Aperture and PSF) photometry failed. |
|
Source, DiaSource |
Centroid algorithm failed; do not trust position. |
|
Object, Source |
Star/galaxy classifier failed; extendedness value unreliable. |
|
Object |
Shape-based star/galaxy classifier failed. |
|
Source, DiaSource |
Shape measurement (second moments) failed. |
|
Object |
HSM Regaussianization shape measurement failed. |
Key points:
Measurement values when flagged:
{band}_psfFlux_flagis set when the algorithm returns a non-finite flux or a flux error. When the flux is finite, a set flag could be related to other types of errors, such aspixelFlags_saturatedCenter = 1.Subflag pattern: Many algorithms provide detailed subflags explaining why the measurement failed (e.g.,
psfFlux_flag_edge,psfFlux_flag_noGoodPixels). If the general flag is set to1, the specific failure reason may be in a subflag, but the general flag alone is sufficient to filter the measurement.Usage rule: If a measured quantity (flux, shape, etc.) is used, the corresponding general flag should be required to be
0. For example, when usingr_psfFlux, requirer_psfFlux_flag = 0.
Deblending flags#
Purpose: Indicate status and quality of the deblending process (separating overlapping sources).
The Source table deblending flags are listed in the table below.
Flag name |
Tables |
Meaning when set to 1 |
|---|---|---|
|
Source |
Number of deblended children. If > 0, this is a parent (blended group) and should not be used as an independent object (to avoid double-counting). Require |
|
Source |
Deblender skipped this source (too complex, too many peaks, or footprint too large). This parent record contains flux from multiple objects. Exclude sources with |
|
Source |
Deblending skipped due to excessive peaks in footprint (often accompanies |
|
Source |
Deblending skipped because parent footprint was too large. |
|
Source |
Source treated as point source during deblending (informational; not a failure). |
|
Source |
Blend had unassigned “stray flux” not allocated to children; photometry may be incomplete. |
|
Source |
Majority of parent footprint was masked; often leads to |
Typical science cut: deblend_nChild = 0 AND deblend_skipped = 0.
This ensures the sample includes only isolated sources or properly deblended children, excluding blended parents and complex cases.
Difference imaging flags#
The DiaSource-specific flags are listed in the table below.
Flag name |
Meaning and recommendation |
|---|---|
|
Detection classified as dipole artifact (imperfect image subtraction). Exclude dipoles ( |
|
Real/bogus score (0–1); higher = more likely real astrophysical source. Note: This is not a boolean flag; it is a continuous floating-point value. Apply a threshold (e.g., |
|
PSF flux on difference image failed. Require |
|
Forced flux on direct (science) image failed. If using science image flux ( |
The ForcedSourceOnDiaObject difference imaging flags are listed in the table below.
Flag name |
Meaning when set to 1 |
|---|---|
|
Forced PSF flux on difference image failed. |
|
Position outside difference image coverage (no template); difference flux invalid. |
Specialized flags#
Other notable flags are listed in the table below.
Flag name |
Tables |
Meaning when set to 1 |
|---|---|---|
|
Object, ForcedSource, ForcedSourceOnDiaObject |
PSF model invalid (no inputs); measurements unreliable. Exclude these sources. |
|
Object, Source |
Blendedness measurement algorithm failed. |
|
Object |
Failed to compute number of coadd input exposures. |
|
DiaSource |
Trailed source (streak) extended off image edge. |
Calibration usage flags#
Pattern: {band}_calib_*
For most science, these flags can be ignored as they pertain to internal use in the calibration process.
These flags indicate whether a source was used in astrometric calibration, photometric calibration, or PSF modeling during single-visit processing.
The common calibration flags are listed in the table below.
Flag name |
Tables |
Meaning when set to 1 |
|---|---|---|
|
Source |
Source used in astrometric (WCS) solution. |
|
Source |
Source used in photometric zeropoint determination. |
|
Source |
Source reserved from photometric calibration (held out for validation). |
|
Source |
Source used for PSF modeling. |
|
Source |
Source reserved from PSF determination. |
|
Source |
Source was a candidate for PSF star selection. |
Important DP1 caveat: In DP1, these flags reflect preliminary single-visit calibration selections and are not updated for final global calibrations (FGCM photometry, refined astrometry). There are known mismatches between single-visit and final calibrations.
These flags are primarily diagnostic.
Calibrator stars can be excluded if needed (e.g., calib_photometry_used = 0 to remove stars used for zeropoint fitting), but the preliminary nature means some true calibrators will not be flagged and vice versa.