crossidentification

Cross-identification is the process of determining when observations from different catalogs, surveys, or instruments correspond to the same real-world object. It is used in astronomy but also in data integration to link records across datasets. It addresses issues of positional accuracy, differing resolutions, time variability, and measurement uncertainties.

In astronomy, cross-identification is essential for creating multi-wavelength or time-domain catalogs. It enables combining photometry, spectroscopy, and astrometry. It supports studies of object properties and evolution. It uses algorithms to compare sky positions within positional uncertainties, sometimes incorporating proper motion and parallax.

Methods: positional matching with error models; probabilistic associations such as likelihood ratio, Bayesian cross-identification; consideration of multiple candidates per target; use of ancillary information (colors, morphology, variability) to select the most probable match; sometimes machine learning.

Challenges: crowded fields, extended sources, source confusion, incomplete or biased catalogs, systematic astrometric offsets, variability causing non-detections in one catalog, proper motion between epochs.

Applications: building unified multi-wavelength catalogs, cross-survey studies, transient event identification, galaxy evolution studies, stellar kinematics.

See also: data matching, record linkage, data fusion, astrometry, catalog cross-matching.