A spurious signal or feature in data caused by instrumental or environmental effects, not a real physical phenomenon.In physics and astronomy, an artifact is a spurious signal or feature in data that arises from instrumental limitations, environmental interference, or data-processing quirks — not from the actual phenomenon under study. Artifacts are the 'noise' that mimics 'signal,' and distinguishing between the two is a core challenge in experimental science. For example, a detector's electronic noise, a cosmic ray striking a sensor, or a telescope's optical distortion can all produce features that look like real astrophysical events.
The term has been used in experimental science for over a century, but its systematic treatment grew with modern detectors and imaging in the mid-20th century. No single inventor; it emerged from practical need.
Artifacts emerge from the imperfect chain between a physical event and recorded data. A sensor may have a nonlinear response, a mirror may scatter light, or software may introduce ringing during image reconstruction. Environmental factors — temperature drift, electromagnetic interference, or even dust on optics — can also imprint false patterns. Scientists identify artifacts by comparing data across multiple instruments, repeating measurements, or using simulations to predict what instrumental effects should look like, then subtracting or masking them.
In astronomy, artifact removal is essential for interpreting images from telescopes like Hubble or JWST, and in medical imaging, it prevents false diagnoses from MRI or CT scan distortions.
Some famous 'discoveries' — like canals on Mars or certain pulsar signals — were later attributed to artifacts of the instruments or the observer's eye.
