The photometric measurements are from a custom built point-spread-function fitting routine (written in C) called tphot, and is based on the algorithms discussed in Tonry (2011) and Sonnett et al. (2013).

The calibrated target image allows a calculation of the PSF shape and aperture correction (called apfit) as a function of the location on the CCD, and applies the apfit flux correction accordingly. That PSF is used on the difference image (when difference image photometry is requested) or the target image (when that mode is requested).

The output of our forced photometry measurements is a space-separated plain text file with the following column entries:

MJD	Modified Julian date of the start of the exposure (all exposures are 30s)
m	AB magnitude corresponding to the flux measurement
dm	error on m (in magnitudes)
uJy	flux in microJanskys
duJy	error on uJy
F	the ATLAS filter (typically c for cyan, or o for orange)
err	tphot error code normally 0 for the 2 parameter fit.
chi/N	reduced chi^2 of the PSF fit
RA	RA at which the PSF is forced (decimal degrees)
Dec	DEC at which the PSF is forced (decimal degrees)
x	x-pixel value at which the PSF is forced
y	y-pixel value at which the PSF is forced
maj	semi-major axis of the PSF fit from tphot (in pixels; 1.86" per pix)
min	semi-minor axis of the PSF fit from tphot (in pixels; 1.86" per pix)
phi	orientation of the PSF, measured counter-clockwise with respect to RA direction (CCD x-axis)
apfit	the aperture correction (in magnitudes) required by tphot
mag5sig	five-sigma limit magnitude
PA_DEG	fitted position angle between North and the detector Y-axis. Usually this will be a number close to zero or 180. If there is not a WCS for the exposure (and therefore no ZP_MAG or MAG5SIG), PA_DEG will contain exactly zero.
Sky	sky mag in 1 sq arcsec
Obs	the ATLAS data file on which the measurements were made

When the forced photometry runs in difference imaging mode, it is measuring the difference (initially in ADU) between the target and the reference frame. This may be significantly positive (excess flux in the target frame over the reference), negative (flux deficit in the target compared to the reference) or close to zero (no significant difference between the two). We calibrate this flux in microJanskys (with an error) and these values are the most meaningful when the user is interested in difference imaging measurements. The photometric calibration comes from the zeropoint in the input image which is based on stars from the ATLAS Refcat2 (Tonry et al. 2018).

An AB magnitude (m) and flux in microJanskys (uJy) are related simply by

m = -2.5*log10(uJy) + 23.9

The error on the magnitude can be determined from a Taylor expansion of log10(uJy +/- duJy) when duJy/uJy is small

The sign of the flux measurement and its error (duJy) are always statistically meaningful in the ATLAS forced photometry. The magnitude, m and its error dm, are of course not meaningful when the flux is small (below signal-to-noise <~ 3). In such cases, if a user wants a 3-sigma upper limit for the magnitude, use

m_3sig > -2.5*log10(3 * duJy) + 23.9

If the flux is negative, then a magnitude is still quoted and denoted with a negative sign. The flux is multiplied by -1 and all the above still applies.

The ATLAS filter bands and their throughput curves can be found in Tonry et al. (2018).

For plotting the output text files, David Young has provided a Python plot script.

If you notice any issues with the data or this website, please report an issue on GitHub or for urgent/security matters email Luke Shingles. This is an open-source volunteer project and feature requests may be considered as time allows.