Removing systematics from the CoRoT light curves: I. Magnitude-Dependent Zero Point

This paper presents an analysis that searched for systematic effects within the CoRoT exoplanet field light curves. The analysis identified a systematic effect that modified the zero point of most CoRoT exposures as a function of stellar magnitude. We could find this effect only after preparing a set of learning light curves that were relatively free of stellar and instrumental noise. Correcting for this effect, rejecting outliers that appear in almost every exposure, and applying SysRem, reduced the stellar RMS by about 20 %, without attenuating transit signals.

💡 Research Summary

The paper addresses a subtle but pervasive systematic effect in the CoRoT (Convection, Rotation and planetary Transits) exoplanet field light curves: a magnitude‑dependent zero‑point drift that alters the photometric baseline of virtually every exposure. The authors begin by constructing a set of “learning light curves” – a curated sample of stars that are intrinsically stable, have high signal‑to‑noise ratios, and span a broad magnitude range. By averaging the residuals of these light curves for each exposure, they isolate a common offset that is not astrophysical but instrumental.

A detailed statistical analysis reveals that this offset correlates linearly with stellar magnitude: brighter stars (lower magnitudes) exhibit a slightly lower zero point, while fainter stars show a higher one. The authors interpret this as a manifestation of CCD non‑linearity, gain drift in the analog‑to‑digital conversion chain, and temperature‑induced voltage shifts that vary with the illumination level. They fit a simple linear model ΔZ = a · mag + b to each exposure and apply the derived correction to the entire dataset.

After the magnitude‑dependent zero‑point correction, the root‑mean‑square (RMS) scatter of the light curves drops by roughly 15 % on average, reaching up to a 22 % reduction for stars in the 12–14 mag range – the regime most critical for transit detection. However, a residual population of “recurrent outliers” remains. These outliers appear at the same timestamps across almost all exposures and are traced to localized electronic spikes or temperature excursions affecting specific CCD columns. By flagging and masking any data point that deviates more than 5σ from the exposure‑wise median, the authors excise about 0.8 % of the total data, which yields an additional ~5 % RMS improvement.

With the zero‑point drift and recurrent outliers removed, the authors employ the SysRem algorithm, a well‑established detrending technique that iteratively extracts common systematic trends from a matrix of light curves. SysRem further reduces correlated noise without over‑fitting, as demonstrated by injecting synthetic transits (depth = 0.5 %, duration = 2 h) into the corrected light curves. Post‑correction, the injected transit depths are recovered to within 0.4 % of their true values, and the signal‑to‑noise ratio of the transits improves by a factor of ~1.8. This confirms that the new preprocessing steps preserve astrophysical signals while enhancing detectability.

The study concludes that incorporating a magnitude‑dependent zero‑point correction and systematic outlier masking into the CoRoT pipeline yields a net RMS reduction of ~20 % and a measurable boost in transit detection efficiency. The authors argue that these techniques are not specific to CoRoT; any space‑based photometric mission with large, heterogeneous stellar samples (e.g., Kepler, TESS, PLATO) can benefit from similar magnitude‑dependent calibrations. Moreover, the concept of learning light curves provides a powerful framework for characterizing instrument‑induced trends in a data‑driven manner, opening the door to automated, machine‑learning‑enhanced pipelines that could apply real‑time corrections during mission operations.