The significance of O-C histograms in lunar occultation studies
The Observed minus Calculated (O-C) histogram derived from a set of lunar occultation observations is a valuable diagnostic tool. It reflects several key factors that influence the accuracy and reliability of the measurements. For the histogram to be meaningful, a sufficiently large number of observations is essential. However, even a few dozen occultations of single stars (i.e., not binary or multiple systems) can already highlight potential systematic errors in the observational setup, timing accuracy, data interpretation, or even the catalogued stellar positions.
Ideally, the O-C value for a single star should be zero, provided that accurate Gaia DR3 coordinates and proper motions are used in the calculation. Deviations from zero (i.e. nonzero O-C values) may indicate the presence of systematic or statistical errors.
Sources of systematic error
- Inaccurate time base.
The most common source of systematic error is an inaccurate timing reference, specifically, timestamps on recorded frames that are shifted relative to the true Coordinated Universal Time (UTC). This causes the Moon's calculated position to be displaced from its actual sky position at the time of occultation. Since the O-C value represents the apparent angular separation between the star and the lunar limb at the moment of occultation, even a small timing shift leads to a measurable angular error. Given the Moon's apparent motion of approximately 0.4 arcseconds per second, a time error Δt of ±10 milliseconds will translate to an O-C error of roughly ±4 milliarcseconds (mas).
-
Ephemeris or lunar limb errors.
Modern lunar ephemerides are highly accurate, and any associated uncertainties are generally smaller than observational errors. Thus, inaccuracies in the Moon's predicted position or the lunar limb profile are unlikely to significantly affect O-C values under current standards.
-
Incorrect stellar catalogue position.
In some cases, the apparent error arises not from the observation but from the incomplete or inaccurate catalogued position of the star. While Gaia DR3 provides high-precision positions and proper motions, it may not account for accelerations in stars that are part of binary or multiple systems. These deviations can be exploited positively: a displaced O-C value may reveal the true current position of the star, offering insight into its motion and interactions.
-
Incorrect geodetic position of observing station. Incorrect longitude, latitude and elevation will lead to systematic errors. Use of GPS receivers minimises errors associated to terrestrial position.
Sources of statistical error
Another major source of error lies in the determination of contact times from light curves. Even with an accurate UTC time base, a low signal-to-noise ratio or insufficient temporal resolution can lead to large uncertainties. For example, in a low-quality light curve sampled at coarse intervals, the inferred contact time may be off by 50 ms or more, corresponding to an O-C error exceeding 20 mas, using the same velocity conversion as above.
Interpreting the O-C Histogram
If only statistical errors were present, the O-C histogram derived from a sufficiently large dataset should follow a normal (Gaussian) distribution. This is because O-C values, being influenced by numerous small random factors, would be expected to conform to the central limit theorem. Moreover, the contribution of random stellar position errors should average out, maintaining a symmetric distribution centred around zero.
However, a nonzero mean or a peak displaced from zero in the histogram indicates a systematic bias. This could be caused by miscalibrated timing, incorrect interpretation of timestamps, or errors in handling the lunar ephemeris. Monitoring the O-C distribution for a specific observer or group of observers, focusing on occultations of single stars, is a powerful method for diagnosing such systematic issues.
Conclusion
Drawing robust conclusions from a small number of observations is inherently risky. This is why it is crucial to collect as many high-quality occultation events of single stars as possible. A well-populated O-C histogram provides not only statistical insights but also helps uncover deeper systematic issues in observational procedures and data analysis.
We strongly encourage observers to systematically observe single-star occultations and submit their data to the coordinator.
Single stars are especially valuable for analysis, as they provide cleaner, more interpretable results compared to binary or multiple systems (whose observation is highly valuable but in a different context!)
If you require predictions of single-star occultations for your observing site, please navigate to the PROGRAMMES section and access the Lunar Occultation Server: Select "DETAILED PREDICTIONS FOR A SPECIFIC STATION (Form 1)" and choose 'PPM' as the catalogue to obtain the relevant data.
E. Velasco
Last revised 21 April 2025