Robust 1-norm periodograms for analysis of noisy non-Gaussian time series with irregular cadences: Application to VLBI astrometry of quasars

Valeri V. Makarov,Sébastien Lambert,Phil Cigan,Christopher DiLullo,David Gordon

2024-05-21

Abstract:Astronomical time series often have non-uniform sampling in time, or irregular cadences, with long gaps separating clusters of observations. Some of these data sets are also explicitly non-Gaussian with respect to the expected model fit, or the simple mean. The standard Lomb-Scargle periodogram is based on the least squares solution for a set of test periods and, therefore, is easily corrupted by a subset of statistical outliers or an intrinsically non-Gaussian population. It can produce completely misleading results for heavy-tailed distribution of residuals. We propose a robust 1-norm periodogram technique, which is based on the principles of robust statistical estimation. This technique can be implemented in weighted or unweighted options. The method is described in detail and compared with the classical least squares periodogram on a set of astrometric VLBI measurements of the ICRF quasar IERS B0642+449. It is uniformly applied to a collection of 259 ICRF3 quasars each with more than 200 epoch VLBI measurements, resulting in a list of 49 objects with quasi-periodic position changes above the $3\sigma$ level, which warrant further investigation.

Instrumentation and Methods for Astrophysics

What problem does this paper attempt to address?

### What problems does this paper attempt to solve?

This paper aims to solve several key problems encountered in the analysis of astronomical time - series data. In particular, when these data have irregular time sampling (irregular cadences), non - Gaussian distribution, and a large number of statistical outliers, how to detect and characterize periodic signals more accurately. Specifically:

1. **The problem of irregular sampling**:
- Astronomical time - series data are usually not uniformly sampled but consist of irregular time intervals, which makes the traditional Fourier transform method no longer applicable.
- The paper proposes a robust periodogram method based on the 1 - norm to deal with this irregular sampling.

2. **The problem of non - Gaussian noise**:
- Many astronomical data sets not only contain random noise but may also exhibit non - Gaussian distribution characteristics, such as heavy - tailed distribution, which means that there are a large number of observation points deviating from the average value in the data.
- The standard Lomb - Scargle periodogram method depends on the least squares solution and is easily disturbed by these outliers or non - Gaussian distribution data points, leading to misleading results.
- The paper introduces the 1 - norm periodogram technique based on robust statistical estimation to reduce the influence of outliers and improve the accuracy of periodic signal detection.

3. **The problem of large - scale data processing**:
- The research involves processing a large amount of single - epoch astrometric data from global Very Long Baseline Interferometry (VLBI) facilities, which cover the position information of thousands of radio sources.
- The proposed method needs to be able to efficiently process such large - scale data sets and ensure the reliability and accuracy of the results.

4. **Application background**:
- The paper pays special attention to the periodic changes in the positions of quasars in the International Celestial Reference Frame (ICRF3), which may be caused by binary black hole systems or other complex structures.
- Through the analysis of 259 ICRF3 quasars, 49 celestial bodies with significant quasi - periodic position changes were found, and these findings provide important clues for follow - up research.

### Method overview

To achieve the above goals, the paper proposes the following improvement measures:

- **1 - norm periodogram**: Construct a periodogram by minimizing the sum of the absolute values of residuals (instead of the sum of squares) to reduce the influence of outliers.
- **Weighted and unweighted options**: Consider the differences in formal errors of different observation points and provide two processing methods: weighted and unweighted.
- **Monte Carlo simulation**: Used to numerically calculate the covariance matrix of periodogram coefficients to ensure the statistical reliability of the results.
- **Global nonlinear optimization methods**: Such as Nelder - Mead, differential evolution or simulated annealing algorithms are used to solve nonlinear optimization problems.

In short, this paper is committed to developing a more robust periodogram analysis method to deal with the complex features in astronomical time - series data, especially when dealing with non - Gaussian noise and irregular sampling.