Evaluate the ICRF3 Axes' Stability via Extragalactic Source Position Time Series

Sun, 01 Jan 2023 00:00:00 +0000

Monitoring the stability of the ICRF3 axes

The apparent positions of extragalactic radio sources can change as their intrinsic structures evolve. These variations can affect the orientation of a celestial reference frame, making regular monitoring of both individual sources and the frame axes necessary.

This study updates our earlier assessment of ICRF3 stability using 7,146 VLBI observing sessions spanning November 1979 to December 2021. The observations were processed with Calc/Solve to obtain position time series for 6,032 extragalactic sources, including all 303 ICRF3 defining sources.

Two complementary assessments

We assess the stability of the ICRF3 axes in two ways:

Long-term spin: Apparent source proper motions are derived from the position time series and fitted with first-degree vector spherical harmonics to estimate the rotation rate of the frame.
Annual orientation variations: Source positions are averaged within one-year windows to construct yearly realizations of ICRF3. Their orientation offsets reveal the temporal scatter of the frame axes.

We also compare solutions obtained by dividing the sources into 4, 8, 12, 16, and 20 subsets. Each subset is treated in a separate global solution to derive its source position time series.

Main findings

The different solutions consistently indicate an axes stability of approximately 10–20 microarcseconds over the full observing span. The annual orientation offsets in the recent observing interval have weighted root-mean-square values below 10 microarcseconds, showing no degradation following the adoption of ICRF3.

Although the formal fits suggest a small spin around the Y-axis, bootstrap resampling gives spin uncertainties of approximately $0.7,\mu\mathrm{as,yr^{-1}}$. This illustrates the importance of accounting for sample-dependent uncertainty when interpreting small apparent rotations.

The agreement between the different subset solutions confirms that the commonly used four-step method is sufficiently robust for this assessment.

Data and code

The source position time series and Jupyter notebooks used to reproduce the analysis are available in the public repository.

Is there an interest in bringing Gaia-CRF into VLBI data reduction?

Tue, 01 Sep 2020 00:00:00 +0000

Using an optical reference frame in VLBI analysis

VLBI and Gaia realize celestial reference frames in the radio and optical bands, respectively. Their different observing geometries offer an opportunity to investigate whether an optical reference frame can help constrain large-scale systematic errors in a radio reference frame.

This study explores the use of Gaia-CRF2 as a celestial reference datum in VLBI data reduction. We examine how the choice of a priori source positions and constraints affects the estimated celestial reference frame, Earth orientation parameters, and terrestrial reference frame.

Four global solutions

We analyse S/X-band VLBI observations from 1979 to 2019 using 250 sources common to the ICRF3 defining-source list and the ICRF3-prototype subset of Gaia DR2.

Four global solutions combine two choices of a priori catalogue with two treatments of defining-source positions:

Solution	A priori positions	Defining-source positions
A	Gaia DR2	Adjusted
B	ICRF3 S/X	Adjusted
C	Gaia DR2	Fixed
D	ICRF3 S/X	Fixed

For the solutions with adjusted positions, a no-net-rotation constraint maintains the orientation relative to the a priori reference frame.

Main findings

Changing the a priori catalogue mainly changes orientation when source positions are adjusted. Solutions A and B differ predominantly by a rotation, without significant glide terms. Both yield a post-fit delay RMS of 26.37 ps.
Earth orientation parameters reflect the frame rotation. Differences between A and B include approximately $2,\mu\mathrm{s}$ in UT1 and offsets of about $80,\mu\mathrm{as}$ and $60,\mu\mathrm{as}$ in the celestial pole coordinates $\mathrm{d}X$ and $\mathrm{d}Y$, respectively.
Fixing source positions also affects dipolar deformation. Relative to fixing positions to ICRF3, fixing them to Gaia DR2 reduces the Y and Z glide components with respect to Gaia, while increasing the X component.
The fit to VLBI observations becomes poorer when positions are fixed to Gaia DR2. The post-fit delay RMS rises to 28.03 ps, compared with 26.44 ps when positions are fixed to ICRF3.
The terrestrial frame is only weakly affected. Station-position differences between A and B are below 0.1 mm. Differences between C and D are approximately 0.5 mm in position and $0.1,\mathrm{mm,yr^{-1}}$ in velocity.

Implications and limitations

The experiment demonstrates that changing the a priori catalogue alone does not remove zonal errors when source positions remain adjustable. Tighter constraints using Gaia positions can alter the large-scale deformation of the estimated radio frame, but also introduce a trade-off in the fit to VLBI observations.

These preliminary results do not establish Gaia-CRF2 as a superior reference datum for VLBI analysis. The potential benefit depends on the accuracy of the optical positions, their systematic errors, and radio–optical offsets. The paper therefore calls for further tests with subsequent Gaia releases.

Conference contribution

This contribution appears on pages 9–14 of Astrometry, Earth Rotation, and Reference Systems in the GAIA era, the Journées 2019 proceedings edited by Christian Bizouard and published in 2020. The PDF link above opens the complete proceedings.

ICRF3 | NJU Astrometry Group

Evaluate the ICRF3 Axes' Stability via Extragalactic Source Position Time Series

Monitoring the stability of the ICRF3 axes

Two complementary assessments

Main findings

Data and code

Is there an interest in bringing Gaia-CRF into VLBI data reduction?

Using an optical reference frame in VLBI analysis

Four global solutions

Main findings

Implications and limitations

Conference contribution