[Paper Review] On Construction of ICRF-2

Zinovy Malkin|ArXiv.org|Nov 16, 2009

Nuclear Physics and Applications1 references3 citations

TL;DR

This paper proposes a two-tiered structure for ICRF-2, comprising a stable 400-source ICRF Core for system realization and a 3,600-source ICRF Extension for densification and calibration. It advocates selecting core sources based on long observation time spans and high session counts to ensure stability, while introducing the End User Error (EUE) and Index of Position Variability (IPV) to better reflect real-world position accuracy beyond nominal precision estimates.

ABSTRACT

In this paper, several issues are considered, related to the construction of the next ICRF generation, ICRF-2. Between them, the following points are touched: ICRF-2 structure, ICRF Core sources selection, and some expected user's requirements.

Motivation & Objective

• To define a robust, stable structure for the next ICRF realization, ICRF-2, to support long-term astrometric and geodetic applications.
• To address the uneven distribution of well-observed radio sources, especially the deficiency in the southern sky, by proposing a systematic source selection strategy.
• To improve user reliability by introducing new error metrics—End User Error (EUE) and Index of Position Variability (IPV)—that reflect actual position accuracy rather than just formal precision.
• To guide future VLBI observations, particularly in the southern hemisphere, by identifying gaps in source coverage and recommending dedicated sessions to complete the ICRF Supplement.

Proposed method

• Proposes a two-list structure: ICRF Core (400 sources) for system stability and ICRF Extension (3,600 sources) for densification and calibration.
• Uses time series analysis of source positions from IVS Analysis Centers, applying 2D velocity, WRMS, and WMADEV (weighted Allan deviation) to assess source stability.
• Prioritizes source selection based on observation time span and number of sessions, arguing these are more reliable indicators than scatter indices alone.
• Applies weighted least-squares estimates with reported uncertainties for position weighting in time series computations.
• Introduces the End User Error (EUE) concept, derived from time series scatter, to quantify real-world position accuracy for users.
• Proposes a continuous Index of Position Variability (IPV) as a dynamic alternative to the current discrete 1–3 classification system.

Experimental results

Research questions

• RQ1 How can the ICRF-2 structure be optimized to ensure long-term system stability and uniform sky coverage?
• RQ2 What criteria should be used to select ICRF-2 Core sources to minimize individual source instability and maximize system robustness?
• RQ3 To what extent do different analysis strategies affect the estimated velocity and scatter of radio sources, and how can this be mitigated in source selection?
• RQ4 How can user-relevant position accuracy be better represented than by formal precision errors alone?
• RQ5 What observational strategy is needed to complete the ICRF Supplement, especially in the southern sky with declination < -40°?

Key findings

• Significant discrepancies in source velocity estimates (e.g., 45 ± 38 μas/yr to 235 ± 50 μas/yr for 0656+082) across different analysis centers indicate that analysis strategy heavily influences stability metrics.
• Despite large time spans (~7.5 years) and many epochs, velocity estimates for the same source vary by factors of several, highlighting the unreliability of scatter indices when derived from inconsistent analysis methods.
• The number of high-quality, compact, and well-observed sources is insufficient to uniformly fill 100 deg² sky cells, especially in the southern hemisphere.
• A two-stage ICRF-2 release strategy is proposed: first release with the best available source per 100 deg² cell, followed by a second release after targeted observations (2009–2010) to improve low-quality candidates.
• The formal position uncertainty in ICRF-Ext.2 (e.g., 0.035 mas for 0923+392) is often underestimated; actual errors can reach up to 1 mas, indicating a need for improved error reporting.
• The proposed End User Error (EUE) and continuous Index of Position Variability (IPV) provide more realistic and user-relevant measures of position accuracy than current precision-based estimates.