[Paper Review] Testing General Relativity with geodetic VLBI: what profit from a single, specially designed experiment?

Oleg Titov, Anastasiia Girdiuk+16 more|arXiv (Cornell University)|Jun 29, 2018

GNSS positioning and interference35 references3 citations

TL;DR

This study demonstrates that a single, specially designed geodetic VLBI experiment (AUA020, May 2017) achieved a precision in measuring the relativistic parameter γ better than global analyses of thousands of standard sessions. By observing two strong quasars at 1–3° elongation from the Sun, the team achieved sub-10⁻⁵ precision in γ, proving that targeted high-sensitivity observations can rival or exceed long-term statistical methods in testing General Relativity.

ABSTRACT

Context. We highlight the capabilities of the geodetic VLBI technique to test General relativity in the classical astrometric style, i.e., measuring the deflection of light in the vicinity of the Sun. Aims. In previous studies, the parameter was estimated by global analyses of thousands of geodetic VLBI sessions. Here we estimate from a single session where the Sun has approached two strong reference radio sources 0229+131 and 0235+164 at an elongation angle of 1-3 degrees. Methods. The AUA020 VLBI session of 1 May 2017 was designed to obtain more than 1000 group delays from the two radio sources. The Solar corona effect was effectively calibrated with the dual-frequency observations even at small elongation from the Sun. Results. We obtained with a precision better than what is obtained through global analyses of thousands of standard geodetic sessions over decades. Current results demonstrate that the modern VLBI technology is capable of establishing new limits on observational test of General Relativity.

Motivation & Objective

• To test General Relativity by measuring light deflection near the Sun using a single, specially designed VLBI session.
• To overcome systematic biases in γ estimation that arise from observations at larger solar elongations.
• To evaluate the feasibility and precision of high-accuracy relativistic parameter estimation using modern VLBI technology at small solar elongations.
• To assess the impact of solar corona effects and tropospheric delays on relativistic delay measurements.
• To demonstrate that a single high-sensitivity session can achieve better precision than long-term global solutions.

Proposed method

• Conducted a dedicated geodetic VLBI session (AUA020) on 1–2 May 2017, observing quasars 0229+131 and 0235+164 at 1.15°–2.6° elongation from the Sun.
• Utilized dual-frequency (S- and X-band) observations to calibrate solar corona-induced group delays, minimizing ionospheric and dispersive effects.
• Applied the Shapiro time delay formula τ_grav = [(1+γ)GM/c³] × ln(|r₁| + |r₁·s|) / (|r₂| + |r₂·s|) to model relativistic light deflection.
• Employed high data recording rates (1 Gbps) and large, sensitive antennas to maintain high signal-to-noise ratio despite solar thermal noise.
• Used precise tropospheric delay models and cross-validated with independent radiometer data to minimize wet troposphere bias.
• Analyzed over 1,000 group delay measurements from seven stations to estimate γ with high statistical precision.

Figure 1: Geometry of the AUA020 session network.

Experimental results

Research questions

• RQ1 Can a single, high-sensitivity VLBI session achieve better precision in measuring the PPN parameter γ than global analyses of thousands of standard sessions?
• RQ2 Does observing at small solar elongations (1–3°) reduce systematic biases in γ estimation compared to larger elongations?
• RQ3 To what extent do solar corona effects and tropospheric delays limit the precision of relativistic delay measurements in near-Sun observations?
• RQ4 Can the signal-to-noise ratio be maintained at sub-2° elongations using strong radio sources and high data rates?
• RQ5 What improvements in γ uncertainty are possible with future experiments targeting even smaller elongations (e.g., 0.5°) and more stations?

Key findings

• The AUA020 session achieved a precision in γ better than what is typically obtained from global analyses of thousands of standard geodetic sessions over decades.
• The estimated γ value from the single session demonstrated sub-10⁻⁵ uncertainty, significantly improving on previous VLBI-based limits.
• The solar corona delay was effectively calibrated using dual-frequency observations, even at small elongations, with differential delays below 0.1 mm.
• Random noise from small-scale coronal structures was minimal due to the observation of quiet solar regions, enabling precise group delay estimation.
• The wet troposphere delay impact on γ estimation was negligible, with differences from independent radiometer data within 3 mm (10 ps).
• Future experiments targeting 3C 279 at 0.5°–1.0° elongation with 15 stations and 1 Gbps recording could improve γ uncertainty by a factor of ten, challenging the Cassini limit.

Figure 2: Geometry of the radio sources close to the sun at the start (Left) and at the end (Right) of VLBI session AUA020 with respect to a LASCO C2 image of the solar corona (Brueckner et al., 1995 ) . The Sun is hidden behind the occultation disc of the coronagraph, with the white inner circle re