Pith. sign in

REVIEW 3 major objections 5 minor 32 references

Detection statistics of the RadioAstron AGN survey

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read A space-VLBI survey finds that two-thirds of bright AGN contain ultra-compact cores with brightness temperatures exceeding the inverse-Compton limit.

desk verdict Solid survey detection statistics let down by an overbroad abstract 'two-thirds' claim and missing uncertainty bars, but the core result stands. read the letter →

arxiv 1909.00785 v1 pith:IWXHS43I submitted 2019-09-02 astro-ph.GA astro-ph.HE

classification astro-ph.GAastro-ph.HE
keywords activegalacticnucleiquasarsgalaxies:jetsradiocontinuum:galaxiesspaceVLBIbrightnesstemperatureinterstellarscattering
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper reports detection statistics from the RadioAstron space-VLBI survey of a complete, flux-density-limited sample of 163 radio-strong AGN at 18, 6, and 1.3 cm. The central result is that about two-thirds of the observed targets produce significant interferometric fringes on ground-space baselines, implying that many AGN cores contain extremely compact, tens-to-hundreds-of-microarcsecond structures of very bright synchrotron emission. If this holds for the whole sample, the measured brightness temperatures exceed the inverse-Compton (Compton catastrophe) limit and sit far above equipartition for most detected sources, which directly challenges standard jet emission models and bears on the 'Doppler factor crisis' from gamma-ray observations. The paper also presents detection fraction versus projected baseline and attributes an excess of 18-cm detections at the longest baselines to interstellar scattering substructure.

What carries the argument

The central instrument is the ground-space interferometer RadioAstron, whose baselines extend to 28 Earth diameters and whose detection sensitivity reaches about 6 mJy at 18 and 6 cm and 60 mJy at 1.3 cm with the largest ground telescopes. The load-bearing observable is a fringe detection: a source counts as detected when the probability of a false detection is below 0.01%, with the signal-to-noise-to-probability calibration derived by fitting a theoretical noise distribution to the low-SNR part of the empirical fringe-SNR distribution. Detection fraction versus projected baseline then maps directly to the angular-size and brightness-temperature distribution of AGN cores, because the maximum measurable brightness temperature depends on baseline length and fringe-visibility accuracy rather than on wavelength.

What would settle it

Re-observing the roughly 16 unobserved low-declination targets with a southern tracking station and uniform short-baseline scheduling would settle the matter: if the detection fraction on baselines beyond a few Earth diameters among these sources falls below about one-third, the claim that two-thirds of the complete sample contains ultra-compact cores would not survive.

Watch

Extended reading notes

Core claim

The paper establishes that two-thirds of the observed complete sample are detected on space VLBI baselines: 30 of 108 sources at 1.3 cm, 95 of 147 at 6 cm, and 84 of 145 at 18 cm, with detection defined by a probability of false detection below 0.01%. The detected sources therefore contain structures with angular sizes of tens to hundreds of microarcseconds and brightness temperatures significantly exceeding the Compton catastrophe limit, with most far above the equipartition value. The fractional detection as a function of projected baseline resembles earlier 5-GHz VSOP and 2-cm VLBA survey curves, consistent with core-jet structure probed at smaller scales. An excess of 18-cm detections at baselines of about 25 Earth diameters or longer is interpreted as evidence of scattering substructure, while interstellar scattering is only weak at 6 cm and absent at 1.3 cm for the typical sightlines sampled.

Load-bearing premise

The two-thirds detection fraction is assumed to represent the complete 163-source sample even though about 10% of targets, mostly low-declination sources, had not been observed by June 2016 and short-baseline K/L-band observations were under-scheduled; if those missing or under-scheduled sources are systematically fainter or more extended, the sample-level prevalence of ultra-compact cores is overestimated.

Editorial extensions

If this is right

  • If two-thirds of radio-strong AGN contain such compact cores, jet emission models must accommodate sustained brightness temperatures above equipartition and often above the inverse-Compton limit.
  • Detections at baselines of about 25 Earth diameters at 18 cm and 15 Earth diameters at 1.3 cm set upper limits of tens of microarcseconds on the angular size of the emitting regions.
  • The excess of 18-cm detections at the longest baselines implies that interstellar scattering substructure, not intrinsic core structure, dominates the most extreme long-baseline detections at that wavelength.
  • The weak scattering at 6 cm and its absence at 1.3 cm mean that brightness-temperature measurements at those bands are largely free of scattering bias for typical mid-Galactic-latitude sightlines.
  • The similarity of the detection histograms to earlier ground and space VLBI surveys suggests a single continuous core-jet brightness distribution from ground to space baselines, with sensitivity rather than physics setting the difference.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • My inference: if the two-thirds detection fraction extends to the unobserved low-declination sources, the number of AGN cores with brightness temperatures above $10^{13}$ K is large enough that the apparent 'Doppler factor crisis' could be resolved by very high Doppler factors in a substantial fraction of jets.
  • My inference: the short-baseline scheduling bias means the true detection fraction at baselines below 2 Earth diameters is probably higher than plotted, so the two-thirds figure is more likely a lower bound on the prevalence of ultra-compact cores.
  • My inference: the same PFD-to-SNR calibration could be applied to multi-epoch observations of individual sources to test whether the ultra-compact component varies, connecting the survey to intraday-variability and scintillation studies.
  • My inference: combining the detection-fraction curves with core-size estimates from scattering should let future space-VLBI missions predict how many AGN will be detectable at baselines beyond 28 Earth diameters.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper presents detection statistics for the RadioAstron AGN survey, a space VLBI Key Science Program targeting a complete VLBI-flux-density-limited sample of 163 radio-strong AGN at 18, 6, and 1.3 cm. Using a PFD threshold of less than 0.01% and an SNR-to-PFD calibration derived from a fit to the low-SNR part of the fringe-SNR distribution, the authors report detection fractions as a function of projected ground-space baseline. They find that about two-thirds of the observed complete sample are detected on space VLBI baselines, with many detected sources showing brightness temperatures above the Compton catastrophe limit, and they note a possible excess of 18 cm detections at the longest baselines, attributed to interstellar scattering substructure.

Significance. If the headline result is stated accurately, this is an important empirical contribution: it demonstrates on a well-defined flux-limited sample that a large fraction of AGN cores contain ultra-compact, very bright structures on space-VLBI scales, directly constraining the prevalence of brightness temperatures in excess of the Compton catastrophe and equipartition limits. The paper's strengths include the sample definition, the multi-band comparison, the explicit false-detection threshold, and the transparent description of the SNR-PFD calibration procedure. The detection counts themselves are straightforward and the qualitative conclusion that many AGN cores are detected on space baselines is robust. However, the precise population-level claim of 'two-thirds' needs to be qualified and quantified before the result can be taken at face value, and the absence of confidence intervals weakens the baseline-dependent comparisons.

major comments (3)
  1. [Abstract and Section 3] The headline 'two-thirds' claim is not stated precisely enough. The abstract says 'Two-thirds of them are indeed detected' after referring to the complete sample of 163 sources, and the Summary repeats 'Two thirds of the targets.' But the body of Section 3 reports 'about two thirds of the observed complete sample,' and the only number matching this in Figure 2 is the C-band result: 95 detections out of 147 sources (64.6%). The K-band and L-band fractions are 30/108 (27.8%) and 84/145 (57.9%), respectively. This is a load-bearing distinction because the paper's central claim is about the prevalence of ultra-compact cores in the complete sample. The abstract and Summary should explicitly state that the two-thirds figure refers to the C-band observed subset, and the corresponding K- and L-band fractions should be reported in the same places.
  2. [Section 2 and Section 3] The manuscript acknowledges that about 10% of the complete sample had not been observed by June 2016 and that these are predominantly low-declination sources, but it does not account for this in the detection statistics. Because scheduling was driven by declination, tracking-station visibility, and ground-telescope constraints, the unobserved sources cannot be assumed exchangeable with the observed subset without justification. The paper should provide a sensitivity analysis or at least a worst-case bound: for the C-band subset, if all 16 unobserved sources are assumed to be non-detections, the fraction drops from 95/147 (64.6%) to 95/163 (58.3%). This bound does not overturn the qualitative conclusion that a large fraction of cores are detected, but it means the precise 'two-thirds' wording is not supported for the full sample, and the manuscript should either add such a bound or restrict all prevalence statements to the observed sample.
  3. [Section 3, Figure 2] Figure 2 displays detection fractions versus projected baseline without confidence intervals, per-bin source counts, or a significance test for the differences between bands. This is particularly important for the two secondary claims in this section: the apparent drop at short baselines (which the text attributes to an observational bias) and the 'possible excess of fractional detections at the longest RadioAstron projected baselines at 18 cm' relative to 6 cm, which is then interpreted as evidence for scattering substructure. Without binomial confidence intervals or a statistical test, the reader cannot judge whether the reported excess is significant or merely noise. At minimum, the figure should include error bars or shaded confidence bands, and the claim about the 18 cm excess should be accompanied by a significance estimate.
minor comments (5)
  1. [Section 2] The text says 'About 10% of the complete sample had not been observed by June 2016,' but the reader is not told the exact per-band numbers of observed sources or whether the same 163 sources are included in each band's denominator. Please clarify that the denominators in Figure 2 (108, 147, 145) are the observed subsets in each band and give the overlap between bands.
  2. [Section 3] The sentence 'RadioAstron has delivered detections in just over one third of the observing segments' is ambiguous because the paper elsewhere discusses detected sources, not segments. Please define what an 'observing segment' is and state how this segment-level detection rate relates to the source-level detection fractions in Figure 2.
  3. [Section 3, Figure 4] The description of the SNR-PFD calibration says the theoretical distribution is fitted to the low-SNR peak and that the parameters depend on the number of spectral channels, correlator integration time, and scan length. It would be helpful to state how many distinct parameter sets were used and how the resulting PFD-SNR mapping varies across them, since only one example is shown.
  4. [Section 3] The phrase 'the apparent drop in the detection fraction at 0 to 2 Earth Diameters ... can be observed in the K- and L-band histograms' is presented as an observational bias, but no quantitative comparison is made with the C-band first bin, which is called 'unbiased.' A short quantitative justification, even a statement of the number of scheduled scans in that bin, would strengthen the claim.
  5. [Summary] The Summary states 'Two thirds of the targets have delivered significant interferometric fringes at space VLBI baselines,' but the body text consistently says 'the observed complete sample.' Please align the Summary with the qualified wording used in Section 3.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the detection statistics are empirical counts, and the SNR-PFD calibration is standard threshold fitting that does not force the reported detection fraction.

full rationale

The paper's central result is a set of empirical detection fractions (95/147 at C-band, 84/145 at L-band, 30/108 at K-band) computed from space-VLBI fringe detections. No derivation reduces to its inputs. The only calibration step is the SNR-to-PFD mapping in Section 3 and Fig. 4, where a theoretical noise distribution is fitted to the low-SNR peak to assign false-detection probabilities; this is standard practice and does not determine whether sources are detected at high SNR or the reported two-thirds fraction. The 'About two thirds of the observed complete sample are detected' statement in Section 3 and the abstract's 'Two-thirds of them' are direct counts, not fitted outputs. Self-citations (Kovalev et al. 2005, 2016; Petrov et al. 2011; etc.) provide methodology and context but are not used to justify the detection fraction. The acknowledged 10% unobserved low-declination sources and short-baseline under-scheduling are possible selection effects, but that is a sampling and representativeness concern, not circularity. No circular step can be exhibited from the paper's own equations or construction.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

This is an observational survey. The central claim rests on the survey data themselves, on standard VLBI calibration theory, on an external catalog for sample completeness, and on prior physical models for interpretation. Free parameters are limited to the detection threshold and the SNR-PFD calibration; neither is adjusted to force the two-thirds fraction. No new physical entities are introduced.

free parameters (2)
  • False-detection threshold = PFD < 0.0001
    Chosen significance level for declaring a detection (Section 3). The detection counts and derived fractions depend on this threshold; it is a hand-set criterion, not fitted to the science result.
  • SNR-PFD calibration fit parameters = not reported
    The theoretical noise distribution is fitted to the low-SNR peak of the empirical SNR distribution (Section 3, Fig. 4) to convert SNR to PFD. The fitted parameters are not tabulated; this calibration is standard but is a data-dependent step.
assumptions (4)
  • standard math The theoretical fringe SNR distribution for the no-signal case (Thompson et al. 2017) correctly describes the low-SNR peak.
    Used in Section 3 to derive the PFD-SNR mapping; if the noise model is wrong, the detection threshold calibration shifts.
  • domain assumption The Radio Fundamental Catalog 2012 defines a complete VLBI-flux-density-limited sample at 8 GHz with correlated flux densities above 600 mJy on baselines longer than 200 Mlambda.
    The sample definition in Section 2 uses this external catalog as the completeness basis; the survey statistics inherit its completeness properties.
  • domain assumption Synchrotron radiation and the inverse Compton and equipartition brightness temperature limits from Kellermann and Pauliny-Toth (1969) and Readhead (1994) are valid physical benchmarks.
    Used in Sections 1 and 3 to interpret high brightness temperatures as physically significant.
  • domain assumption The interstellar scattering substructure model of Johnson and Gwinn (2015) explains the excess of long-baseline 18 cm detections.
    Invoked in Section 3 as the interpretation of the 18 cm excess; the survey itself does not directly measure scattering substructure.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Detection statistics of the RadioAstron AGN survey." pith.science (2026). https://pith.science/paper/IWXHS43I

@misc{pith2026190900785,
  author       = {Pith},
  title        = {Pith review of: Detection statistics of the RadioAstron AGN survey},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IWXHS43I}},
  note         = {Machine review of arXiv:1909.00785}
}
abstract

The largest Key Science Program of the RadioAstron space VLBI mission is a survey of active galactic nuclei (AGN). The main goal of the survey is to measure and study the brightness of AGN cores in order to better understand the physics of their emission while taking interstellar scattering into consideration. In this paper we present detection statistics for observations on ground-space baselines of a complete sample of radio-strong AGN at the wavelengths of 18, 6, and 1.3 cm. Two-thirds of them are indeed detected by RadioAstron and are found to contain extremely compact, tens to hundreds of $\mu$as structures within their cores.

Figures

Figures reproduced from arXiv: 1909.00785 by the authors.

Figure 1
Figure 1. Redshift distribution of the complete VLBI-flux-density limited sample of 163 [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Fraction of detected sources versus projected ground-space VLBI baselines [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Examples of RadioAstron interferometric fringes, C-band observations of 0716+714. Left panel: Detection with PFD ă 10´100 at projected baseline 7.0 ED, SRT–Effelsberg, SNR=138; middle panel: detection with PFD “ 10´8 at projected base￾line 21.9 ED, SRT–Effelsberg, SNR=7.3; Right panel: non-detection with PFD “ 0.07 at projected baseline 22.0 ED, SRT–Noto, SNR=5.3 3. Space VLBI data analysis and detection results The… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Low SNR part of the empirical distribution for the fringe SNR from the results [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 16 canonical work pages

  1. [3]

    ApJ 658, 232–244

    Relativistic Beaming and the Intrinsic Properties of Extragalactic Radio Jets. ApJ 658, 232–244. doi:10.1086/511063, arXiv:astro-ph/0611642. Dodson, R., Fomalont, E.B., Wiik, K., et al.,

  2. [8]

    Discovery of Substructure in the Scatter-Broadened Image of Sgr A*

    Discov- ery of Substructure in the Scatter-broadened Image of Sgr A*. ApJL 794, L14. doi: 10.1088/2041-8205/794/1/L14, arXiv:1409.0530. Gwinn, C.R., Popov, M.V., Bartel, N., et al.,

  3. [12]

    14 Liu, J., Bignall, H., Krichbaum, T., et al.,

    doi: 10.3847/0004-6256/152/1/12, arXiv:1603.03882. 14 Liu, J., Bignall, H., Krichbaum, T., et al.,

  4. [16]

    Cosmic Research 52, 393–402

    The RadioAstron project: Measurements and analysis of basic parameters of space tele- scope in flight in 2011-2013. Cosmic Research 52, 393–402. doi: 10.1134/ S0010952514050074. Kovalev, Y.Y., Kardashev, N.S., Kellermann, K.I., et al.,

  5. [17]

    ApJL 820, L9

    RadioAs- tron Observations of the Quasar 3C273: A Challenge to the Brightness Temperature Limit. ApJL 820, L9. doi: 10.3847/2041-8205/820/1/L9, arXiv:1601.05806. Kovalev, Y.Y., Kellermann, K.I., Lister, M.L., et al.,

  6. [20]

    Journal of Astronomical Instrumentation 6, 1750004–131

    Software Correla- tor for Radioastron Mission. Journal of Astronomical Instrumentation 6, 1750004–131. doi: 10.1142/S2251171717500040, arXiv:1706.06320. Linfield, R.P., Levy, G.S., Edwards, C.D., et al.,

  7. [27]

    The Micro- Arcsecond Scintillation-Induced Variability (MASIV) Survey. II. The First Four Epochs. ApJ 689, 108–126. doi: 10.1086/592485, arXiv:0808.1140. Murphy, D.W.,

  8. [29]

    The high brightness temperature of B0529+483 revealed by RadioAstron and implications for interstellar scattering

    The high brightness temperature of B0529+483 revealed by RadioAstron and impli- cations for interstellar scattering. MNRAS 474, 3523–3534. doi: 10.1093/ mnras/stx2991, arXiv:1711.06713. Popov, M.V., Bartel, N., Gwinn, C.R., et al.,

Show all 32 references
  1. [30]

    MNRAS 465, 978–985

    PSR B0329+54: sub- structure in the scatter-broadened image discovered with RadioAstron on baselines up to 330 000 km. MNRAS 465, 978–985. doi: 10.1093/mnras/ stw2353, arXiv:1609.04008. Readhead, A.C.S.,

  2. [35]

    Pilipenko, S.V., Kovalev, Y.Y., Andrianov, A.S., et al.,

    doi:10.1088/ 0004-6256/142/2/35, arXiv:1101.1460. Pilipenko, S.V., Kovalev, Y.Y., Andrianov, A.S., et al.,

  3. [49]

    Lovell, J.E.J., Rickett, B.J., Macquart, J.P., et al.,

    doi: 10.3390/galaxies6020049, arXiv:1804.09289. Lovell, J.E.J., Rickett, B.J., Macquart, J.P., et al.,

  4. [96]

    12 Horiuchi, S., Fomalont, E.B., Taylor, W.K., et al.,

    doi:10.3847/0004-637X/822/ 2/96, arXiv:1501.04449. 12 Horiuchi, S., Fomalont, E.B., Taylor, W.K., et al.,

  5. [104]

    Kardashev, N.S., Khartov, V.V., Abramov, V.V., et al.,

    doi:10.3847/1538-4357/aadcff, arXiv:1808.08966. Kardashev, N.S., Khartov, V.V., Abramov, V.V., et al.,

  6. [180]

    1088/0004-637X/805/2/180, arXiv:1502.05722

    doi:10. 1088/0004-637X/805/2/180, arXiv:1502.05722. Johnson, M.D., Kovalev, Y.Y., Gwinn, C.R., et al.,

  7. [273]

    doi: 10.1093/mnras/stx710, arXiv:1703.07976

    MNRAS 468, 4478–4493. doi: 10.1093/mnras/stx710, arXiv:1703.07976. Lister, M.L.,

  8. [501]

    doi: 10.1086/ 521382, arXiv:astro-ph/0702008

    ApJ 669, 862–883. doi: 10.1086/ 521382, arXiv:astro-ph/0702008. Cohen, M.H., Lister, M.L., Homan, D.C., et al.,

  9. [1969]

    ApJL 155, L71–L78

    The Spectra of Opaque Radio Sources. ApJL 155, L71–L78. doi: 10.1086/180305. Kovalev, Y.A., Vasil’kov, V.I., Popov, M.V., et al.,

  10. [1989]

    II - Brightness temperatures exceeding the inverse Compton limit

    VLBI using a telescope in Earth orbit. II - Brightness temperatures exceeding the inverse Compton limit. ApJ 336, 1105–1112. doi: 10.1086/167081. Lipunov, V., Kornilov, V., Gorbovskoy, E., et al.,

  11. [1990]

    ApJ 358, 350–358

    15 GHz space VLBI observations using an antenna on a TDRSS satellite. ApJ 358, 350–358. doi:10.1086/168992. Linfield, R.P., Levy, G.S., Ulvestad, J.S., et al.,

  12. [1994]

    ApJ 426, 51–59

    Equipartition brightness temperature and the in- verse Compton catastrophe. ApJ 426, 51–59. doi: 10.1086/174038. Shklovskii, I.S.,

  13. [2000]

    PASJ 52, 975–982

    Dual- Frequency VSOP Observations of AO 0235+164. PASJ 52, 975–982. arXiv:arXiv:astro-ph/0007347. Gwinn, C.R., Kovalev, Y.Y., Johnson, M.D., Soglasnov, V.A.,

  14. [2004]

    The VSOP 5 GHz Active Galactic Nucleus Survey. IV. The Angular Size/Brightness Temperature Distribution. ApJ 616, 110–122. doi: 10.1086/424811, arXiv:arXiv:astro-ph/0407069. Johnson, M.D., Gwinn, C.R.,

  15. [2005]

    Sub- Milliarcsecond Imaging of Quasars and Active Galactic Nuclei. IV. Fine-Scale Structure. AJ 130, 2473–2505. doi: 10.1086/497430, arXiv:arXiv:astro-ph/0505536. 13 Kutkin, A.M., Pashchenko, I.N., Lisakov, M.M., et al.,

  16. [2007]

    ApJL 664, L71–L74

    An Exceptional Very High Energy Gamma-Ray Flare of PKS 2155-304. ApJL 664, L71–L74. doi: 10.1086/520635, arXiv:0706.0797. Albert, J., Aliu, E., Anderhub, H., et al.,

  17. [2008]

    The VSOP 5 GHz Active Galactic Nucleus Survey. V. Imaging Results for the Remaining 140 Sources. ApJS 175, 314–355. doi: 10.1086/525025, arXiv:0710.5707. Edwards, P.G., Kovalev, Y.Y., Ojha, R., et al.,

  18. [2010]

    Advances in Astronomy 2010, ID 349171 (6 pages)

    Master Robotic Net. Advances in Astronomy 2010, ID 349171 (6 pages). doi: 10.1155/2010/ 349171, arXiv:0907.0827. Lisakov, M.M., Kovalev, Y.Y., Savolainen, T., Hovatta, T., Kutkin, A.M.,

  19. [2013]

    RadioAstron

    “RadioAstron”-A telescope with a size of 300 000 km: Main param- eters and first observational results. Astronomy Reports 57, 153–194. doi:10.1134/S1063772913030025, arXiv:1303.5013. Kellermann, K.I., Pauliny-Toth, I.I.K.,

  20. [2014]

    Ground-based and Airborne Telescopes V, p

    The RadioAstron Green Bank Earth Station, in: Proceedings of the SPIE. Ground-based and Airborne Telescopes V, p. 91450B. doi: 10.1117/12.2056761. Frey, S., Gurvits, L.I., Altschuler, D.R., et al.,

  21. [2015]

    Cosmic Research 53, 216–225

    Fakerat software in the international interferometric radioastron project with very long ground-space bases. Cosmic Research 53, 216–225. doi: 10.1134/S0010952515030090, arXiv:1404.2430. 16

  22. [2016]

    ApJL 820, L10

    Extreme Brightness Temperatures and Refractive Substructure in 3C273 with RadioAstron. ApJL 820, L10. doi: 10.3847/2041-8205/820/1/L10, arXiv:1601.05810. Johnson, M.D., Narayan, R., Psaltis, D., et al.,

  23. [2017]

    PASA 34, e021

    PKS 1954-388: Ra- dioAstron Detection on 80,000 km Baselines and Multiwavelength Obser- vations. PASA 34, e021. doi:10.1017/pasa.2017.16, arXiv:1705.02067. Ford, H.A., Anderson, R., Belousov, K., et al.,

  24. [2018]

    MNRAS 475, 4994–5009

    The ex- treme blazar AO 0235+164 as seen by extensive ground and space ra- dio observations. MNRAS 475, 4994–5009. doi: 10.1093/mnras/sty144, arXiv:1801.04892. Levy, G.S., Linfield, R.P., Edwards, C.D., et al.,

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.