Pith. sign in

REVIEW 3 major objections 4 minor 89 references

Symbiotic star candidates in Gaia Data Release 3

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A machine-learning search of Gaia Data Release 3 yields 1,674 new symbiotic star candidates, 25 with X-ray emission.

desk verdict Useful matched-filter catalog, but the machine-learning validation is circular, so the 1,674 count and the 25 X-ray picks are not out-of-sample tested. read the letter →

arxiv 2506.20505 v2 pith:L2FKXW5Y submitted 2025-06-25 astro-ph.SR astro-ph.GA

Samantha E. Ball , Benjamin C. Bromley , Scott J. Kenyon This is my paper
classification astro-ph.SRastro-ph.GA
keywords symbioticstarsGaiaDR3machine-learningclassifierspectralprojectioncoefficientX-raycounterpartsredgiantsbinarycandidatecatalog
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

Symbiotic stars are interacting binaries in which a cool red giant feeds a hot compact companion, usually a white dwarf; only a few hundred are known, yet the Milky Way is expected to hold thousands. This paper reports a search of Gaia Data Release 3 that uses the low-resolution spectra of 171 known symbiotics to build a mean spectral template, then trains a machine-learning classifier on the match to that template together with Gaia colors, magnitudes, and astrometric quality. Applied to bright, well-measured sources in a red-giant region of the color-magnitude diagram, the classifier returns 1,674 new candidates, 25 of which have X-ray counterparts and are singled out as the most compelling targets for follow-up. If confirmed, these candidates would substantially expand the known symbiotic population and sharpen our picture of the evolutionary path from binary giants to supernova progenitors.

What carries the argument

The load-bearing object is the spectral projection coefficient S of Eq. (3). Each source's BP/RP spectrum is normalized to unit total flux, continuum-subtracted with a smoothing filter, and scaled to unit pointwise standard deviation; S is then the inner product of that residual spectrum with the catalog-average residual spectrum over 600–775 nm, the window containing Hα and the TiO bands. Values S>1 couple more strongly to the symbiotic template than the average training star, and the paper uses S>1 as a pre-cut before classification. The random-forest classifier combines S with $bp-rp$, $G$-band magnitude, parallax and its signal-to-noise, astrometric goodness-of-fit, the renormalized unit weight error (ruwe), total proper motion, and absolute magnitude $M_G$; these features encode the cool-giant locus in the color-magnitude plane and the astrometric jitter expected from binary orbital motion at a few astronomical units. The classifier's role is to reject the many red giants, subgiants, and reddened Be or O stars that pass the S>1 cut but are not true symbiotics.

What would settle it

Take medium-resolution optical spectra of the 25 X-ray-matched candidates; if a majority show no Hα emission, no TiO absorption, and no blue excess from a hot companion, the claim that these are symbiotic stars is falsified. A sharper control is to run the same classifier on random GDR3 red giants that pass S>1 but lie outside the red-giant zone; a similar confirmation rate would show the selection zone, not the symbiotic template, is driving the identifications.

Watch

Extended reading notes

Core claim

The central claim is that the defining spectral signature of symbiotic stars—Hα emission from ionized wind material, TiO absorption from the cool giant, and a blue excess from the hot component—can be compressed into a single projection coefficient S, the normalized overlap between an individual Gaia BP/RP spectrum and the mean spectrum of 171 known symbiotics. A random-forest classifier trained on S together with Gaia colors, magnitudes, parallax, proper motion, and astrometric quality indicators separates symbiotics from the general red-giant population; on the paper's test set it recovers all known symbiotics with very few false positives. When the classifier is run on the full Gaia archive under explicit brightness, parallax-signal-to-noise, color-magnitude, and S>1 selection, it yields 1,674 candidates after removing known symbiotics. Twenty-five of these candidates also appear in X-ray catalogs, and the paper presents them, along with a few high-S, high-astrometric-jitter stars, as the most promising discoveries for spectroscopic confirmation.

Load-bearing premise

The 171 known symbiotic stars used to construct the spectral template and train the classifier must be representative of all symbiotic stars that Gaia can see, including dusty, heavily reddened, and D-type systems; if that sample is biased, then the S>1 cut and the machine-learning decision boundary inherit the bias and the 1,674 count will reflect selection choices rather than the true population.

Editorial extensions

If this is right

  • If even a fraction of the 1,674 candidates are confirmed, the known symbiotic population would grow several-fold, narrowing the long-standing gap between the ~300 confirmed systems and predicted Milky Way populations of $10^3$–$10^5$.
  • The 25 X-ray-matched candidates are concrete, prioritized targets: follow-up spectroscopy on them directly tests both the machine-learning selection and the link between accretion and X-ray emission in symbiotics.
  • The method shows that one carefully chosen spectral summary statistic, combined with public astrometry and photometry, can mine rare binaries from a billion-star archive without dedicated high-resolution spectra.
  • The classifier's ranking recovers known symbiotics among the candidates (48 known systems, with the first 30 judged true symbiotics), so the probability ranking can be used to schedule observations from most to least promising.
  • The high-probability subset—182 candidates at P>90%, 36 at P≥95%—provides an immediate shortlist for deeper follow-up before the full catalog is examined.

Reading between the lines

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

  • If the X-ray candidates confirm as symbiotics, their space density in the Gaia-selected volume could recalibrate population-synthesis rates for the symbiotic channel to Type Ia supernovae, a link the paper raises but does not quantify.
  • The same template-and-classifier pipeline could be rerun on future Gaia releases with longer astrometric baselines; the longer baseline should sharpen the ruwe signal for the roughly two-year orbital periods that presently sit near the 34-month sensitivity window.
  • A testable extension is to drop the S>1 pre-cut and feed the full spectral coefficients to the classifier; this would reveal whether dusty D-type or heavily reddened symbiotics, which may not match the mean S-type-dominated template, are being lost at the pre-selection stage.
  • The paper notes candidates skew bluer and fainter than known symbiotics and that 60% lie within 100 pc of the Galactic plane; a plausible inference is that reddened Be and young O-type stars contaminate the fainter tail, so lower-probability candidates should show a lower spectroscopic confirmation rate than the X-ray-selected subset.
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 / 4 minor

Summary. The paper uses Gaia DR3 astrometry, photometry, and low-resolution BP/RP spectra to construct a catalog of symbiotic star candidates. From 248 known symbiotics identified in SIMBAD, the authors define an HR-diagram selection zone, build a mean spectral template, and introduce a projection coefficient S (Eq. 3). They draw a random comparison sample of 47,693 field stars, train a Random Forest classifier on 1,500 positive rows generated by duplicating the 171 in-zone known symbiotics and 1,500 field-star negatives, and then mine the full GDR3 archive for sources with S > 1. The pipeline yields 1,674 candidates, of which 25 have X-ray counterparts and are presented as particularly compelling targets. The authors state clearly that spectroscopic follow-up is required and acknowledge that many candidates may be subgiants or reddened Be/O stars.

Significance. If the candidate list is reliable, it would be a substantial expansion of the known symbiotic population and would provide a useful set of prioritized targets for spectroscopic confirmation. The paper is honest in framing the output as a candidate catalog, it makes the full table available as an ancillary file, and it documents the main heuristic choices (selection zone, projection wavelength range, S threshold) in a reproducible way. However, the quantitative performance claims rest on a validation procedure that is compromised by direct overlap between the training and test rows, so the stated 1.0 accuracy and the ML probability ranking are not yet established. The central catalog, stripped of the ML-validation language, may still be a useful starting point, but the current presentation overstates what has been demonstrated.

major comments (3)
  1. [Section 3.1, Figure 6] The reported classifier performance is not a valid out-of-sample measurement. The 1,500 positive rows are produced by duplicating the same 171 known symbiotics with replacement, and the positive and negative rows are then mixed and split in half. Roughly 750 of the positive test rows are therefore exact feature-vector duplicates of positive training rows, and the feature set includes RA, Dec, and proper motions, which makes row-level memorization trivial for a Random Forest. The 1.0 accuracy, zero false negatives, and three false positives in Figure 6 therefore measure memorization of training instances, not generalization to new sources. Since the ML probability P is used to rank the 1,674 candidates and to define the 25 high-promise X-ray candidates in Section 4.1 and Table 3, this defect is load-bearing for the abstract's central claim.
  2. [Section 4, paragraph 1] The check using the 48 known symbiotics that appear in the mined S > 1 list is also an in-sample check. These 48 sources come from the same 171-star catalog used to build the template and train the classifier, so their high ranking only shows that the classifier recognizes training material. It does not demonstrate accuracy on genuinely new objects. The authors should provide an object-level out-of-sample evaluation, for example by withholding a random subset of the 171 known symbiotics from both the template construction and the classifier training, or by training on one independent symbionic catalog and testing on another, and then reporting contamination and completeness on the withheld objects.
  3. [Section 2.4 and Section 3.2] The final candidate count is almost entirely determined by the S > 1 matched-filter cutoff rather than by the machine-learning step: of 1,790 S > 1 sources, the classifier rejects only 68. Because S is defined as a projection onto the mean spectrum of the very same 171 known symbiotics, the 1,674 number is the tail of a matched-filter distribution whose center was constructed from the input catalog. This makes the catalog's completeness and contamination sensitive to how representative those 171 stars are of the full symbiotic population. The paper itself notes in Section 4 that most candidates are bluer and fainter than known symbiotics and may be subgiants or reddened Be/O stars with H-alpha emission, which is exactly the population that an out-of-sample test must separate. The authors should quantify the false-positive rate using a clean test sample of non-symbiotic stars drawn from the same selection zone, and should report how the candidate list changes when the template is built from a subset of the known symbiotics or from S-type and D-type stars separately.
minor comments (4)
  1. [Section 3.1] The description of the train/test construction is confusing: the text first calls the 1,500 field stars a 'test dataset' and then says the combined 3,000-row set is mixed and split in half. Please clarify that the split is row-level, not object-level, and define which rows are used for training versus testing.
  2. [Section 3.1 feature list] Including RA, Dec, and proper motions as classifier features is physically questionable for identifying symbiotic status and makes positional memorization easier. Please justify these features or remove them from the classifier and show that the results do not change.
  3. [Section 4, Section 5] There are several typographical errors: 'coeffecient' in Section 5, 'emcompasses' and 'fortuitiously' in Sections 4 and 3.1, and 'Savitsky-Golay' in Section 2.4 should be 'Savitzky-Golay'. Figure 1's caption also contains 'GRD3' instead of 'GDR3'.
  4. [Table 3 and Figure 10] The X-ray luminosities in Table 3 and Figure 10 are derived from three surveys with different energy bands and are said to be order-of-magnitude estimates; please state the assumed distances and band corrections explicitly in the table note or figure caption so that the values can be reproduced.

Circularity Check

2 steps flagged · score 6.0 of 10

The reported 1.0 classifier accuracy is an artifact of same-star train/test duplication, so the ML-based ranking and the 1,674-candidate catalog rest on an in-sample matched filter rather than out-of-sample prediction.

  1. fitted input called prediction [Section 3.1, 'A machine-learning classifier' and Fig. 6 confusion matrix]
    "In creating the training set, we use the 171 known symbiotic stars within our selected region. ... we copied the full 171 over into a new dataset at least once, from there we randomly selected from the 171, allowing duplicates, to add up to a list of 1500 total stars. ... The combined test/train sets totals to 3000 sources. This list was then randomly mixed and split in half (1500 sources) to be tested by the classifier. ... Overall, the matrix reflects perfect recall (no missed symbiotics) and high precision, with a 1.0 accuracy."

    The positive rows are generated by duplicating the same 171 known symbiotics, then randomly mixed and split in half; therefore roughly half of the positive test rows are exact duplicate feature vectors of positive training rows. A Random Forest can memorize these exact rows, especially with RA, Dec, and proper motions as features, so the reported 1.0 accuracy, zero false negatives, and only three false positives are forced by construction rather than measuring out-of-sample generalization. The later 'test' on 48 known symbiotics found in the mined list is also in-sample, because those 48 come from the same 171-star training catalog. The classifier probability P used to rank the 1,674 candidates and to define the 25 high-promise X-ray sources is thereby unvalidated for new Gaia sources.

  2. self definitional [Section 2.4, Eq. (3); Section 3.2, mining cut; Section 4, candidate count]
    "Si≡β(f i, ¯f)=∑j fij ¯fj, (3) ... The angular braces denote the catalog average. ... we further limit our operation to sources with S >1, thus focusing on the peak of the projection coefficient distribution for known symbiotic stars. As roughly anticipated from the 47k stars, this selection nets us 1,790 candidates ... After removing the known symbiotics, the ML classifier identified 1,674 stars as candidates with a formal probability P>50% of being a symbiotic star."

    S is defined as the dot product of each source's spectrum with the mean spectrum of the very same 171 known symbiotics that define the catalog and train the classifier. The candidate cut S>1 therefore selects the tail of a matched-filter distribution whose kernel was fit to the input sample; the 1,674-source catalog is not an independent prediction but a thresholded projection of the known-sample template. The ML step is nearly a pass-through (1,790 S>1 sources minus 48 known leaves 1,742, and the classifier rejects only 68), so the central 'prediction' reduces to the S>1 matched filter built from the training sample itself.

full rationale

The paper is transparent that its method is template matching and supervised classification, and it does not rely on self-citations as load-bearing evidence. However, the circularity lies in the validation and in the construction of the candidate metric. The reported 1.0 classifier accuracy is a tautology because the positive test rows are duplicates of positive training rows drawn from the same 171 known symbiotics. The S>1 selection criterion is defined as projection onto the mean spectrum of that same 171-star sample, so the resulting 1,674-candidate catalog is the tail of a matched-filter distribution whose kernel was fit to the input, not an independent out-of-sample prediction. The X-ray cross-matches provide some external information for the 25 high-promise sources, but those sources are pre-selected by the unvalidated classifier probability P>=80, so the selection ranking remains unvalidated. These factors make the central claims partially circular: the ML performance claim reduces by construction, and the candidate count is essentially a thresholded projection of the training template. Absent clean out-of-sample validation, the 1.0 accuracy and the P-based rankings should not be presented as evidence of generalization. Score 6 reflects one or more predictions reducing by construction while the underlying template-matching idea retains some independent content.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities. Its central output is a catalog whose size and composition depend on several hand-tuned thresholds and on the representativeness of the 171 known symbiotics; these are the effective free parameters of the search.

free parameters (6)
  • Selection zone boundaries = 1 < BP-RP < 7 mag; -6 < M_G < min(1.8*BP-RP - 3.4) mag
    Chosen in Eq. (1) to enclose a strong majority of the 248 known symbiotics while excluding hot supergiants; a hand-tuned cut.
  • Projection coefficient wavelength range = 600 to 775 nm
    Chosen in Section 2.4 to focus on H-alpha and TiO bands while giving both roughly equal weight.
  • Mining threshold on projection coefficient = S > 1
    Set to the peak of the known-symbiotic S distribution (Sections 2.5 and 3.2); yields roughly 1,500 candidates in the pilot random draw and 1,790 in the full archive.
  • Red giant zone boundary = M_G < 2.1*BP-RP - 5 mag
    Adopted in Eq. (4) so that over 90% of known symbiotics lie in the region; used to interpret candidate demographics.
  • Positive training sample size = 1500 (from 171 sources with duplication)
    The 171 known symbiotics are copied and randomly selected into a 1500-star positive set (Section 3.1); the size is arbitrary and the duplication causes train/test leakage.
  • Random Forest hyperparameters = grid-tuned, values not reported
    Described only as 'parameter grid' in Section 3.1; specific settings are not provided, limiting exact reproduction.
assumptions (5)
  • domain assumption The 171 known symbiotic stars with Gaia BP/RP spectra within the selection zone are representative of the full symbiotic star population in GDR3.
    The S template, the S>1 threshold, and the ML decision boundary are all fitted to these 171 stars; if they miss D-type or heavily reddened symbiotics, the catalog is biased. Invoked in Sections 2.2 and 3.1.
  • domain assumption The 47,693 random Gaia sources in the selection zone are a representative sample of non-symbiotic field stars.
    Used as the negative training class and to calibrate the S distribution (Sections 2.5 and 3.1). If the random draw is contaminated by unknown symbiotics or biased by the random_index ordering, the negative class is mis-specified.
  • domain assumption Gaia BP/RP low-resolution spectra, after continuum subtraction and normalization, retain enough information to distinguish symbiotic stars through the projection coefficient S.
    The entire search rests on the template construction in Section 2.4; weak or distorted spectra would make S a noisy feature.
  • domain assumption Astrometric ruwe >1.3 is a useful binarity indicator for symbiotic stars with orbital periods comparable to the 34-month Gaia baseline.
    Used to justify ruwe as an ML feature and to interpret the X-ray candidates in Section 2.3 and Figure 10.
  • domain assumption Positional cross-matches with ROSAT (30 arcsec), XMM-Newton (5 arcsec), and eROSITA (20 arcsec) correctly associate X-ray sources with the optical counterparts.
    Used to identify 25 high-promise candidates in Section 4.1; mismatches would inflate false positives.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Symbiotic star candidates in Gaia Data Release 3." pith.science (2026). https://pith.science/paper/L2FKXW5Y

@misc{pith2026250620505,
  author       = {Pith},
  title        = {Pith review of: Symbiotic star candidates in Gaia Data Release 3},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L2FKXW5Y}},
  note         = {Machine review of arXiv:2506.20505}
}
abstract

Symbiotic stars, binary pairs with a cool giant fueling accretion onto a hot compact companion, offer unique insights to our understanding of stellar evolution. Yet, only a few hundred symbiotic stars are confirmed. Here, we report on a new search for symbiotic star candidates in Gaia Data Release 3 (GDR3), based entirely on the archive's astrometric, photometric, and spectroscopic information. To begin our search, we identified known symbiotic stars in GDR3 and assessed their absolute magnitude and colors, which are dominated by the cool giant. We also considered measures of astrometric quality that might be affected by binary motion in these systems. Finally, from those sources with Gaia spectroscopic data, we built a low-resolution spectral template that characterizes the unique features of these systems, including H$\alpha$ emission from interaction with the giant's wind and radiation from the hot star. We then queried the full GDR3 archive for sources with spectroscopic data that are bright (< 17 mag in G-band), have modest relative parallax uncertainties (< 20%), and fall within a region of color-magnitude space characteristic of red giants, keeping only sources with spectra that quantitatively match our template. A machine-learning algorithm, trained on known symbiotic stars, produced a new catalog of 1,674 sources. From cross-matches with infrared and X-ray surveys, we present 25 of these sources as particularly compelling candidates for new symbiotic stars.

Figures

Figures reproduced from arXiv: 2506.20505 by the authors.

Figure 1
Figure 1. — A Hertzsprung-Russel diagram for the 248 symbiotic stars highlighted in this study. Each point is derived from GRD3 parameters. Most of the sources (224, violet markers) are con￾firmed in Akras et al. (2019a, Tables 6+7 therein), while the re￾maining 24 were identified here with SIMBAD (blue-green mark￾ers). The larger points indicate that the GDR3’s parallax error is less than 20%. The shape of the symbol indicat… view at source ↗
Figure 2
Figure 2. — Color-color diagrams of known symbiotic stars in our catalog of 248 sources. In the left plot, J and K magnitudes are from SIMBAD, while the G-band magnitudes are from the phot g mean mag column in the GDR3 main source catalog. On the right are WISE colors. The symbol color of each point designates the extinction (ag gspphot), provided in the Gaia archive, unless it is not provided (gray points). cess as a sign of… view at source ↗
Figure 3
Figure 3. — The GDR3 astrometric error parameter ruwe as a func￾tion of parallax. The data is consistent with binary stars at a separation of a couple of astronomical units. Nearby sources have higher overall astrometric errors than distant stars, since the or￾bital displacement on the plane of the sky is greater. If the orbital separation were much smaller or much larger than a few astronomi￾cal units, then orbital motion wo… view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: — Distribution of projection coefficient S from a random draw of over 47,000 sources from Gaia (yellow-tinted histogram) and from our sample of symbiotic stars (violet-tinted histogram). Sources with S ≳ 0 corresponds to potential symbiotic star candi￾dates in the rand…
Figure 6
Figure 6. Figure 6: — Confusion matrix of our classifier. The y-axis repre￾sents the true labels, with the top row corresponding to true non￾symbiotic stars and the bottom row corresponding to true symbi￾otic stars. The x-axis represents the predicted labels, where the left column indicat…
Figure 8
Figure 8. Figure 8: — GDR BP+RP spectra of symbiotic star candidates. The thick purple curve is the mean spectrum of known symbiotic stars, the green curve is the average spectrum of all candidate stars, while the thin curves are individual spectra. bility sources (P ≥ 95%). The basic spe…
Figure 9
Figure 9. Figure 9: — Optical, near infrared and WISE colors for symbiotic star candidates, similar to [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: — [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

89 extracted references · 23 canonical work pages

  1. [1]

    2023, MNRAS, 519, 6044, doi:10.1093/mnras/stad096

    Akras, S. 2023, MNRAS, 519, 6044, doi:10.1093/mnras/stad096

    doi:10.1093/mnras/stad096 2023

  2. [2]

    R., Alvarez-Candal, A., & Pereira, C

    Akras, S., Gon¸ calves, D. R., Alvarez-Candal, A., & Pereira, C. B. 2021, MNRAS, 502, 2513, doi:10.1093/mnras/stab195

    doi:10.1093/mnras/stab195 2021

  3. [3]

    2019a, ApJS, 240, 21, doi:10.3847/1538-4365/aaf88c

    Ramos-Larios, G. 2019a, ApJS, 240, 21, doi:10.3847/1538-4365/aaf88c

    doi:10.3847/1538-4365/aaf88c

  4. [4]

    2019b, MNRAS, 483, 5077, doi:10.1093/mnras/sty3359

    Ramos-Larios, G. 2019b, MNRAS, 483, 5077, doi:10.1093/mnras/sty3359

    doi:10.1093/mnras/sty3359

  5. [5]

    Allen, D. A. 1978, MNRAS, 184, 601, doi:10.1093/mnras/184.3.601 —. 1980, MNRAS, 192, 521, doi:10.1093/mnras/192.3.521 —. 1984, PASA, 5, 369, doi:10.1017/S1323358000017215

    work page doi:10.1093/mnras/184.3.601 1978

  6. [6]

    1981, Space Sci

    Aschenbach, B., Br¨ auninger, H., Briel, U., et al. 1981, Space Sci. Rev., 30, 569, doi:10.1007/BF01246075

    doi:10.1007/bf01246075 1981

  7. [7]

    2023, A&A, 674, A32, doi:10.1051/0004-6361/202243790

    Babusiaux, C., Fabricius, C., Khanna, S., et al. 2023, A&A, 674, A32, doi:10.1051/0004-6361/202243790

    doi:10.1051/0004-6361/202243790 2023

  8. [8]

    Bailer-Jones, C. A. L., Rybizki, J., Fouesneau, M., Demleitner, M., & Andrae, R. 2021, AJ, 161, 147, doi:10.3847/1538-3881/abd806

    doi:10.3847/1538-3881/abd806 2021

Show all 89 references
  1. [9]

    Bath, G. T. 1977, MNRAS, 178, 203, doi:10.1093/mnras/178.2.203 Belczy´ nski, K., Miko/suppress lajewska, J., Munari, U., Ivison, R. J., &

    1977 doi

  2. [10]

    2000, A&AS, 146, 407, doi:10.1051/aas:2000280

    Friedjung, M. 2000, A&AS, 146, 407, doi:10.1051/aas:2000280

    2000 doi

  3. [11]

    J., Espey, B

    Birriel, J. J., Espey, B. R., & Schulte-Ladbeck, R. E. 2000, ApJ, 545, 1020, doi:10.1086/317851

    2000 doi

  4. [12]

    Blind, N., Boffin, H. M. J., Berger, J. P., et al. 2011, A&A, 536, A55, doi:10.1051/0004-6361/201118036

    2011 doi

  5. [13]

    V., & Bajkova, A

    Bobylev, V. V., & Bajkova, A. T. 2016, Baltic Astronomy, 25, 261, doi:10.1515/astro-2017-0128

    2016 doi

  6. [14]

    2000, A&AS, 143, 33, doi:10.1051/aas:2000331

    Bonnarel, F., Fernique, P., Bienaym´ e, O., et al. 2000, A&AS, 143, 33, doi:10.1051/aas:2000331

    2000 doi

  7. [15]

    Boyarchuk, A. A. 1969, Commmunications of the Konkoly Observatory Hungary, 65, 395

    1969

  8. [16]

    2022, ApJ, 938, 110, doi:10.3847/1538-4357/ac8e04

    Brout, D., Scolnic, D., Popovic, B., et al. 2022, ApJ, 938, 110, doi:10.3847/1538-4357/ac8e04

    2022 doi

  9. [17]

    2018, A&A, 616, L3, doi:10.1051/0004-6361/201833633

    Bujarrabal, V., Alcolea, J., Miko/suppress lajewska, J., Castro-Carrizo, A., & Ramstedt, S. 2018, A&A, 616, L3, doi:10.1051/0004-6361/201833633

    2018 doi

  10. [18]

    H., & Snow, Jr., T

    Buss, Jr., R. H., & Snow, Jr., T. P. 1988, ApJ, 335, 331, doi:10.1086/166931

    1988 doi

  11. [19]

    R., et al

    Castro-Ginard, A., Penoyre, Z., Casey, A. R., et al. 2024, A&A, 688, A1, doi:10.1051/0004-6361/202450172

    2024 doi

  12. [20]

    2025, arXiv e-prints, arXiv:2506.09352, doi:10.48550/arXiv.2506.09352

    Chen, J., Wang, L., Li, Y.-B., et al. 2025, arXiv e-prints, arXiv:2506.09352, doi:10.48550/arXiv.2506.09352

    2025 doi arXiv

  13. [21]

    M., Wright, E

    Cutri, R. M., Wright, E. L., Conrow, T., et al. 2013, Explanatory Supplement to the AllWISE Data Release Products, Explanatory Supplement to the AllWISE Data Release Products, by R. M. Cutri et al. De Angeli, F., Weiler, M., Montegriffo, P., et al. 2023, A&A, 674, A2, doi:10.1...

    2013 doi

  14. [22]

    2023, A&A, 674, A13, doi:10.1051/0004-6361/202244242

    Eyer, L., Audard, M., Holl, B., et al. 2023, A&A, 674, A13, doi:10.1051/0004-6361/202244242

    2023 doi

  15. [23]

    2024, A&A, 684, A121, doi:10.1051/0004-6361/202348278 Gaia Collaboration, Vallenari, A., Brown, A

    Freund, S., Czesla, S., Predehl, P., et al. 2024, A&A, 684, A121, doi:10.1051/0004-6361/202348278 Gaia Collaboration, Vallenari, A., Brown, A. G. A., et al. 2023a, A&A, 674, A1, doi:10.1051/0004-6361/202243940 Gaia Collaboration, Montegriffo, P., Bellazzini, M., et al. 2023b, ...

    2024 doi

  16. [24]

    N., Frankel, N., Zari, E., Xiang, M., & Rix, H.-W

    Garzon, D. N., Frankel, N., Zari, E., Xiang, M., & Rix, H.-W. 2024, ApJ, 972, 155, doi:10.3847/1538-4357/ad61e8

    2024 doi

  17. [25]

    Gilmore, G., Wyse, R. F. G., & Kuijken, K. 1989, ARA&A, 27, 555, doi:10.1146/annurev.aa.27.090189.003011

    1989

  18. [26]

    A., et al

    Godoy-Rivera, D., Mathur, S., Garc´ ıa, R. A., et al. 2025, A&A, 696, A243, doi:10.1051/0004-6361/202348735

    2025 doi

  19. [27]

    2019, ApJ, 887, 93, doi:10.3847/1538-4357/ab5362

    Finkbeiner, D. 2019, ApJ, 887, 93, doi:10.3847/1538-4357/ab5362

    2019 doi

  20. [28]

    2013, Acta Astron., 63, 405, doi:10.48550/arXiv.1312.6063

    Gromadzki, M., Miko/suppress lajewska, J., & Soszy´ nski, I. 2013, Acta Astron., 63, 405, doi:10.48550/arXiv.1312.6063

    2013 doi arXiv

  21. [29]

    A., Vasquez-Torres, D

    Guerrero, M. A., Vasquez-Torres, D. A., Rodr´ ıguez-Gonz´ alez, J. B., Toal´ a, J. A., & Ortiz, R. 2025, A&A, 693, A203, doi:10.1051/0004-6361/202451715

    2025 doi

  22. [30]

    1999, ApJ, 522, 487, doi:10.1086/307608

    Hachisu, I., Kato, M., & Nomoto, K. 1999, ApJ, 522, 487, doi:10.1086/307608

    1999 doi

  23. [31]

    2024, Monthly Notices of the Royal Astronomical Society, 529, 3630, doi:10.1093/mnras/stae738

    Healy, S., Horiuchi, S., Colomer Molla, M., et al. 2024, Monthly Notices of the Royal Astronomical Society, 529, 3630, doi:10.1093/mnras/stae738

    2024 doi

  24. [32]

    2021, ApJ, 915, 105, doi:10.3847/1538-4357/ac03b1

    Heo, J.-E., Lee, H.-W., Angeloni, R., Palma, T., & Di Mille, F. 2021, ApJ, 915, 105, doi:10.3847/1538-4357/ac03b1

    2021 doi

  25. [33]

    Hunsch, M., Schmitt, J. H. M. M., & Voges, W. 1998, A&AS, 127, 251, doi:10.1051/aas:1998347

    1998 doi

  26. [34]

    A., Zasowski, G., et al

    Imig, J., Holtzman, J. A., Zasowski, G., et al. 2025, arXiv e-prints, arXiv:2507.17629, doi:10.48550/arXiv.2507.17629

    2025 doi arXiv

  27. [35]

    2001, A&A, 365, L1, doi:10.1051/0004-6361:20000036

    Jansen, F., Lumb, D., Altieri, B., et al. 2001, A&A, 365, L1, doi:10.1051/0004-6361:20000036

    2001 doi

  28. [36]

    2023, Research in Astronomy and Astrophysics, 23, 105012, doi:10.1088/1674-4527/ace9b2

    Jia, Y., Guo, S., Zhu, C., et al. 2023, Research in Astronomy and Astrophysics, 23, 105012, doi:10.1088/1674-4527/ace9b2

    2023 doi

  29. [37]

    Kenyon, S. J. 1986, The symbiotic stars (Cambridge University Press)

    1986

  30. [38]

    J., & Fernandez-Castro, T

    Kenyon, S. J., & Fernandez-Castro, T. 1987, AJ, 93, 938, doi:10.1086/114379

    1987 doi

  31. [39]

    J., Livio, M., Mikolajewska, J., & Tout, C

    Kenyon, S. J., Livio, M., Mikolajewska, J., & Tout, C. A. 1993, ApJ, 407, L81, doi:10.1086/186811

    1993 doi

  32. [40]

    J., & Truran, J

    Kenyon, S. J., & Truran, J. W. 1983, ApJ, 273, 280, doi:10.1086/161367

    1983 doi

  33. [41]

    J., & Webbink, R

    Kenyon, S. J., & Webbink, R. F. 1984, ApJ, 279, 252, doi:10.1086/161888

    1984 doi

  34. [42]

    2022, A&A, 657, A7, doi:10.1051/0004-6361/202142146

    Kervella, P., Arenou, F., & Th´ evenin, F. 2022, A&A, 657, A7, doi:10.1051/0004-6361/202142146

    2022 doi

  35. [43]

    R., Rocha-Pinto, H

    Laversveiler, M., Gon¸ calves, D. R., Rocha-Pinto, H. J., & Merc, J. 2025, A&A, 698, A155, doi:10.1051/0004-6361/202451548

    2025 doi

  36. [44]

    2016, ApJ, 833, 75, doi:10.3847/1538-4357/833/1/75

    Lee, Y.-M., Lee, D.-S., Chang, S.-J., et al. 2016, ApJ, 833, 75, doi:10.3847/1538-4357/833/1/75

    2016 doi

  37. [45]

    2023, ApJ, 956, 26, doi:10.3847/1538-4357/aced8a

    Li, L., Wang, S., Chen, X., & Jiang, Q. 2023, ApJ, 956, 26, doi:10.3847/1538-4357/aced8a

    2023 doi

  38. [46]

    K., & Han, Z

    Liu, Z.-W., R¨ opke, F. K., & Han, Z. 2023, Research in Astronomy and Astrophysics, 23, 082001, doi:10.1088/1674-4527/acd89e L¨ u, G., Yungelson, L., & Han, Z. 2006, MNRAS, 372, 1389, doi:10.1111/j.1365-2966.2006.10947.x

    2023

  39. [47]

    B., Sokoloski, J

    Lucy, A. B., Sokoloski, J. L., Luna, G. J. M., et al. 2024, arXiv e-prints, arXiv:2412.00855, doi:10.48550/arXiv.2412.00855

    2024 doi arXiv

  40. [48]

    Luna, G. J. M., Sokoloski, J. L., Mukai, K., & Nelson, T. 2013, A&A, 559, A6, doi:10.1051/0004-6361/201220792 Symbiotic star candidates 13

    2013 doi

  41. [49]

    Magrini, L., Corradi, R. L. M., & Munari, U. 2003, in Astronomical Society of the Pacific Conference Series, Vol. 303, Symbiotic Stars Probing Stellar Evolution, ed. R. L. M

    2003

  42. [50]

    Mikolajewska, & T

    Corradi, J. Mikolajewska, & T. J. Mahoney, 539, doi:10.48550/arXiv.astro-ph/0208085

    doi arXiv

  43. [51]

    2025, arXiv e-prints, arXiv:2504.16825, doi:10.48550/arXiv.2504.16825

    Merc, J. 2025, arXiv e-prints, arXiv:2504.16825, doi:10.48550/arXiv.2504.16825

    2025 doi arXiv

  44. [52]

    Merc, J., & Boffin, H. M. J. 2025, A&A, 695, A61, doi:10.1051/0004-6361/202553789

    2025 doi

  45. [53]

    2019, Research Notes of the American Astronomical Society, 3, 28, doi:10.3847/2515-5172/ab0429

    Merc, J., G´ alis, R., & Wolf, M. 2019, Research Notes of the American Astronomical Society, 3, 28, doi:10.3847/2515-5172/ab0429

    2019 doi

  46. [54]

    A., Toal´ a, J

    Merc, J., Guerrero, M. A., Toal´ a, J. A., & Ortiz, R. 2025, A&A, 699, A117, doi:10.1051/0004-6361/202554671

    2025 doi

  47. [55]

    2020, A&A, 644, A49, doi:10.1051/0004-6361/202039132

    Merc, J., Miko/suppress lajewska, J., Gromadzki, M., et al. 2020, A&A, 644, A49, doi:10.1051/0004-6361/202039132

    2020 doi

  48. [56]

    W., & Humason, M

    Merrill, P. W., & Humason, M. L. 1932, PASP, 44, 56, doi:10.1086/124191

    1932 doi

  49. [57]

    Mikolajewska, J., & Kenyon, S. J. 1992, MNRAS, 256, 177, doi:10.1093/mnras/256.2.177

    1992 doi

  50. [58]

    2025, A&A, 695, A227, doi:10.1051/0004-6361/202453359

    Mulato, L., Merc, J., Charbonnel, S., et al. 2025, A&A, 695, A227, doi:10.1051/0004-6361/202453359

    2025 doi

  51. [59]

    2019, arXiv e-prints, arXiv:1909.01389, doi:10.48550/arXiv.1909.01389 —

    Munari, U. 2019, arXiv e-prints, arXiv:1909.01389, doi:10.48550/arXiv.1909.01389 —. 2025, Contributions of the Astronomical Observatory Skalnate Pleso, 55, 47, doi:10.31577/caosp.2025.55.3.47

    2019 doi arXiv

  52. [60]

    1992a, ApJ, 397, L87, doi:10.1086/186551 —

    Munari, U., & Renzini, A. 1992a, ApJ, 397, L87, doi:10.1086/186551 —. 1992b, ApJ, 397, L87, doi:10.1086/186551

    doi

  53. [61]

    2021, MNRAS, 505, 6121, doi:10.1093/mnras/stab1620

    Munari, U., Traven, G., Masetti, N., et al. 2021, MNRAS, 505, 6121, doi:10.1093/mnras/stab1620

    2021 doi

  54. [62]

    2000, A&AS, 143, 23, doi:10.1051/aas:2000169

    Ochsenbein, F., Bauer, P., & Marcout, J. 2000, A&AS, 143, 23, doi:10.1051/aas:2000169

    2000 doi

  55. [63]

    1946, ApJ, 104, 362, doi:10.1086/144866 —

    Payne-Gaposchkin, C. 1946, ApJ, 104, 362, doi:10.1086/144866 —. 1964, The galactic novae (Dover, New York)

    1946 doi

  56. [64]

    A., Males, J

    Pearce, L. A., Males, J. R., Weinberger, A. J., et al. 2022, MNRAS, 515, 4487, doi:10.1093/mnras/stac2056

    2022 doi

  57. [65]

    Penoyre, Z., Belokurov, V., & Evans, N. W. 2022, MNRAS, 513, 5270, doi:10.1093/mnras/stac1147

    2022 doi

  58. [66]

    2006, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference

    Predehl, P., Hasinger, G., B¨ ohringer, H., et al. 2006, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference

    2006

  59. [67]

    6266, Space Telescopes and Instrumentation II: Ultraviolet to Gamma Ray, ed

    Series, Vol. 6266, Space Telescopes and Instrumentation II: Ultraviolet to Gamma Ray, ed. M. J. L. Turner & G. Hasinger, 62660P, doi:10.1117/12.670249

    doi

  60. [68]

    L., Wright, N

    Quintana, A. L., Wright, N. J., & Garc´ ıa, J. M. 2025, Monthly Notices of the Royal Astronomical Society, 538, 1367, doi:10.1093/mnras/staf083

    2025 doi

  61. [69]

    C., Oudmaijer, R

    Radley, I. C., Oudmaijer, R. D., Vioque, M., & Dodd, J. M. 2025, MNRAS, 539, 1964, doi:10.1093/mnras/staf618

    2025 doi

  62. [70]

    G., Filippenko, A

    Riess, A. G., Filippenko, A. V., Challis, P., et al. 1998, AJ, 116, 1009, doi:10.1086/300499

    1998 doi

  63. [71]

    2023, A&A, 674, A14, doi:10.1051/0004-6361/202245591

    Rimoldini, L., Holl, B., Gavras, P., et al. 2023, A&A, 674, A14, doi:10.1051/0004-6361/202245591

    2023 doi

  64. [72]

    Schmid, H. M. 1989, A&A, 211, L31

    1989

  65. [73]

    M., Bazzon, A., Milli, J., et al

    Schmid, H. M., Bazzon, A., Milli, J., et al. 2017, A&A, 602, A53, doi:10.1051/0004-6361/201629416

    2017 doi

  66. [74]

    Schmitt, J. H. M. M., H¨ unsch, M., Schneider, P. C., et al. 2024, A&A, 688, A9, doi:10.1051/0004-6361/202449181

    2024 doi

  67. [75]

    1923, Harvard College Observatory Bulletin, 783, 3

    Shapley, H. 1923, Harvard College Observatory Bulletin, 783, 3

    1923

  68. [76]

    G., & Torres, G

    Stassun, K. G., & Torres, G. 2021, ApJ, 907, L33, doi:10.3847/2041-8213/abdaad

    2021 doi

  69. [77]

    2011, A&A, 529, A152, doi:10.1051/0004-6361/201015967

    Sturm, R., Haberl, F., Greiner, J., et al. 2011, A&A, 529, A152, doi:10.1051/0004-6361/201015967

    2011 doi

  70. [78]

    1941, ApJ, 93, 356, doi:10.1086/144272

    Swings, P., & Struve, O. 1941, ApJ, 93, 356, doi:10.1086/144272

    1941 doi

  71. [79]

    2019, ApJ, 878, 21, doi:10.3847/1538-4357/ab1ea5

    Ting, Y.-S., & Rix, H.-W. 2019, ApJ, 878, 21, doi:10.3847/1538-4357/ab1ea5

    2019 doi

  72. [80]

    A., Coriat, M., Traulsen, I., et al

    Webb, N. A., Coriat, M., Traulsen, I., et al. 2020, A&A, 641, A136, doi:10.1051/0004-6361/201937353

    2020 doi

  73. [81]

    F., Livio, M., Truran, J

    Webbink, R. F., Livio, M., Truran, J. W., & Orio, M. 1987, ApJ, 314, 653, doi:10.1086/165095

    1987 doi

  74. [82]

    L., & Allen, D

    Webster, B. L., & Allen, D. A. 1975, MNRAS, 171, 171, doi:10.1093/mnras/171.1.171

    1975 doi

  75. [83]

    2000, A&AS, 143, 9, doi:10.1051/aas:2000332

    Wenger, M., Ochsenbein, F., Egret, D., et al. 2000, A&AS, 143, 9, doi:10.1051/aas:2000332

    2000 doi

  76. [84]

    E., Giommi, P., & Angelini, L

    White, N. E., Giommi, P., & Angelini, L. 2000, VizieR Online Data Catalog: The WGACAT version of ROSAT sources (White+ 2000), VizieR On-line Data Catalog: IX/31. Originally published in: Laboratory for High Energy Astrophysics (LHEA/NASA), Greenbelt (2000)

    2000

  77. [85]

    L., Hill, J

    Whiting, M. L., Hill, J. B., Bromley, B. C., & Kenyon, S. J. 2023, AJ, 165, 193, doi:10.3847/1538-3881/acc526

    2023 doi

  78. [86]

    L., Eisenhardt, P

    Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868, doi:10.1088/0004-6256/140/6/1868

    2010 doi

  79. [87]

    2024, ApJ, 962, 126, doi:10.3847/1538-4357/ad20ec Yungel’son, L., Livio, M., Tutukov, A., & Kenyon, S

    Xu, X.-j., Shao, Y., & Li, X.-D. 2024, ApJ, 962, 126, doi:10.3847/1538-4357/ad20ec Yungel’son, L., Livio, M., Tutukov, A., & Kenyon, S. J. 1995, ApJ, 447, 656, doi:10.1086/175908

    2024 doi

  80. [88]

    2023, ApJS, 264, 14, doi:10.3847/1538-4365/ac9dfa

    Zhang, R., & Yuan, H. 2023, ApJS, 264, 14, doi:10.3847/1538-4365/ac9dfa

    2023 doi

  81. [89]

    K., Speagle, J

    Zucker, C., Saydjari, A. K., Speagle, J. S., et al. 2025, arXiv e-prints, arXiv:2503.02657, doi:10.48550/arXiv.2503.02657 This paper was built using the Open Journal of As- trophysics LATEX template. The OJA is a journal which provides fast and easy peer review for new papers ...

    2025 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.