{"id":"2977ffa8-a254-49a7-a6a1-39ab5e3556cc","arxiv_id":"2607.06357","paper_version":1,"verdict":"CONDITIONAL","confidence":"UNKNOWN","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":4,"one_line_summary":"UV spectroscopy of the 44 nearest white dwarfs reveals a 2–6% temperature discrepancy between UV and optical model fits, six UV-only metal detections, and a 30% planetary debris accretion rate.","lead":"This paper analyzed 44 white dwarfs within 13 parsecs of the Sun using Hubble UV spectroscopy combined with optical and infrared data. It found that UV-based temperature measurements systematically exceed optical-only measurements for cool white dwarfs, and that UV spectra reveal metal pollution in six stars where optical spectra showed nothing.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"The UV-optical Teff discrepancy claim hinges on ruling out wavelength-dependent STIS G230L flux calibration systematics, which the Monte Carlo test in §5.1 does not address.","rationale":"The reader correctly identified model reliability as a concern, but conflated DQ/He-atmosphere model issues (C I gf adjustments, DQ low-mass bias from §3.2.2 and §4.2.2) with the H-atmosphere Teff discrepancy claim, which uses a different model code (Tremblay et al. 2013 vs. Koester 2010). The DQ issues are real and the reader is right to flag them, but they primarily affect the DQ parameter determinations and DC carbon limits, not the central H-atmosphere Teff discrepancy. The more directly load-bearing concern for the central claim is whether wavelength-dependent STIS G230L flux calibration systematics could produce the observed Teff offset. The paper's Monte Carlo test addresses random, overall flux-level uncertainties but not wavelength-dependent calibration slopes. This is a genuine gap, though the paper has partial mitigating evidence from the sign reversal at high Teff and the Balmer line comparison. The claim is about the existence of a discrepancy between two fitting approaches, which is robustly demonstrated by the data regardless of which Teff is 'correct.' The interpretation as a model deficiency is reasonable but would be strengthened by ruling out wavelength-dependent calibration effects. The verdict remains CONDITIONAL because the quantitative DQ results and DC carbon limits are model-dependent as the reader notes, but the central H-atmosphere Teff discrepancy observation stands as a genuine finding even with this concern.","tokens_in":48547,"tokens_out":4486,"duration_ms":362800,"concrete_test":"Split each STIS G230L spectrum into blue (1570-2400 Å) and red (2400-3180 Å) halves and repeat the hybrid fits independently for each half for the H-atmosphere DAs with Teff < 10,000 K. If both halves yield Teff values 2-6% above the optical/IR-only fit with consistent sign, wavelength-dependent flux calibration is unlikely to be the cause. If the offset differs significantly in magnitude or sign between the two halves, a wavelength-dependent G230L calibration systematic is contributing to the observed discrepancy.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the 2-6% Teff offset between UV-inclusive and optical/IR-only fits reflects model atmosphere deficiencies rather than observational systematics. The Monte Carlo test in §5.1 perturbs the overall STIS flux level by up to 3% and Gaia magnitudes by 0.01 mag, finding these do not significantly increase parameter uncertainties. However, this test treats flux calibration uncertainties as random perturbations to the overall flux level. The Teff determination from UV data depends primarily on the slope of the UV continuum across the G230L wavelength range (1570-3180 Å), not on the absolute flux level. A wavelength-dependent flux calibration error in STIS G230L — a systematic slope across the grating's wavelength range — would directly bias the derived Teff but would not be captured by the Monte Carlo test as designed. Such wavelength-dependent systematics could be correlated across all G230L observations since they share the same CALSPEC standards and reduction pipeline. The paper does have mitigating evidence: the offset reverses sign at higher Teff where different gratings (E140M/E230M) are used, and a similar offset exists between Balmer line fits and hybrid fits (independent of STIS calibration). However, the Balmer line fits carry their own known ~2% systematic offset (acknowledged in §3.1), and the high-Teff reversal could reflect different model opacity regimes rather than ruling out calibration effects. The reader's concern about DQ model issues (C I gf adjustments, low DQ masses) is valid for the He-atmosphere results but does not directly bear on the H-atmosphere Teff discrepancy, which uses different model codes (Tremblay et al. 2013 vs. Koester 2010).","agreement_with_reader":"partial"},"referee_report":{"model":"glm-5.2","summary":"This paper presents a comprehensive multi-wavelength spectroscopic and photometric analysis of the 44 confirmed white dwarfs within 13 pc of the Sun, combining HST/STIS and COS UV spectroscopy with ground-based optical spectroscopy and Gaia/2MASS/WISE photometry. The authors employ a hybrid fitting method to derive atmospheric parameters for each star, tailored to its spectral type. The key findings include: (1) a systematic 2–6% upward Teff offset for H-atmosphere WDs with Teff < 10,000 K when UV spectra are included in fits compared to optical/IR-only fits; (2) re-classification of three He-rich WDs as metal-enriched based on UV Mg detections; (3) a 30% metal-enrichment fraction and 32% multiplicity fraction; (4) no measurable difference in hydrogen content between DQ and DC WDs, but significantly lower carbon upper limits in DCs; and (5) identification of model deficiencies in reproducing UV carbon lines in DQs and the UV-to-IR SED of cool H-atmosphere WDs.","tokens_in":48775,"tokens_out":2599,"duration_ms":215159,"significance":"This is a valuable volume-limited study of the nearest white dwarfs, providing a uniform UV-to-IR analysis of a sample that is complete to 13 pc. The identification of a systematic Teff offset between UV-inclusive and optical-only fits is an important empirical result that directly tests white dwarf atmosphere models. The paper properly applies the recently identified H3+ partition function correction, validates results against independent mass measurements (microlensing for WD0426+588 and WD1142-645; astrometric for Sirius B and 40 Eri B), and is transparent about model failures (DQ UV carbon lines, low DQ masses from UV fitting). The new UV-only metal detections and the multiplicity fraction measurement add further value. The sample size is necessarily limited by the volume, but the data quality is high and the analysis is thorough.","major_comments":[{"comment":"§5.1, Monte Carlo test (Fig. 16 discussion): The central claim that the 2–6% Teff offset reflects model atmosphere deficiencies rather than observational systematics is not fully supported by the error analysis as designed. The Monte Carlo test perturbs the overall STIS flux level by up to 3% and Gaia magnitudes by 0.01 mag, but Teff determination from UV data depends primarily on the slope of the UV continuum across the G230L wavelength range (1570–3180 Å), not on the absolute flux level. A wavelength-dependent (i.e., slope) calibration error in STIS G230L would directly bias the derived Teff but would not be captured by the Monte Carlo test as designed, since the perturbations are applied uniformly across the spectrum. The paper does note mitigating evidence (the offset reverses sign at higher Teff where different gratings are used, and a similar offset exists between Balmer line andhy","section":null},{"comment":"§3.2.2, DQ fitting and Table 8: The artificial reduction of C I gf values by a factor of ~30 to match UV observations is acknowledged transparently, but the resulting DQ parameters in Table 8 (masses as low as 0.357 M☉ for WD0208-510, 0.470 M☉ for WD1142-645) are inconsistent with both optical-only fits (Coutu et al. 2019) and the model-independent microlensing mass for WD1142-645 (0.56 ± 0.08 M☉; McGill et al. 2023). The paper states that 'including STIS UV data in DQ fitting might bias the solutions towards low Teff and masses' (§4.2.2), yet these parameters are presented in Table 8 without a flag indicating their likely unreliability. The authors should clarify which DQ parameters should be considered trustworthy and which are affected by the known UV model issues, particularly for readers who may use Table 8 without reading the full discussion.","section":null},{"comment":"§5.1, Fig. 16: The claim of a 'systematic' 2–6% Teff offset is based on approximately 15 DA WDs with G230L data below 10,000 K (from Table 4, counting stars with both 'with UV' and 'without UV' entries). The paper does not provide a quantitative statistical test (e.g., a mean offset with uncertainty, or a significance level) for this offset. Given the small sample size and the acknowledged ~2% systematic offset between photometric and spectroscopic Teff solutions (§3.1), a more rigorous statistical characterization of the offset and its significance would strengthen the central claim.","section":null}],"minor_comments":[{"comment":"Table 2: The spectral type for WD1748+708 is listed as 'DXH' with composition '–', but the text (§4.3) discusses it as having a helium-rich or carbon-rich atmosphere. The dash in the composition column is ambiguous.","section":null},{"comment":"§3.2.2: The DQpec Swan band shift parameter is described as a=0.1, but the text also mentions values of 1.6 (theoretical), 0.2 (empirical calibration from Kowalski 2010 and Blouin & Dufour 2019), and 0.1 (this work). It would help to briefly justify why 0.1 was chosen over 0.2 for this sample specifically.","section":null},{"comment":"Fig. 2: The Gaia XP spectra are shown but not incorporated into the fits. While the reasoning is explained, it would be useful to add a brief note in the figure caption reminding the reader that XP spectra are for visual comparison only.","section":null},{"comment":"Table 4: The footnote symbols (∗, ‡, §) indicating which photometry was used are defined, but it would be helpful to also indicate in the table or caption which stars have COS vs. STIS data, as this affects the wavelength coverage and fitting approach.","section":null},{"comment":"§4.2.4, WD0046+051: The text mentions using 'quasi-static profiles described by Walkup (1982)' for Si II lines broadened by neutral helium. A brief explanation of what these profiles are and why they were needed (as opposed to the unified profiles used for other lines) would improve clarity.","section":null},{"comment":"§5.4.2: The enrichment fraction of 30 ± 8% is compared to 45 ± 6% from Ould Rouis et al. (2024), and stated to be consistent at 1.5σ. It would be useful to explicitly state the expected direction of the offset if enrichment properties are age-independent (i.e., the 13 pc sample should have a higher fraction due to longer diffusion timescales), to help the reader evaluate the degeneracy discussion that follows.","section":null},{"comment":"Table 11: The projected separation for G203-47 is listed as 0.05 au, which seems very small for a resolved system. Please verify this value.","section":null},{"comment":"§5.6: The TESS variability results are presented without error bars on the rotation periods. Adding typical uncertainties would be useful for future comparisons.","section":null}],"recommendation":"major_revision","confidential_remarks":"The paper is a substantial observational study that will be a valuable reference for the local white dwarf population. The main concern is that the central claim (UV-optical Teff discrepancy as a model deficiency) rests on a Monte Carlo test that does not address the most relevant systematic (wavelength-dependent flux calibration). The authors have partial mitigating evidence (Balmer line comparison, high-Teff reversal), but the argument needs to be made more carefully. The DQ fitting issues are handled transparently but the presentation of potentially unreliable parameters in Table 8 without flags is a concern. I believe these issues can be addressed within the scope of a revision and do not undermine the overall value of the work. The self-citation pattern (O'Brien et al. 2024 for the 40 pc sample) is appropriate given the sequential nature of the survey program."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for a careful and constructive report. The referee correctly identifies the paper's main contributions and its transparent treatment of model failures. We address each major comment below. In summary: (1) we agree that the Monte Carlo test does not capture wavelength-dependent calibration errors and will revise the text to make this limitation explicit; (2) we agree that Table 8 should flag DQ parameters affected by known UV model issues and will add such flags; and (3) we will add a quantitative statistical characterization of the Teff offset. One point—the suggestion that a slope calibration error in STIS G230L could fully explain the offset—we address with counter-arguments but acknowledge as a legitimate alternative hypothesis that should be discussed more thoroughly.","responses":[{"response":"The referee raises a valid point. Our Monte Carlo test was designed to assess the impact of known absolute flux calibration uncertainties (at the 1–3% level, as characterized by Elms et al. 2024) and Gaia photometric uncertainties, and we agree that it does not capture wavelength-dependent (slope) calibration errors within the G230L bandpass. We acknowledge that a slope error across the 1570–3180 Å range could in principle bias the derived Teff in a manner not captured by our test. We will revise the text to state this limitation explicitly. However, we maintain that the offset is more likely dominated by model deficiencies for the following reasons, which we will also make clearer in the revised text. First, the sign of the offset reverses at higher Teff where different gratings (E140M/E230M, COS) are used; a G230L-specific slope calibration error would not naturally produce this reversal. Second, a similar Teff offset exists between Balmer line fits and photometric fits (a well-known ~2% offset documented by Tremblay et al. 2019; Genest-Beaulieu & Bergeron 2019; and others), which is entirely independent of UV data and thus cannot be attributed to STIS calibration. Third, the STIS G230L flux calibration has been validated against CALSPEC standards at the 1–2% level across the full wavelength range (Bohlin et al. 2019; Elms et al. 2024), and a slope error large enough to produce a 2–6% Teff offset would likely have been detected in those calibration checks. We agree, however, that we cannot fully rule out a contribution from a residual wavelength-dependent calibration error, and we will revise the text to present this as an alternative hypothesis that our current test cannot exclude.","revision_made":"partial","referee_comment":"§5.1, Monte Carlo test (Fig. 16 discussion): The central claim that the 2–6% Teff offset reflects model atmosphere deficiencies rather than observational systematics is not fully supported by the error analysis as designed. The Monte Carlo test perturbs the overall STIS flux level by up to 3% and Gaia magnitudes by 0.01 mag, but Teff determination from UV data depends primarily on the slope of the UV continuum across the G230L wavelength range (1570–3180 Å), not on the absolute flux level. A wavelength-dependent (i.e., slope) calibration error in STIS G230L would directly bias the derived Teff but would not be captured by the Monte Carlo test as designed, since the perturbations are applied uniformly across the spectrum."},{"response":"We agree with this comment. The DQ parameters derived from UV-inclusive fits are known to be biased low in mass, as we discuss in §4.2.2, but Table 8 does not currently flag this for readers who may use the table without reading the full discussion. We will add a column or footnote to Table 8 explicitly flagging the DQ parameters that are affected by known UV model issues (specifically WD0208-510, WD0435-088, WD1142-645, and WD2140+207, which have masses significantly below both optical-only fits and independent mass measurements). We will also add a note directing readers to §4.2.2 for discussion of the UV model issues affecting these parameters. For WD1142-645 specifically, we will note the 2σ tension with the microlensing mass from McGill et al. (2023). We will clarify that for DQ white dwarfs, the optical-only parameters from Coutu et al. (2019) should be preferred for applications requiring reliable masses, while the UV-inclusive parameters are presented to document the model discrepancies.","revision_made":"yes","referee_comment":"§3.2.2, DQ fitting and Table 8: The artificial reduction of C I gf values by a factor of ~30 to match UV observations is acknowledged transparently, but the resulting DQ parameters in Table 8 (masses as low as 0.357 M☉ for WD0208-510, 0.470 M☉ for WD1142-645) are inconsistent with both optical-only fits (Coutu et al. 2019) and the model-independent microlensing mass for WD1142-645 (0.56 ± 0.08 M☉; McGill et al. 2023). The paper states that 'including STIS UV data in DQ fitting might bias the solutions towards low Teff and masses' (§4.2.2), yet these parameters are presented in Table 8 without a flag indicating their likely unreliability."},{"response":"We agree that a quantitative statistical characterization would strengthen the claim and will add this to the revised manuscript. We will compute the mean and median Teff offset (ΔTeff/Teff = (Teff,hybrid − Teff,phot)/Teff,hybrid) for the DA white dwarfs with G230L data below 10,000 K, along with the standard error on the mean and a significance test (e.g., a one-sample t-test against zero offset). We will also quantify how this offset compares to the known ~2% photometric-spectroscopic Teff offset. We note that the known ~2% offset between photometric and spectroscopic (Balmer line) Teff solutions is a separate effect that is also visible in Fig. 16 (red points), and the UV-vs-photometric offset (blue points) is of comparable or larger magnitude but in the same direction. We will discuss the relationship between these two offsets explicitly. The small sample size is an inherent limitation of a volume-limited study, and we will state this clearly, but the uniformity of the data quality and analysis is a compensating strength.","revision_made":"yes","referee_comment":"§5.1, Fig. 16: The claim of a 'systematic' 2–6% Teff offset is based on approximately 15 DA WDs with G230L data below 10,000 K. The paper does not provide a quantitative statistical test (e.g., a mean offset with uncertainty, or a significance level) for this offset. Given the small sample size and the acknowledged ~2% systematic offset between photometric and spectroscopic Teff solutions (§3.1), a more rigorous statistical characterization of the offset and its significance would strengthen the central claim."}],"tokens_in":48678,"tokens_out":1529,"duration_ms":169195,"standing_objections":[]},"desk_editor":{"model":"glm-5.2","letter":"Two things matter here. First, the systematic 2–6% UV-vs-optical Teff offset for H-atmosphere WDs below 10,000 K is a new empirical result with real downstream consequences for cooling ages and chronology. Second, the UV-only metal detections in three previously featureless He-atmosphere WDs are clean new discoveries that raise the local planetary accretion fraction and show optical surveys miss real pollutants. Both are worth the community's attention. The paper does a lot else well. The hybrid UV+photometric fitting of the complete 13 pc volume is a substantial undertaking, executed carefully. The H3+ partition function correction is properly applied. Cross-checks against independent mass measurements — microlensing for WD0426+588 and WD1142-645, astrometric for Sirius B and 40 Eri B — are the right thing to do and the agreement is mostly reasonable. The 30% accretion fraction and 32% multiplicity fraction are useful demographic numbers for a volume-limited sample. The stress-test concern about wavelength-dependent STIS G230L flux calibration systematics does not land as a fatal problem. The Monte Carlo in §5.1 perturbs the overall flux level, not the slope across the grating, which is a fair point. But the Teff offset reverses sign at higher temperatures where different gratings are used, and a similar offset exists between Balmer line fits and hybrid fits that is independent of STIS calibration entirely. These two facts together make it hard to attribute the discrepancy purely to a G230L slope error. The more likely explanation is model opacity systematics, which is what the authors argue. The soft spots are in the He-atmosphere results, not the H-atmosphere ones. The artificial reduction of C I gf values by a factor of ~30 to match UV line strengths is a post-hoc adjustment without independent physical justification, and the authors say so themselves. The DQpec Swan band shift parameter is empirically tuned with acknowledged non-universality. The systematically low DQ masses from UV fitting — inconsistent with optical fits and 2σ off the microlensing mass for WD1142-645 — confirm that DQ models have unresolved problems in the UV. These issues do not invalidate the empirical detections or the H-atmosphere Teff discrepancy, but they mean the DQ parameters and DC carbon limits should be treated as model-dependent preliminary values. This paper is for white dwarf modelers and observers working on cool atmospheres and planetary accretion demographics. It deserves a serious referee who can evaluate the atmosphere model details and push the authors to strengthen the DQ discussion or scope it down. Recommend accept for peer review.","headline":"UV-optical Teff discrepancy for cool DAs is a genuine new finding; DQ UV modeling issues are real but secondary","tokens_in":49709,"tokens_out":634,"would_cite":true,"duration_ms":189019,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"glm-5.2","headline":"UV spectra expose a systematic flaw in white dwarf temperature models","keywords":["white dwarfs","ultraviolet spectroscopy","model atmospheres","effective temperature","planetary debris accretion","spectral energy distribution","local stellar neighborhood"],"falsifier":"If the 2–6 percent UV-versus-optical temperature offset were an artifact of STIS flux calibration rather than a model deficiency, it would disappear when independently calibrated UV spectra (e.g., from a different instrument or epoch) are fitted with the same models. Conversely, if improved opacity calculations incorporating the H3+ correction and updated Lyman-alpha profiles eliminate the offset, the discrepancy would be resolved as a modeling gap rather than new physics.","tokens_in":48673,"feed_emoji":"紫外","tokens_out":1291,"duration_ms":205334,"temperature":0.7,"pith_summary":"This paper analyzes Hubble Space Telescope ultraviolet spectroscopy of all 44 confirmed white dwarfs within 13 parsecs of the Sun, combining UV data with optical and infrared photometry to determine stellar parameters. The central finding is that for hydrogen-atmosphere white dwarfs cooler than 10,000 K, fitting the UV spectrum yields effective temperatures 2–6 percent higher than fitting optical and infrared photometry alone. This offset is not a statistical fluctuation or a calibration artifact; it persists after accounting for flux calibration uncertainties and after switching between 1D and 3D model atmospheres. The authors conclude that current white dwarf atmosphere models cannot self-consistently reproduce the full ultraviolet-to-infrared spectral energy distribution for these stars. The UV data also reveal photospheric metals in six white dwarfs where optical spectra show nothing, reclassify three helium-rich stars as metal-enriched based on UV magnesium detections, and find that 30 percent of the nearest white dwarfs show spectroscopic evidence of accreted planetary debris. The paper further documents that UV carbon line strengths in DQ white dwarfs are overpredicted by models by roughly a factor of thirty, requiring artificial suppression of oscillator strengths to match observations, and that UV fitting of DQs produces systematically low masses inconsistent with optical fits and microlensing measurements.","feed_headline":"UV spectra reveal white dwarf models can't fit the full spectrum","feed_subtitle":"Fitting UV data raises nearby white dwarf temperatures 2–6 percent, exposing a systematic gap in atmosphere models for the Sun's coolest恒星残骸","key_machinery":"The central mechanism is the hybrid spectrophotometric fit: flux-calibrated HST/STIS UV spectra are combined with Gaia, 2MASS, and WISE photometry and compared to model atmosphere grids. By running the fit twice—once with UV data included and once without—the authors isolate a wavelength-dependent systematic offset in the best-fit effective temperature, surface gravity, and mass. The discrepancy is attributed to inadequacies in UV opacity sources (Lyman-alpha broadening, collision-induced absorption, H3+ partition functions) rather than observational error.","core_discovery":"The core discovery is a systematic, model-level discrepancy in cool hydrogen-atmosphere white dwarfs: incorporating UV spectra into the fit raises the derived effective temperature by 2–6 percent compared to optical/IR-only fits, and no existing atmosphere model can simultaneously match the UV and optical/infrared continuum. This is demonstrated on a complete, volume-limited sample of the 44 nearest white dwarfs, where the UV data also expose previously hidden metals, planetary debris signatures, and failures in carbon line modeling.","pith_inferences":["If the UV-optical temperature offset stems from an opacity source that is temperature-dependent, it may change sign or magnitude at temperatures above 10,000 K; the paper notes a hint of such a reversal in the hottest DAs, which would further constrain the physical origin of the discrepancy.","The fact that both the H3+ partition function correction and the residual UV-optical discrepancy point to the same cool-temperature regime suggests that additional unmodeled molecular or pressure-dependent opacities in the UV remain to be identified; upcoming JWST near-IR spectroscopy of the same sample could test whether the discrepancy extends into the infrared.","The model-independent microlensing mass for WD 1142-645 (0.56 solar masses) being 2-sigma away from the UV-derived mass (0.47 solar masses) provides an external check that could be exploited more systematically: if more microlensing masses become available for DQ white dwarfs, they could directly quantify the UV model bias without relying on either UV or optical atmosphere fits."],"forward_implications":["If the 2–6 percent temperature offset is real and systematic, derived masses, cooling ages, and space densities for the majority of nearby white dwarfs may need revision, affecting Galactic archaeology and stellar initial-to-final mass relations.","The overprediction of UV carbon line strengths by a factor of ~30 in DQ models indicates a fundamental gap in atomic data or line-broadening theory for dense helium plasmas, which biases DQ mass determinations low when UV data are included.","Six metal-polluted white dwarfs were identified solely through UV spectroscopy, meaning optical-only surveys underestimate the fraction of white dwarfs accreting planetary debris, currently measured at 30 percent for the local volume.","The 32 percent multiplicity fraction for nearby white dwarfs, combined with the difficulty of detecting close white dwarf companions at larger distances, suggests that binary fraction estimates from magnitude-limited surveys are systematically incomplete."],"fun_headline_variants":["UV data raises temperatures of nearby cool white dwarfs by up to 6%","Existing white dwarf models fail to fit UV and optical data simultaneously","UV spectra expose temperature gap in cool nearby white dwarf models","Optical-only fits underestimate temperatures of nearby cool white dwarfs","Hubble UV data uncover temperature gaps and hidden metals in white dwarfs"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The analysis assumes that at least one of the two fitting regimes—UV-inclusive or optical/IR-only—reliably recovers the true stellar parameters, but the paper itself shows that both regimes have unresolved model issues: UV carbon lines require artificial weakening by a factor of ~30, UV-derived DQ masses conflict with microlensing measurements, and the optical/IR fits may suffer from unmodeled infrared opacities. The central temperature-offset claim is robust as a detection, ","fun_headline_variants_meta":{"raw":{"variants":["UV data raises temperatures of nearby cool white dwarfs by up to 6%","Existing white dwarf models fail to fit UV and optical data simultaneously","UV spectra expose temperature gap in cool nearby white dwarf models","Optical-only fits underestimate temperatures of nearby cool white dwarfs","Hubble UV data uncover temperature gaps and hidden metals in white dwarfs","UV spectroscopy reveals systemic model gaps in nearest white dwarfs"]},"model":"glm-5.2","effort":"high","cost_usd":0.0,"raw_usage":{"total_tokens":1359,"prompt_tokens":585,"completion_tokens":774,"prompt_tokens_details":null},"tokens_in":585,"tokens_out":774,"duration_ms":66522,"temperature":1.0,"reasoning_tokens":771,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-08T08:21:55.873065+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"If the 2–6 percent UV-versus-optical temperature offset were an artifact of STIS flux calibration rather than a model deficiency, it would disappear when independently calibrated UV spectra (e.g., from a different instrument or epoch) are fitted with the same models. Conversely, if improved opacity calculations incorporating the H3+ correction and updated Lyman-alpha profiles eliminate the offset, the discrepancy would be resolved as a modeling gap rather than new physics.","supporting_citations":[],"review_version":1}