{"id":"df694fdb-1b36-45d3-8921-2b4ff76f0a6e","arxiv_id":"2506.16769","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"InPTA's second data release provides reprocessed times of arrival, dispersion measures, and timing ephemerides for 27 millisecond pulsars observed with uGMRT over 7.5 years.","lead":"The Indian Pulsar Timing Array has released seven years of radio timing data for 27 millisecond pulsars, including some of the highest precision dispersion measure measurements reported. The release is designed to sharpen future nanohertz gravitational wave searches when combined with other pulsar timing arrays.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Table 3 DM uncertainties appear to be raw formal errors, not EFAC-calibrated: for J1939+2134 the stated 0.3e-5 pc/cm^3 precision would inflate by roughly 14x under the paper's own T2EFAC, directly weakening the 'highest precision DM estimates' claim.","rationale":"The reader's weakest assumption was that a fixed template built from one high-S/N epoch remains valid across the full 7.5-year baseline, with the J1713+0747 profile event as the cautionary example. That is a genuine accuracy risk and could bias both ToAs and DMs. I flag a different, more directly advertised weakness: the precision numbers in Table 3 are not calibrated by the paper's own EFAC analysis, so the 'highest precision DM estimates' claim is not yet demonstrated. The two concerns are related, since template errors would also inflate the apparent scatter and hence the fitted EFACs, but they are distinct: the reader's concern is about bias, mine is about overstated precision. My proposed test is cheap, decisive, and uses only quantities already in the paper: rescale the DM errors with the quoted T2EFAC values and see whether Table 3 survives. If the scaled errors are substantially larger, the headline precision claim and the significance of some solar-wind DM variations need qualification. I retain the reader's CONDITIONAL verdict because the dataset can still be useful for IPTA even with raw formal errors, provided the precision claim is recalibrated or explicitly caveated, and provided the dataset is actually released publicly.","tokens_in":32065,"tokens_out":8756,"duration_ms":100631,"concrete_test":"For at least two pulsars, e.g., J1939+2134 and J1909-3744, recompute the epoch-wise B3+5 DM uncertainties after multiplying each sub-banded ToA error by the corresponding T2EFAC value from Section 5, or re-fit the DM with the error-scaled ToA file. Compare the new median/min DM uncertainties to Table 3. If the scaled values exceed the quoted values by more than a factor of two (and by roughly 14 for J1939+2134), then the paper should either downgrade the 'highest precision' wording or explicitly label Table 3 as raw formal errors pending noise calibration. A complementary check is to inspect the DR2 code on the public repository to confirm whether any EFAC scaling is already applied inside DMCalc before the Table 3 entries are generated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in the abstract is a precision claim: the reprocessed dataset provides 'some of the highest precision DM estimates so far' and enables 'interesting solar wind related DM variations.' The quantitative support for this claim is Table 3, which lists median and minimum DM uncertainties as low as 0.3e-5 pc/cm^3 for J1909-3744 and J1939+2134. However, Section 5 shows that the sub-banded ToAs on which these DM fits are based require T2EFAC scaling factors of 0.3-7 (and about 14 for J1939+2134 band 3) to bring reduced chi-squared to unity. Section 4.4, which describes how the Table 3 DM uncertainties are produced, mentions no EFAC or equivalent white-noise calibration, and Section 5 explicitly defers proper white- and red-noise modeling to a later paper. Since a DM is the frequency slope of sub-banded ToAs, its fitted uncertainty scales linearly with the ToA error scaling: a 14x T2EFAC would turn J1939+2134's quoted 0.3e-5 pc/cm^3 into roughly 4e-5 pc/cm^3, which is not obviously better than other PTAs' published DM precisions. The same issue affects the significance of the claimed solar-wind DM variations, since those detections use the same DM error bars. The concern is not that the processing is wrong, but that the headline 'highest precision' assertion is supported by uncalibrated formal errors rather than validated measurement uncertainties.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents InPTA Data Release 2, a reprocessed seven-year dataset of sub-banded times of arrival (ToAs), epoch-wise dispersion measures (DMs), and initial timing ephemerides for 27 millisecond pulsars observed with the uGMRT in two simultaneous frequency bands (300–500 MHz and 1260–1460 MHz). The authors describe their full pipeline: noise-free template generation, per-pulsar optimization of sub-band count, iterative estimation of a fiducial DM, DMX-based deterministic timing, and the production of DM time series. The central claims are that the dual-band, low-frequency dataset provides some of the highest-precision DM estimates published so far and that annual DM variations attributable to the solar wind are visible in several pulsars. The paper is written as a data-release paper, with detailed documentation of the observing setup, exclusions, outlier flagging, and known limitations such as the J1643–1224 scatter-broadening bias and the J1713+0747 profile change.","tokens_in":32457,"tokens_out":3374,"duration_ms":37183,"significance":"If the precision and calibration claims hold, this dataset is genuinely valuable for the PTA community: it adds a low-frequency, high-cadence, simultaneous dual-band component to IPTA DR3, and its per-epoch DMX measurements could improve chromatic-noise modeling in combined GW searches. The paper is unusually transparent about processing choices: band-4 exclusion, J0900–3144 band-3 exclusion, J2302+4442 band-5 exclusion, J1713+0747 truncation at MJD 59309, and the use of FDJUMPDM for J1643–1224 are all disclosed. The pipeline is reproducible in principle, with the code and data to be released on GitHub. The main significance risk is that the headline precision claim rests on formal DM uncertainties that are not yet calibrated by the EFAC factors the paper itself reports.","major_comments":[{"comment":"The DM uncertainties quoted in Table 3 are formal errors from a weighted fit to sub-banded ToA errors, but Section 5 states that the same ToAs require T2EFAC scaling factors of 0.3–7, and about 14 for PSR J1939+2134 band 3 with 200 MHz bandwidth and CDP, to bring reduced chi-squared to unity. Because a DM fit is a linear function of the ToA uncertainties, a T2EFAC of 14 would inflate the quoted 0.3e-5 pc/cm^3 minimum for J1939+2134 to roughly 4e-5 pc/cm^3, which is no longer outstanding relative to other PTA DM measurements. The abstract's claim of 'some of the highest precision DM estimates so far' is therefore not yet supported by validated measurement uncertainties. I ask the authors to either propagate the EFACs into the DM uncertainties before quoting them in Table 3 and the abstract, or to explicitly label Table 3 as reporting uncalibrated formal errors and to temper the precision claim accordingly. This also affects the significance of the claimed solar-wind DM variations, which use the same error bars.","section":"Section 4.4 and Table 3 vs. Section 5"},{"comment":"The analysis uses a single high-S/N epoch per pulsar to build the noise-free template that is then cross-correlated with all data across the full 7.5-year baseline. The paper itself acknowledges that the J1713+0747 profile-shape event at MJD 59320–59321 required truncating that pulsar's data, which demonstrates that profile instability is a real hazard for this method. Yet no explicit test of profile stability over time is reported for the other 26 pulsars, and Section 5 argues that FD parameters are unnecessary without a quantitative demonstration that the sub-band selection removes all frequency-dependent profile evolution over the full time span. I recommend adding a stability check, for example by comparing templates built from early and late epochs or by examining per-epoch profile residuals, or at minimum a clear statement of the implied systematic risk for the DM and ToA precision claims.","section":"Section 4.1 and Section 6"},{"comment":"The fiducial DM is estimated from the same template-epoch data that is used to align the templates, and all DMX values are then measured relative to this fiducial DM. Any systematic error in the fiducial DM is a constant offset that does not affect timing residuals or GW sensitivity, so this is not circular for the timing products. However, the DM time series and the absolute DM precision claims in Table 3 do depend on the fiducial DM uncertainty, which is not reported anywhere. I ask the authors to state the uncertainty on each fiducial DM, or to explicitly note that all quoted DM uncertainties are relative to the fiducial value.","section":"Section 4.3 and Section 4.4"}],"minor_comments":[{"comment":"In the Figure 6 panel for PSR J2124–3358, the fiducial DM is listed as 24.35335, which is identical to the value listed for J1944+0907 and differs from the value 4.59288 shown in Figure 33 of Appendix 2; this appears to be a typo and should be corrected.","section":"Figure 6 and Appendix 2"},{"comment":"The manuscript includes a Figure 13 without a caption in the provided text; please ensure every figure has a caption and that cross-references are correct.","section":"Appendix 2"},{"comment":"The sentence beginning 'In the case of PSRs J0613−0200, the scaling factor decreases to approximately 0.2...' has a grammatical issue ('In the case of PSRs J0613−0200' should be 'In the case of PSR J0613−0200') and should be reworded.","section":"Section 5"},{"comment":"Equation (1) is dimensionally inconsistent as written (S/N is set equal to an expression involving T_peak/T_sys with non-obvious units); please add a short definition of each symbol and state that this is a proportionality relation rather than an exact equality.","section":"Section 4.2 and Appendix 1"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid data-release description with unusually candid handling of systematics. The main issue is that the central precision claim is supported by formal rather than calibrated uncertainties; this is fixable within the manuscript's scope and does not require new observations. I would not reject on the basis of the EFAC issue because the dataset itself is clearly valuable and the authors already have the T2EFAC values needed to rescale. One editorial observation for the editor: the reference list is dominated by InPTA's own prior papers, but this is common in data-release papers and I do not view it as a problem for the recommendation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the take: this is a genuinely useful data-release paper, but the abstract's precision claim is ahead of the actual error budget. The stress-test note holds up. Table 3 lists DM uncertainties down to 0.3e-5 pc/cm^3 for J1909-3744 and J1939+2134, but those come from the DMCalc fits before any EFAC scaling. The paper itself reports T2EFAC factors up to ~7 and ~14 for J1939+2134 band 3, and explicitly defers white-noise modeling to a later paper. Since the DM uncertainty is fit from sub-banded ToA errors, it scales roughly linearly with those error bars: a 14x EFAC turns the quoted 0.3e-5 into ~4e-5 pc/cm^3, which is no longer obviously better than other PTAs' published DM precision. The solar-wind DM variation detections use the same error bars, so their significance is also provisional. This is the one load-bearing weakness. It doesn't invalidate the data release—the ToAs, ephemerides, and DM time series are still valuable—but the 'highest precision' claim should be reworded or supported by calibrated uncertainties.\n\nWhat's actually new: a 7.5-year reprocessed dataset for 27 MSPs with sub-banded ToAs, epoch-wise DMs, and timing ephemerides, extending InPTA DR1. The methodological increments are genuine: band equalization in template generation, a more formal sub-band optimization with Anderson-Darling tests, and reprocessing early PRESTO epochs with PINTA. The paper is transparent about exclusions (band 4, J0900-3144 band 3, J2302+4442 band 5, J1713+0747 truncation) and names the J1643-1224 scatter-broadening bias and its mitigation with FDJUMPDM. That is good practice and I believe the processing is internally consistent.\n\nThe other soft spots are minor. The single-template assumption across 7.5 years is a real risk, but the paper acknowledges it and the J1713 event shows they handle profile changes sensibly. The fiducial DM being derived from the template epoch itself is a calibration convention, not circular reasoning. Self-citation of InPTA's own pipeline is expected for a data-release paper. One practical issue: the dataset is promised at a GitHub URL but not verifiably public yet; acceptance should be conditional on the release actually being there and the precision claim being fixed.\n\nBottom line: this paper deserves a serious referee. It is a substantive contribution to IPTA DR3, and the flaws are fixable with a comparison table of DM uncertainties against prior published values and a clear statement about EFAC calibration. I'd want the revised version before trusting the 'highest precision' language, but the data themselves are worth having.","headline":"Solid InPTA data release, but the headline DM precision claim rests on uncalibrated formal errors; fix that and the dataset is a strong IPTA DR3 input.","tokens_in":33257,"tokens_out":3349,"would_cite":true,"duration_ms":30835,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The Indian Pulsar Timing Array's second data release reprocesses 7.5 years of dual-band observations of 27 millisecond pulsars into sub-banded times of arrival and epoch-resolved dispersion measures, reporting some of the most precise…","keywords":["pulsar timing array","millisecond pulsars","dispersion measure","times of arrival","uGMRT","solar wind","interstellar medium","gravitational wave background"],"falsifier":"Re-time the full dataset for any one pulsar using a template built from the latest epoch instead of the selected template epoch; if the resulting ToA and DM time series shift by more than the quoted uncertainties or develop new annual artifacts, the single-template assumption fails. A sharper check is to search the released residuals of PSR J1713+0747 for an unmodeled step at MJD 59309, the truncation point imposed by its profile-change event, since an unmodeled step there would show that the dataset is not internally consistent.","tokens_in":31879,"feed_emoji":"📡","tokens_out":8906,"duration_ms":81087,"temperature":0.7,"pith_summary":"This paper is a data release: it presents 7.5 years of simultaneous 300–500 MHz and 1260–1460 MHz observations of 27 millisecond pulsars, reprocessed end to end into sub-banded times of arrival, dispersion measures, and timing ephemerides. The central claim is that this reprocessing yields some of the most precise dispersion-measure estimates ever published for these pulsars, with uncertainties reaching $10^{-6}$ pc cm$^{-3}$, and that the resulting epoch-resolved DMs remove interstellar propagation noise well enough to sharpen pulsar-timing-array sensitivity to nanohertz gravitational waves. The paper also reports annual solar-wind DM variations in five pulsars and a possible discrete DM jump in one pulsar, and it argues that the dataset is ready to be folded into the next combined international pulsar timing data release. A sympathetic reader would care because cleaner dispersion measures directly attack the main chromatic noise source that limits current gravitational-wave background searches.","feed_headline":"Pulsar array release pushes dispersion measures to record precision","feed_subtitle":"Cleaner dispersion measures mean less interstellar noise in the hunt for nanohertz gravitational waves.","key_machinery":"The load-bearing mechanism is the combination of a noise-free, frequency-resolved template built from one carefully selected high-S/N epoch per pulsar, a per-pulsar optimization of the number of frequency sub-bands, and the DMCalc pipeline that cross-correlates each epoch's sub-banded data against that template to produce sub-banded ToAs and epoch-wise DMs. A fiducial DM, estimated iteratively by aligning concurrent band 3 and band 5 template epochs, fixes the reference for both bands so that frequency-dependent delays are measured rather than modeled. The epoch-wise DMs are written into the ephemeris as DMX values representing instantaneous measured DMs, which is conceptually different from the usual piecewise-linear DMX fitting, and this is what lets the timing analysis absorb interstellar dispersion without fitting extra frequency-dependent parameters.","core_discovery":"The paper establishes that simultaneous dual-band timing at 300–500 MHz and 1260–1460 MHz, with a single high-signal-to-noise template epoch per pulsar and per-epoch dispersion-measure fits, can produce dispersion-measure time series precise to parts in $10^{-5}$ to $10^{-6}$ pc cm$^{-3}$ for many millisecond pulsars. It shows that applying these epoch-resolved DMs directly as DMX delays, rather than as piecewise-linear fits, leaves timing residuals with post-fit weighted RMS as low as 1.095 microseconds for PSR J1909−3744. It further claims that the reprocessed dataset cleanly reveals annual solar-wind electron-content variations in PSRs J0034−0534, J0613−0200, J1744−1134, J1909−3744, and J2145−0750, and a candidate DM jump in PSR J1125+7819 near MJD 59741. On this basis the paper argues that the data release will characterize interstellar-medium noise better than the previous release and increase the sensitivity of upcoming nanohertz gravitational-wave searches.","pith_inferences":["If the claimed dispersion-measure precision transfers to other low-frequency arrays, the same dual-band epoch-DM scheme could be adopted there to separate interstellar and solar-wind delays without relying on external solar wind models.","The five pulsars with annual solar-wind signatures provide a testable solar-wind monitor: comparing their DM curves with spacecraft-based solar-wind density measurements should either confirm the calibration or reveal residual unmodeled delays.","A concrete prediction for the next combined data release is that pulsars with the largest chromatic noise, such as PSR J1909−3744, should show the greatest improvement in gravitational-wave detection statistics once these DMs are included.","The candidate DM jump in PSR J1125+7819 at MJD 59741, if confirmed by independent telescopes, would be a rare probe of compact structure in the ionized interstellar medium."],"forward_implications":["The released ToAs, DMs, and ephemerides for 27 pulsars enter the next combined international data release, so gravitational-wave searches inherit a better-characterized interstellar-medium noise term.","Epoch-wise DMX values measured from simultaneous dual-band data replace fitted piecewise-linear DM models, which should remove a class of chromatic timing artifacts in the residuals.","Five pulsars show annual solar-wind DM modulations, giving a direct multi-year record of solar-wind electron content along those lines of sight.","The per-pulsar sub-band optimization eliminates the need for frequency-dependent timing parameters even at low radio frequencies.","The template-epoch strategy means any future pulse-profile shape-change event, like the one that truncates PSR J1713+0747 data at MJD 59309, must be detected and handled explicitly to preserve the precision claims."],"supporting_citations":[{"why":"The prior InPTA data release whose template-generation and fiducial-DM procedure this paper extends and whose data it reprocesses.","marker":"Tarafdar et al. (2022)"},{"why":"Defines DMCalc, the pipeline that produces the sub-banded ToAs and epoch-wise DMs central to the release.","marker":"Krishnakumar et al. (2021)"},{"why":"Defines PINTA, the standardized data-reduction pipeline used to bring the oldest observations to a common format.","marker":"Susobhanan et al. (2021)"},{"why":"Supplies the coherent dedispersion pipeline used for band 3 data so that low-frequency profiles are not smeared by dispersion.","marker":"De & Gupta (2016)"},{"why":"Provides the Fourier-domain Markov-chain Monte Carlo cross-correlation method used to estimate ToAs and earlier ephemerides for folding.","marker":"Verbiest et al. (2016)"},{"why":"Explains and mitigates the scatter-broadening bias that produces the band 3 versus band 3+5 DM offset seen for PSR J1643−1224.","marker":"Singha et al. (2024)"},{"why":"One of the most recent ephemeris products used as a starting point for the timing solutions in this release.","marker":"EPTA Collaboration et al. (2023b)"}],"fun_headline_variants":["Dual-band pulsar timing achieves record dispersion-measure precision","InPTA's second release: 27 pulsars, 7 years, sharper DM time series","Solar wind effects on five pulsars revealed in InPTA's updated dataset","Dual-band uGMRT timing yields sub-microsecond residual pulsar fits","New InPTA data release sharpens interstellar medium noise for pulsar timing"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes that one noise-free pulse template built from a single high-signal observation stays a valid cross-correlation reference for the full 7.5-year baseline; if a pulsar's pulse shape drifts or jumps, as happened for PSR J1713+0747 in April 2021, the measured arrival times and dispersion measures inherit that drift.","fun_headline_variants_meta":{"raw":{"variants":["Dual-band pulsar timing achieves record dispersion-measure precision","InPTA's second release: 27 pulsars, 7 years, sharper DM time series","Solar wind effects on five pulsars revealed in InPTA's updated dataset","Dual-band uGMRT timing yields sub-microsecond residual pulsar fits","New InPTA data release sharpens interstellar medium noise for pulsar timing"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000589,"raw_usage":{"total_tokens":2803,"prompt_tokens":1023,"completion_tokens":1780,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":639,"completion_tokens_details":{"reasoning_tokens":1676}},"tokens_in":639,"tokens_out":1780,"duration_ms":15302,"temperature":1.0,"reasoning_tokens":1676,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:20:29.058745+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-time the full dataset for any one pulsar using a template built from the latest epoch instead of the selected template epoch; if the resulting ToA and DM time series shift by more than the quoted uncertainties or develop new annual artifacts, the single-template assumption fails. A sharper check is to search the released residuals of PSR J1713+0747 for an unmodeled step at MJD 59309, the truncation point imposed by its profile-change event, since an unmodeled step there would show that the dataset is not internally consistent.","supporting_citations":[{"cited_title":"2022, , 39, e053","cited_arxiv_id":null,"evidence_quote":"The prior InPTA data release whose template-generation and fiducial-DM procedure this paper extends and whose data it reprocesses."},{"cited_title":"A., Manoharan , P","cited_arxiv_id":null,"evidence_quote":"Defines DMCalc, the pipeline that produces the sub-banded ToAs and epoch-wise DMs central to the release."},{"cited_title":"C., et al","cited_arxiv_id":null,"evidence_quote":"Defines PINTA, the standardized data-reduction pipeline used to bring the oldest observations to a common format."},{"cited_title":"P., Lentati, L., Hobbs, G., et al","cited_arxiv_id":null,"evidence_quote":"Provides the Fourier-domain Markov-chain Monte Carlo cross-correlation method used to estimate ToAs and earlier ephemerides for folding."},{"cited_title":"C., Krishnakumar , M","cited_arxiv_id":null,"evidence_quote":"Explains and mitigates the scatter-broadening bias that produces the band 3 versus band 3+5 DM offset seen for PSR J1643−1224."}],"review_version":2}