{"id":"dd2e9b19-b065-496f-beb5-07ebf607b362","arxiv_id":"1908.01636","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Ground calibration and end-to-end simulations predict CHEOPS will measure Earth-size and Neptune-size transit radii with 5% and 2% precision, meeting mission requirements.","lead":"CHEOPS, an ESA exoplanet follow-up telescope, reached about 15 parts per million photometric stability in ground tests and is expected to measure Earth-size and Neptune-size transit depths with signal-to-noise of 10 and 25. The paper is a pre-launch calibration report that translates lab measurements and end-to-end simulations into mission performance predictions.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 5% Earth-radius precision assumes no stellar granulation/activity; for a V=9 G2V these sources add tens of ppm over hours, comparable to the 100 ppm transit depth, so the quoted science precision is likely an optimistic instrument-only bound.","rationale":"The reader's weakest assumption identifies the same load-bearing issue: the simulated precision numbers omit stellar granulation and activity, which are not negligible for a V=9 G2V at the 10 ppm level claimed for the Earth-transit case. This is the single most direct control on the abstract's headline 5% Earth-size precision, since the residual in Case 1 is 10.2 ppm and an Earth transit depth is only about 100 ppm. The paper itself acknowledges the omission in Section 5.1.2 and the simulator has the capability to include such effects, but they were switched off, making the quoted number a best-case instrument limit. The reader's other concern, the ground-calibration 15 ppm noise floor, is real but secondary: the radii claims come from the simulator, not from that measurement, and the simulator's residual is already lower than 15 ppm. A concrete simulation test with granulation and spot noise would settle whether the 5% claim degrades materially; this is an addressable correctness risk rather than an internal inconsistency. The paper is otherwise careful, with an independent DRP cross-check supporting internal consistency, so the appropriate verdict remains CONDITIONAL as the reader concluded.","tokens_in":15383,"tokens_out":4629,"duration_ms":47383,"concrete_test":"Rerun the Case-1 CHEOPSim transit with realistic stellar noise: add a granulation component with about 20 ppm rms and a 10-minute correlation time (or a solar granulation power spectrum scaled to V=9 in the CHEOPS passband), plus a spot modulation of about 50 ppm with a 25-day rotation period and a finite spot lifetime, then repeat the same MCMC radius recovery. If the 68% credible interval on Rp/R* exceeds about 10%, the abstract's 5% precision claim must be qualified as instrument-only or restricted to a quieter/fainter target.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claims are the 5% Earth-radius and 2% Neptune-radius precisions from end-to-end simulations, but those simulations are not end-to-end in the astrophysical sense: Table 1 sets `Activity none` and Section 5.1.2 explicitly states `Stellar photosphere effects, like spots or granulation, were not considered`. For a V=9 G2V observed in the 330-1100 nm CHEOPS passband, solar-type granulation contributes roughly 10-30 ppm rms on timescales from minutes to hours, and active regions add quasi-periodic variability of tens to hundreds of ppm. The Case-1 fit residual is only 10.2 ppm against a 100 ppm Earth transit depth; adding even 20 ppm of stellar noise in quadrature raises the depth uncertainty to about 22 ppm, degrading Rp/R* precision from about 5% to 10% or more. The Neptune case (2550 ppm depth, 51.7 ppm residual) is much less sensitive, so the 2% claim is comparatively robust. The 5% Earth-size claim is therefore an instrument-plus-photon-noise lower bound, not a predicted science precision, and the abstract states it without the stellar-noise caveat. The ground-calibration 15 ppm measurement is a separate concern; even if accepted, it does not rescue the Earth-size number because the simulator's 10.2 ppm residual is already below the measured ground stability.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports on ground-based calibration of the CHEOPS payload and uses those calibration products to predict in-flight photometric performance. From a 27.35-hour uniformly illuminated sequence, the authors derive a 15 ppm photometric stability floor over 5 hours after correcting for source variability and a temperature-dependent optical-table effect. They then use the CHEOPSim end-to-end simulator, fed with the measured calibration products, to simulate transits of an Earth-size planet around a V=9 G2V star and a Neptune-size planet around a V=12 K5V star. Aperture photometry, background estimation, centroiding, and MCMC transit fits yield quoted precisions on Rp/R* of 5% and 2%, respectively, with SNR on transit depth of 10 and 25. The analysis is compared to the independent CHEOPS data reduction pipeline and found consistent.","tokens_in":15595,"tokens_out":5869,"duration_ms":58156,"significance":"If the central claims hold, the paper demonstrates that CHEOPS meets its top-level photometric requirements: 15 ppm on-ground stability, 20 ppm in 6 hours for an Earth-size transit, and 85 ppm in 3 hours for a Neptune-size transit. These numbers are important for the exoplanet community because they set expectations for the precision of radii measurements from CHEOPS follow-up. The work has real strengths: the ground calibration is detailed (flat-field synthesis from narrow-band filters, gain/dark/bias stability, PSF characterization), the end-to-end simulation uses calibration-derived detector parameters rather than ad hoc values, and the comparison with an independent pipeline adds confidence. However, the headline precision numbers are presented without a prominent caveat that the simulated noise model omits stellar granulation and activity, and the 15 ppm floor is measured after fitting a temperature correction to the same dataset. The abstract's wording therefore overstates what is demonstrated.","major_comments":[{"comment":"The end-to-end simulation sets 'Activity none' and explicitly excludes spots and granulation, and Section 5.1.2 also omits cosmic rays and smearing. The quoted 5% precision on Rp/R* for an Earth-size planet orbiting a V=9 G2V star is therefore an instrument-plus-photon-noise lower bound, not a predicted science precision. For a solar-type star, granulation contributes roughly 10-30 ppm rms on timescales from minutes to hours, which is comparable to the ~100 ppm Earth-transit depth. Adding only 20 ppm of stellar noise in quadrature to the reported 10.2 ppm residual raises the depth uncertainty to about 22 ppm, degrading the expected radius precision from 5% to roughly 10%. The abstract and Section 6 present the 5% as an expected in-flight performance without this caveat. The Neptune 2% claim is more robust because the 2550 ppm depth is much larger, but it is affected by the same omission. I recommend either adding stellar noise to the simulation or explicitly relabeling the quoted precisions as instrument-limited.","section":"Section 5.1.2, Table 1"},{"comment":"The temperature-flux decorrelation coefficient used to correct the light curve is fit to the same 27.35-hour dataset from which the 15 ppm stability is then measured, and Section 4.3 states that the assumption allowing the lamp-variation segments to be discarded could not be confirmed by a repeated measurement. This is an in-sample fit: the correction is not validated on independent data, so it may absorb part of the correlated noise and bias the reported floor low. The 15 ppm value should be presented as a conditional, bench-corrected lower limit with an explicit statement that no independent confirmation was obtained, or it should be supported by a validation subset left out of the fit. The current wording of the abstract ('on-ground photometric stability ... is found to be of the order of 15 parts per million') does not convey this caveat.","section":"Sections 4.2.3 and 4.3"},{"comment":"The simulated residual noise for the Earth-size case is 10.2 ppm over 6 h, which is below the ground-measured stability of 15 ppm over 5 h (full-frame, after discarding lamp variations) and 20 ppm in the 8x8 sub-aperture extraction. The paper does not explain how the in-flight simulation yields a lower noise than the ground-calibrated floor. If the ground floor is dominated by bench effects that are absent in orbit, that should be stated explicitly; otherwise the simulator appears to omit a correlated-noise component that the ground data show is present. This discrepancy is load-bearing for the credibility of the simulated precision numbers and should be addressed directly.","section":"Section 5.3 vs Section 4.3"}],"minor_comments":[{"comment":"The phrase 'by mean of end-to-end simulation' should read 'by means of end-to-end simulation.'","section":"Abstract"},{"comment":"In Case 2, the planet radius is typeset as '1 RÈ', which appears to be a LaTeX error; the text states a Neptune-size planet, so the value should be specified in Earth radii or equivalent.","section":"Table 1"},{"comment":"The sentence 'the three-dimension nature of the atmospheres has been taken into account' should read 'the three-dimensional nature of the atmospheres.'","section":"Section 5.1.2"},{"comment":"The caption contains a typo: 'obliqblack dashed line' should be 'oblique black dashed line.'","section":"Figure 10 caption"},{"comment":"The noise formula using 'ne−' is ambiguous; define the symbol as the total number of electrons in the photometric aperture, not the per-pixel electron count.","section":"Section 4.3"},{"comment":"The conclusion states that the results 'cover all the effects related to the instrument' but then lists in-orbit effects expected to have marginal impact; a sentence reconciling the simulated 10.2 ppm with the measured 15 ppm floor would help the reader interpret the headline numbers.","section":"Section 6"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a calibration/performance paper from the CHEOPS instrument team, and the companion papers (CHEOPSim and DRP) are cited as in press. The referee report is based solely on the submitted text. The strongest concern is that the abstract's headline numbers (5% Earth-size radius precision and 15 ppm stability) are not qualified by the acknowledged limitations: no stellar noise in the simulation and an in-sample temperature decorrelation fit. These are fixable by rewording the claims and possibly by adding a stellar-noise term to the simulation, so the paper is a major revision rather than a reject. I would also encourage the authors to explicitly discuss the apparent inconsistency between the 10.2 ppm simulated residual and the measured 15-20 ppm ground floor."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth reading. This is the pre-launch CHEOPS calibration paper, giving the first public quantitative performance predictions for the mission. The ground-calibration work is careful: flat-field synthesis validated at 0.07%, bias/dark/gain stability characterized, a 27-hour uniformly illuminated sequence analyzed with Allan variance, and a clear end-to-end simulation (CHEOPSim) cross-checked against the DRP. The 15 ppm/5 h stability and the 20 ppm/5 h aperture-window numbers are credible as instrument floor measurements, and the paper is unusually candid about the lab's own systematic problems (temperature-sensitive feedback fiber, open-loop operation, inability to repeat runs).\n\nThe main quantitative claims hold up as instrument performance. The simulated transit fits give 10.2 ppm residual for the Earth case and 51.7 ppm for the Neptune case, consistent with photon noise plus known calibration terms. The 2% Neptune radius precision is robust, and the DRP cross-check (9.1/51.6 ppm) is good validation. So the paper deserves a serious referee.\n\nSoft spots, in proportion. First, the abstract and conclusions state that planet-to-star radius ratios can be measured to 5% for an Earth-size planet without the caveat that the simulation set Activity=none and explicitly excluded spots and granulation (Section 5.1.2). For a V=9 G2V, granulation alone is tens of ppm over an hour, comparable to the 100 ppm Earth transit depth. Add that in quadrature and 5% becomes ~10% or worse. The headline number is an instrument-plus-photon lower bound, not an end-to-end science precision; this should be reframed or flagged. Second, the 15 ppm floor comes after fitting the temperature-flux slope to the same dataset and discarding lamp-variation data; the authors admit they could not repeat the measurement. That is a genuine in-sample correction, though the slope converges with binning and the residual is stable, so it is a caveat rather than a fatal flaw. The Neptune case is much less affected by stellar noise.\n\nCitation pattern is clean. The companion papers (CHEOPSim, DRP) are the mission's own software, which is legitimate in a mission-characterization paper. Bottom line: accept with revision. The instrument performance claims are solid; the science-precision numbers need a stellar-noise caveat or a fuller noise model. For exoplanet and space-instrumentation readers, and for anyone using CHEOPS photometry, this is definitely worth a reading group slot and likely a citation.","headline":"Honest, careful pre-launch calibration report; the instrument floor measurements are credible, but the abstract's 5% Earth-radius precision omits stellar granulation/activity and should be labeled an instrument-only bound.","tokens_in":16307,"tokens_out":1971,"would_cite":true,"duration_ms":19792,"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":"Pre-launch tests show CHEOPS can measure exoplanet radii to 2% and 5%.","keywords":["CHEOPS","exoplanet transit photometry","space telescope calibration","photometric stability","end-to-end simulation","planetary radii","CCD calibration","transit depth precision"],"falsifier":"Observe a known constant bright star with CHEOPS in orbit over 5 hours and measure the photometric noise: if the scatter exceeds the ground-calibrated 15 ppm floor (e.g., >20 ppm) on a star without detected variability, then the simulated transit precision and the 5% Earth-radius claim would not hold in practice.","tokens_in":15097,"feed_emoji":"🛰️","tokens_out":6145,"duration_ms":55285,"temperature":0.7,"pith_summary":"This paper establishes that the CHEOPS space telescope, from ground calibration alone, is stable enough to detect and measure small transiting planets around bright stars. The authors measure an on-ground photometric stability of about 15 parts per million over five hours, and through an end-to-end simulation they predict that CHEOPS can determine planet-to-star radius ratios with 2% precision for a Neptune-size planet around a K-dwarf star and 5% precision for an Earth-size planet around a Sun-like star. These numbers correspond to signal-to-noise ratios on the transit depth of 25 and 10, and they meet the mission's science requirements. The result matters because it shows that a carefully calibrated space photometer can characterize small planets discovered by surveys like TESS, providing the radii needed to turn masses into densities.","feed_headline":"2% for Neptunes, 5% for Earths: CHEOPS clears pre-launch tests","feed_subtitle":"Ground calibration plus end-to-end simulations predict transit-depth signal-to-noise of 25 and 10.","key_machinery":"The central mechanism is the pairing of a comprehensive ground calibration campaign with an end-to-end simulator, CHEOPSim, that turns measured instrument parameters into realistic synthetic science images. The calibration provides flat fields synthesized from narrowband measurements for any stellar spectrum, a gain model with temperature and voltage sensitivities, a measured PSF, and a long-duration stability sequence. CHEOPSim then injects these calibration products into simulated frames that include jitter, background stars, stray light, orbital interruptions, and the detector's readout chains. The photometric analysis extracts aperture photometry, corrects for background and centroid, and fits transits with a quadratic limb-darkening model using MCMC, producing the quoted radius precision.","core_discovery":"The central claim is that CHEOPS meets its pre-launch photometric requirements. Ground measurements using uniformly illuminated frames, corrected for a temperature-dependent calibration lamp drift, yield a residual noise floor of about 15 ppm over 5 hours, and a sub-aperture analysis reaches 20 ppm in 5 hours as required. Feeding the calibration products—flat fields, gain model, PSF, and noise—into the CHEOPSim end-to-end simulator, the authors generate synthetic transit light curves and fit them with a Markov chain Monte Carlo transit model. For a V=9 G2V star hosting an Earth-size planet (100 ppm transit depth), the recovered radius ratio has a 5% uncertainty; for a V=12 K5V star with a Neptune-size planet, the radius ratio is determined to 2%. These simulated residuals, 10.2 ppm and 51.7 ppm respectively, are close to the photon noise limit, and the authors conclude that the instrument is compliant with its design goal of detecting an Earth-size transit around a Sun-like star.","pith_inferences":["The quoted precision assumes a perfectly quiet star; real G2V stars exhibit granulation and starspot noise that can reach tens of ppm on hour timescales, comparable to the 100 ppm Earth transit depth, so the 5% radius precision may degrade when those sources are included.","The temperature-correlated drift discovered in the calibration bench (a feedback-fibre index change) shows that even a grounded 'super-stable' source can introduce systematics; in flight, thermal variations of the telescope or optics could analogously require monitoring with onboard temperature sensors.","If the in-orbit noise floor turns out to be higher than 15 ppm—due to unmodelled stray light, focusing changes in zero gravity, or cosmic-ray hits—the error budget would need revision, but the margin between measured stability and the 20 ppm requirement provides some headroom.","The single-transit detection capability implies CHEOPS could confirm long-period TESS candidates that are observed only once, since the Earth-transit case reaches high signal-to-noise despite gaps."],"forward_implications":["CHEOPS can achieve the mission requirement of detecting an Earth-size transit around a Sun-like star at V=9, since the 10.2 ppm residual is phonton-noise limited and below the 20 ppm budget.","A single transit may be enough to detect an Earth-size planet around a bright Sun-like star, given the high signal-to-noise despite data gaps from orbital interruptions.","Neptune-size planets around K-dwarfs can have their radii measured to 2% from one transit, with the precision limited more by degeneracies with impact parameter than by photon noise.","The independent analysis pipeline produces results consistent with the official CHEOPS data reduction pipeline, indicating the performance does not depend on a particular reduction algorithm.","The flat-field synthesis method allows correcting the detector response for any stellar spectrum to about 0.07% rms, which is sufficient for the required photometric precision."],"supporting_citations":[{"why":"Defines the CHEOPS mission design and the two core science requirements (20 ppm over 6 hours and SNR 30 for a Neptune) against which the simulated performances are judged.","marker":"Broeg et al. 2013"},{"why":"Describes the payload architecture, including the 32-cm telescope and CCD, whose parameters feed the calibration and the CHEOPSim model.","marker":"Fortier et al. 2014"},{"why":"Presents CHEOPSim, the end-to-end simulator that generates the synthetic science frames used to compute the radius precision numbers.","marker":"Futyan et al. 2019"},{"why":"Supplies the official data reduction pipeline whose light curves are compared with the authors' independent extraction, validating the simulation results.","marker":"Hoyer et al. 2019"},{"why":"Reports the ground calibration bench and its integration, the source of the photometric stability measurements and the PSF data.","marker":"Wildi et al. 2015a"},{"why":"Describes the super-stable light source with 3 ppm short-term stability used for the long-duration calibration sequences.","marker":"Wildi et al. 2015b"},{"why":"Provides the CCD gain model and sensitivity to temperature and voltage that are used to correct the measured and simulated data.","marker":"Deline et al. 2017"}],"fun_headline_variants":["CHEOPS pre-launch tests predict 2% Neptune, 5% Earth radius precision","15 ppm stability: CHEOPS passes pre-launch photometry checks","Simulations show CHEOPS ready to detect Earth-size transits","CHEOPS calibration yields 2% Neptune, 5% Earth transit depths","CHEOPS meets requirements: 5% Earth, 2% Neptune radius ratios"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quoted precision numbers assume the target star is completely quiet, with no starspots or granulation noise; the simulations set stellar activity to 'none', so real stellar variability could add noise comparable to the planet's transit signal.","fun_headline_variants_meta":{"raw":{"variants":["CHEOPS pre-launch tests predict 2% Neptune, 5% Earth radius precision","15 ppm stability: CHEOPS passes pre-launch photometry checks","Simulations show CHEOPS ready to detect Earth-size transits","CHEOPS calibration yields 2% Neptune, 5% Earth transit depths","CHEOPS meets requirements: 5% Earth, 2% Neptune radius ratios"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000533,"raw_usage":{"total_tokens":2588,"prompt_tokens":990,"completion_tokens":1598,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":606,"completion_tokens_details":{"reasoning_tokens":1493}},"tokens_in":606,"tokens_out":1598,"duration_ms":10257,"temperature":1.0,"reasoning_tokens":1493,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:07:07.922190+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Observe a known constant bright star with CHEOPS in orbit over 5 hours and measure the photometric noise: if the scatter exceeds the ground-calibrated 15 ppm floor (e.g., >20 ppm) on a star without detected variability, then the simulated transit precision and the 5% Earth-radius claim would not hold in practice.","supporting_citations":[{"cited_title":"2013, in European Physical Journal Web of Conferences, V ol","cited_arxiv_id":null,"evidence_quote":"Defines the CHEOPS mission design and the two core science requirements (20 ppm over 6 hours and SNR 30 for a Neptune) against which the simulated performances are judged."},{"cited_title":"2014, in Space Telescopes and Instrumen- tation 2014: Optical, Infrared, and Millimeter Wave, V ol","cited_arxiv_id":null,"evidence_quote":"Describes the payload architecture, including the 32-cm telescope and CCD, whose parameters feed the calibration and the CHEOPSim model."},{"cited_title":"2019, A&A, (in press) Gaia Collaboration, Prusti, T., de Bruijne, J","cited_arxiv_id":null,"evidence_quote":"Presents CHEOPSim, the end-to-end simulator that generates the synthetic science frames used to compute the radius precision numbers."},{"cited_title":"2019, A&A, (in press)","cited_arxiv_id":null,"evidence_quote":"Supplies the official data reduction pipeline whose light curves are compared with the authors' independent extraction, validating the simulation results."},{"cited_title":"2017, in Society of Photo- Optical Instrumentation Engineers (SPIE) Conference Series, V ol","cited_arxiv_id":null,"evidence_quote":"Provides the CCD gain model and sensitivity to temperature and voltage that are used to correct the measured and simulated data."}],"review_version":1}