{"id":"dc275605-deda-4037-86ab-75bc0a1081b3","arxiv_id":"2501.12460","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A simulation predicts that paired-visit scheduling could double WFST's known-NEO tracklet yield and increase blind NEO discoveries by about half, yielding roughly 1,800 tracklet-capable and 600+ discovered NEOs per year.","lead":"This paper simulates a year of observations with China's Wide Field Survey Telescope and shows that a simple scheduling change, observing each sky tile twice per night, could roughly double the number of near-Earth asteroids for which the telescope can report tracklets. The result matters because it offers a concrete, low-cost way to improve planetary-defense monitoring without changing the telescope's other science goals.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The absolute discovery counts hinge on an unmeasured transfer of LSST's trailing-loss coefficient c=0.42 to WFST; the paper itself flags this as uncertain, so the 1800/600 predictions need a quantitative sensitivity check.","rationale":"The reader's weakest-assumption analysis identifies exactly the same load-bearing concern: the LSST-derived trailing-loss coefficient c=0.42 is applied to WFST without measurement (Section 2.3.1, Equations 2-3), and it enters before the scheduling comparison, with the paper itself acknowledging possible error. My reading of the full text confirms this is the most consequential unverified input. The paper's own sensitivity discussion in Section 4 demonstrates that trailing loss changes the recovered NEO counts by a factor of several, so the uncertainty in c is not a small correction; it directly affects the absolute claims of about 1800 tracklets and more than 600 blind-search discoveries, especially in the H=22-25 range where fast, faint objects dominate. I also considered the alternative concern that the 'found' count depends on the companion HOPS pipeline's linking criteria rather than on an end-to-end pipeline run. That is a real limitation, but the paper explicitly defines 'found' in terms of those criteria and cites the pipeline, so the claim is internally consistent as a search-capability estimate; the unmeasured c is more directly quantitative and more central to the abstract's headline numbers. Because the reader already arrived at a CONDITIONAL verdict on this basis, my stress-test does not move the verdict: the relative cadence conclusions are robust, but the absolute predictions require either a WFST-specific trailing-loss measurement or a clearly quantified sensitivity analysis. Verdict remains CONDITIONAL, equivalently UNCHANGED relative to the reader's recommendation.","tokens_in":13389,"tokens_out":4188,"duration_ms":47628,"concrete_test":"Re-run the one-year simulation under otherwise identical settings while sweeping c over a physically plausible range, e.g., 0.2, 0.42, and 0.7, and optionally varying theta between 0.6 and 0.9 arcsec; report the resulting 'in tracklets' and 'found' totals. If the 1800/600 numbers shift by less than 10% across the range, the concern is minor; if they shift by more than 30%, the abstract's absolute claims must be conditioned on a WFST-specific measurement of c. An independent check would be to measure c directly from trailed star images or known main-belt asteroids in WFST commissioning data and compare the recovered trailing-loss curve with Equation (2).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing assumption is in Section 2.3.1: Equation (2) uses c=0.42, 'derived for LSST,' applied directly to WFST without measurement. The paper candidly says 'this choice might introduce some error,' but does not bound that error. This matters more than a typical calibration nuisance because Section 4 shows trailing loss is a dominant effect: removing it or imposing a 2 deg/day speed cut changes the number of findable NEOs by 'several times.' The reason is that the faintest NEOs (H=22-25), which dominate the totals, also have the highest apparent speeds, so they are preferentially removed by the trailing-loss term. An error in c, or in the assumed seeing theta=0.75 arcsec used in Equation (3), therefore propagates directly into the headline numbers of about 1800 known-orbit tracklets and more than 600 blind-search discoveries. The concern is limited to absolute predictions: the relative comparison between Baseline, Twinbase, and Twinaway is driven by cadence and should be robust to a multiplicative detection-efficiency factor. Still, the abstract presents the absolute numbers as a prediction, so the unverified transfer of c should be treated as a condition on that prediction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents end-to-end mock observations of one year of WFST survey operations, combining the Granvik NEO population model, orbital propagation, a detection model with limiting magnitudes and trailing losses, and a scheduler that generates three cadence schemes (Baseline, Twinbase, Twinaway). The central claims are that the Twinbase/Twinaway schemes improve the search capability for known and unknown NEOs by approximately 100% and 50% relative to Baseline, and that at a clear-day ratio of 0.7 the optimized scheme can produce tracklets for about 1800 NEOs with known orbits and find more than 600 NEOs in blind search. The paper also proposes a compact effective field-of-view model and argues that trailing loss is a dominant effect for small, fast NEOs.","tokens_in":1359,"tokens_out":1190,"duration_ms":49758,"significance":"If the results hold, the paper provides a practical, quantitative basis for WFST survey scheduling and identifies an important modeling issue (trailing loss) for small-NEO searches. The relative comparison between scheduling schemes is internally consistent and supported by ten random realizations with quoted 1-sigma scatter, and the compact eFoV model is a reasonable improvement over a purely random acceptance model. The main new quantitative predictions, however, are conditional on an unmeasured transfer of the LSST trailing-loss coefficient to WFST, and the absolute discovery numbers should be read with that caveat until a sensitivity analysis is supplied.","major_comments":[{"comment":"The trailing-loss coefficient c = 0.42 is taken from LSST and applied to WFST with no measurement, and the paper itself states that 'this choice might introduce some error.' This is not a minor calibration issue: Section 4 shows that removing trailing loss or imposing a 2 deg/day speed limit changes the number of findable NEOs by 'several times,' and the faintest NEOs (H = 22-25), which dominate the totals, are precisely the ones with high apparent speeds and the largest trailing losses. An error in c, or in the assumed seeing theta = 0.75 arcsec in Eq. (3), therefore propagates directly into the headline numbers of roughly 1800 tracklets and more than 600 blind-search discoveries. I request a quantitative sensitivity study over a plausible range of c (e.g., 0.2-0.6) or, better, a measured or simulated estimate for WFST's PSF and charge-transfer properties; without this, the absolute predictions in the abstract are not fully supported.","section":"Section 2.3.1, Eqs. (2)-(3)"},{"comment":"The 'known' tracklet counts are computed for the full Granvik model population, i.e., under the assumption that every modeled NEO has a known orbit. The abstract says 'if their orbits are known,' which is a clear condition, but the paper does not state how this hypothetical relates to the actual currently known NEO catalog, which is far smaller than the model population. Since the 100% improvement claim for known NEOs is a central result, the authors should explicitly quantify how many of the ~1800 tracklet objects would correspond to already-known asteroids and how many are hypothetical known-orbit objects; otherwise the 'known' terminology risks being misinterpreted as a prediction about the current MPC catalog.","section":"Section 4, Table 3"}],"minor_comments":[{"comment":"The summary contains a typo: 'rugular survey' should be 'regular survey.'","section":"Section 5"},{"comment":"The notation 'F LI' with a space is awkward; use a consistent subscript or symbol, e.g., FLI.","section":"Section 2.3.1 and Table 1"},{"comment":"The scheme name 'Twinaway ' has a trailing space in the text; please fix the formatting.","section":"Section 3.4"},{"comment":"The table note and column headers are somewhat confusing: clarify that the entries are ratios of NEO counts (random eFoV / compact eFoV), not absolute numbers.","section":"Table 2"},{"comment":"The sentence 'For the Atiras type, they can hardly be found' should be reworded to 'Atira-type objects can hardly be found' for clarity.","section":"Section 4"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for PASP and addresses a timely operational question. The main concern is the unquantified transfer of the LSST trailing-loss coefficient; this is a load-bearing assumption for the absolute predictions but is fixable with sensitivity runs. I would not reject on this basis, and the relative scheme comparison is likely robust."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The bottom line: the relative comparison between scheduling schemes is the real contribution, and it holds up. The absolute numbers are softer than the abstract implies. The paper's own caveat about the trailing-loss coefficient is not just boilerplate; it deserves a sensitivity analysis before anyone quotes 1800 and 600 as predictions.\n\nWhat's new and good: the paired-visit idea (Twinbase) is simple, physically motivated, and likely actionable for WFST's survey planners. The compact eFoV model is a genuine improvement over the random eFoV model; the authors demonstrate that the random model overestimates tracklets by about 40% in the baseline because of overlapping tile fields, which is a real correction. Running 10 realizations and quoting 1-sigma scatter is honest. The paper also correctly notes that Atiras need twilight observations and are essentially missed by the regular survey, which is a useful boundary on the conclusions.\n\nSoft spots, in proportion: the stress-test note lands. The c=0.42 trailing-loss coefficient is taken from LSST and applied to WFST in Eq. 2, and the paper says \"this choice might introduce some error\" but does not bound it. Because the faintest NEOs (H=22-25) dominate the counts and also have the highest apparent speeds, an error in c or in the assumed seeing shifts the absolute predictions directly. The relative gains of Twinbase versus Baseline, however, are driven by cadence and should be robust to a multiplicative detection-efficiency factor; I agree with the stress-test note on that distinction. A sensitivity sweep over c (and perhaps seeing) would be a cheap and necessary addition.\n\nMinor points: the \"found\" criterion depends on the HOPS pipeline described in a companion paper by overlapping authors, which is a mild self-reliance but not circular, since the discovery counts are outputs, not inputs. No code is released; given the number of free parameters, that limits reproducibility but is common in this literature. The atmospheric extinction values and limiting magnitudes are taken from WFST characterization papers, which is appropriate.\n\nWho gets value: anyone planning WFST's survey cadence, and small-body modelers comparing survey strategies. It is a solid simulation study with a concrete operational recommendation, not a fundamental advance.\n\nRecommendation: yes, send it to peer review. Ask for the sensitivity analysis on c and a clearer statement of the HOPS criteria. The relative result is publishable now; the absolute prediction needs its uncertainty quantified.","headline":"Paired-visit scheduling roughly doubles WFST's NEO yield in simulation, but the headline 1800/600 counts are conditional on an unmeasured trailing-loss coefficient.","tokens_in":14193,"tokens_out":2100,"would_cite":true,"duration_ms":23390,"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 paper shows that a modest change to WFST's nightly schedule—giving each sky tile two visits instead of one—roughly doubles the number of known near-Earth objects that get tracklets and increases blind discoveries by about half, with…","keywords":["near-Earth objects","survey scheduling","mock observations","tracklets","trailing loss","effective field of view","Wide Field Survey Telescope"],"falsifier":"Measure WFST's actual trailing-loss coefficient by imaging artificial or real fast-moving sources (or trailed stars) and comparing detected magnitudes to the c=0.42 prediction, then rerun the one-year mock survey; a measured c that differs materially would change the 1,800/600 tracklet and discovery numbers.","tokens_in":13183,"feed_emoji":"☄️","tokens_out":4137,"duration_ms":37875,"temperature":0.7,"pith_summary":"The paper argues that the Wide Field Survey Telescope (WFST), a general-purpose northern sky survey facility not dedicated to asteroid hunting, can still become a productive near-Earth object (NEO) discovery and monitoring machine if its nightly schedule is slightly rearranged. Through mock observations that combine a debiased NEO population model, a detection model, and the actual WFST survey plan, the authors test three scheduling schemes. They find that forcing each sky tile to be visited twice in a night, instead of once, roughly doubles the number of known NEOs that get tracklets and increases blind discoveries by about half. With a realistic 0.7 clear-night fraction, they predict about 1,800 known NEOs with tracklets and more than 600 blind finds in one year. They also show that how the effective field of view is modeled matters, and that trailing loss from fast-moving objects must be included to avoid overestimating NEO counts.","feed_headline":"Paired visits double WFST's known-NEO tracklets","feed_subtitle":"Mock survey predicts 600 blind NEO finds and 1,800 tracklets per year with a smarter schedule.","key_machinery":"The load-bearing mechanism is the scheduling metric called 'recent repeated observations' in the greedy tile-group selector. In the baseline, a tile group that has already been observed that night is avoided, so most tiles get a single visit; the new schemes reset this metric to favor a second visit to the same tile group on the same night, producing paired exposures that are the minimum requirement for forming a tracklet. A supporting piece of machinery is the compact effective field of view model, which replaces the random accept/reject eFoV with a close-packed CCD array of matching filling factor, so that repeated detections of the same object are correlated with its position rather than drawn randomly. The simulation also uses a trailing-loss formula borrowed from LSST to dim fast-moving sources.","core_discovery":"The central claim is that a scheduling scheme that gives each survey tile paired visits on the same night, called Twinbase (and a variant Twinaway that also avoids repeating the previous day's tiles), improves WFST's search capability for known and unknown NEOs by approximately 100% and 50% relative to the baseline schedule. Running one-year mock observations with a 0.7 clear-day ratio, the paper reports 1,824.5 known-orbit NEOs with tracklets and 642.6 blind-discovered NEOs for the best scheme, versus 893.6 and 427.3 in the baseline. A second finding is that the commonly used random effective-field-of-view model overestimates the number of unknown NEOs found, by about 10%, because overlapping tile pointings can create false tracklets; the paper's compact eFoV model avoids this. The paper also demonstrates that ignoring trailing loss would overestimate NEO detections severalfold, especially for small (faint, fast) objects.","pith_inferences":["The paired-visit scheduling principle could be transferred to any survey telescope that currently visits fields once per night, potentially boosting its NEO yield without extra observing time.","The random-vs-compact eFoV discrepancy suggests that other simulations using random field-of-view acceptance may over-report tracklet counts when adjacent tile pointings overlap.","If WFST's first-year real tracklet counts fall well short of the predicted 1,800/600, the most likely culprit is the unmeasured trailing-loss coefficient, making an empirical measurement of c a high-value early activity.","The same mock-observation pipeline could be extended to other solar system populations, such as main-belt asteroids or Jupiter Trojans, to optimize WFST's cadence for those sciences."],"forward_implications":["WFST's regular survey can contribute roughly 1,800 NEO tracklets per year for objects with known orbits, sufficient for MPC submission, without dedicating the telescope exclusively to asteroids.","Blind searches with WFST can find more than 600 NEOs per year, including small objects with absolute magnitude up to 25 (roughly 25 meters across at assumed albedo 0.25).","Potentially hazardous asteroids are found preferentially: the optimized schemes roughly double the number of PHAs with tracklets, from 132 to 279 per year.","The gain comes almost entirely from WFS tiles receiving paired same-night visits; the DHS component contributes fewer NEOs because of its smaller sky coverage.","Survey schedulers should plan for paired visits and account for trailing loss when predicting moving-object yields."],"supporting_citations":[{"why":"Supplies the debiased NEO population model with 802,000 simulated objects, the statistical base population for the mock survey.","marker":"Granvik et al. 2018"},{"why":"Defines the baseline greedy tile-group scheduling strategy that the paper modifies into Twinbase and Twinaway.","marker":"Chen et al. 2023"},{"why":"Provides the trailing-loss formula, the c=0.42 coefficient, and the detection fading function used in the WFST detection model.","marker":"Jones et al. 2018"},{"why":"Describes the HelioLinC-based search pipeline whose tracklet-finding criteria define what counts as a discovered NEO in the simulation.","marker":"Wang et al. 2025"},{"why":"Supplies the Lenghu site seeing (0.75 arcseconds) and the 0.7 clear-day ratio used for the main weather scenario.","marker":"Deng et al. 2021"},{"why":"Gives the WFST limiting magnitudes as a function of moon phase and airmass, which set the detectability threshold for simulated sources.","marker":"Lei et al. 2023"}],"fun_headline_variants":["WFST pairing scheme doubles known-NEO detections","Mock survey: WFST finds 600 blind NEOs yearly with new schedule","WFST twin-visit scheme: 100% more known NEOs, 50% more blind","Paired visits boost WFST NEO finds: 100% known, 50% blind"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The simulation takes the trailing-loss coefficient c=0.42 from LSST and applies it to WFST without measuring it on WFST, so the predicted counts for faint fast NEOs could shift if WFST's point-spread function or charge-transfer smearing behaves differently.","fun_headline_variants_meta":{"raw":{"variants":["WFST pairing scheme doubles known-NEO detections","Mock survey: WFST finds 600 blind NEOs yearly with new schedule","WFST twin-visit scheme: 100% more known NEOs, 50% more blind","Paired visits boost WFST NEO finds: 100% known, 50% blind"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00097,"raw_usage":{"total_tokens":4152,"prompt_tokens":1002,"completion_tokens":3150,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":618,"completion_tokens_details":{"reasoning_tokens":3061}},"tokens_in":618,"tokens_out":3150,"duration_ms":22029,"temperature":1.0,"reasoning_tokens":3061,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T17:10:20.482558+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure WFST's actual trailing-loss coefficient by imaging artificial or real fast-moving sources (or trailed stars) and comparing detected magnitudes to the c=0.42 prediction, then rerun the one-year mock survey; a measured c that differs materially would change the 1,800/600 tracklet and discovery numbers.","supporting_citations":[{"cited_title":"A Heliocentric-orbiting Objects Processing System (HOPS) for the Wide Field Survey Telescope: Architecture, Processing Workflow, and Preliminary Results","cited_arxiv_id":"2501.17472","evidence_quote":"Describes the HelioLinC-based search pipeline whose tracklet-finding criteria define what counts as a discovered NEO in the simulation."}],"review_version":1}