{"id":"55205b89-34df-4f53-8025-33a1a6b6a7b5","arxiv_id":"1908.00457","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A 3D radar-based snowfall climatology for Antarctica shows precipitation concentrated on the periphery and suggests orographic lifting with an inferred vertical wind near 0.02 m/s.","lead":"Using CloudSat radar data, this paper maps the three-dimensional structure of snowfall over Antarctica from 2007 to 2010, revealing strong contrasts between coastal areas and the high plateau. The authors also compare observed snowfall and temperature to a simple model of air lifted over slopes, suggesting a typical vertical wind of about 0.02 m/s.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 0.02 m/s vertical wind claim depends on dropping horizontal advection in Eq. B3, and the paper offers no quantitative test that the simplified budget actually explains the observed scatter.","rationale":"The reader's weakest_assumption identifies the same load-bearing issue: the derivation of Eq. 1 drops horizontal advection in Appendix B, and the inferred w is therefore not uniquely tied to topographic lifting. My stress-test pass confirms this is the structurally weak point of the central mechanism claim. However, I do not see a reason to move the verdict from CONDITIONAL to REJECT: the paper's primary contribution, the 3D CloudSat precipitation climatology, is independent of the simplified budget model, and the mechanism claim is presented as a hypothesis with a plausible order-of-magnitude check. The correct disposition is still CONDITIONAL, requiring the authors to quantify the advection balance and the fit quality; thus the verdict should be UNCHANGED relative to the reader's verdict. The concrete test I propose would settle the concern: if the full budget shows horizontal advection is small, the w=0.02 m/s claim gains real support; if not, the claim should be reframed as an effective diagnostic parameter rather than a physical vertical velocity.","tokens_in":20267,"tokens_out":1669,"duration_ms":15876,"concrete_test":"Re-derive the precipitation-temperature relation from the full steady-state moisture budget rather than the truncated form: use ERA-Interim/ERA5 (or the ECMWF-AUX fields already collocated with CloudSat) to compute u*dq/dx + v*dq/dy, w*dq/dz, and the residual Ev-Pr term, averaged over the same 2007-2010 period and regions. Compare the magnitude of the horizontal advection term to the vertical term; if |u*dq/dx + v*dq/dy| is not small relative to |w*dq_sat/dz| for w=0.01-0.02 m/s, the w inferred from Eq. 1 is not a true vertical wind. Additionally, fit w by least squares to the observed joint precipitation-temperature distributions and report a skill score (e.g., log-likelihood or RMS in log precipitation); if the best-fit w differs from 0.02 m/s by more than a factor of two, the headline value is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central mechanism claim (abstract: precipitation is largely dependent on advection of air masses along topographic slopes with an average vertical wind of 0.02 m/s) rests on Eq. 1, derived in Appendix B from the moisture budget Eq. B3 by assuming steady state, saturation, zero evaporation, and purely vertical motion. The reduction from Eq. B4 to Eq. B5 drops u*dq/dx + v*dq/dy, the horizontal advection terms, without justification. In the Antarctic coastal zone, where the largest precipitation rates occur, cold-air outbreaks and synoptic systems bring strong horizontal moisture gradients; horizontal advection of moisture can be comparable to or larger than the vertical advection term. If the retained w*dq_sat/dz term is not the dominant balance, the fitted vertical wind w is not a physically meaningful measure of topographic lifting. The paper's support for the 0.02 m/s value is also only visual: the text states that the observed distribution 'follows' the analytical curve with w=0.01-0.02 m/s, but no quantitative skill score, regression, likelihood, or residual analysis is provided. An independent and much simpler concern is that Eq. 1 is integrated over the CloudSat column with the observed in-column temperatures; the fitted w thus absorbs all model error and any horizontal advection contribution. The load-bearing assumption is therefore not the dataset construction but the physical interpretation of the fitted w, and this is exactly where the manuscript is least secure.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript constructs the first multi-year, three-dimensional CloudSat snowfall climatology over Antarctica (2007–2010) on a 1° latitude × 2° longitude grid with 240 m vertical bins, using the 2C-SNOW-PROFILE product and co-located ECMWF-AUX temperatures. It documents strong contrasts between peripheral areas and the plateau, regional zonal structures, sea-ice influence over western ice shelves, and histograms/scatterplots of precipitation versus temperature at multiple heights. The paper then derives a saturated-lifting analytical relation (Eq. 1) and, by visual comparison of the curve family to the observed scatter, concludes that precipitation is largely controlled by topographic advection with an average vertical wind of 0.01–0.02 m/s. This mechanism claim is the stated central finding.","tokens_in":20502,"tokens_out":6541,"duration_ms":67958,"significance":"The three-dimensional climatology itself is a valuable community resource: it extends the existing 2D Palerme climatologies, provides vertical-profile diagnostics, and is accompanied by uncertainty estimates from ground-radar comparisons. If the 0.02 m/s result were rigorously established, it would be a compact and useful diagnostic for evaluating model precipitation processes over Antarctica. However, the mechanism claim is currently much weaker than the dataset description: the physical derivation omits horizontal advection without justification, and the agreement is assessed visually rather than quantitatively. The paper therefore needs either a substantial reanalysis-based budget analysis or a careful reframing of Eq. (1) as a conditional-mean diagnostic rather than a retrieved vertical velocity.","major_comments":[{"comment":"The reduction from the full steady-state moisture budget to w ∂q_sat/∂z = −Pr silently drops the horizontal advection terms u∂q/∂x + v∂q/∂y. The text justifies this as “assuming a purely vertical motion”, but the proposed mechanism is topographic lifting of air that is advected horizontally; if u = v = 0 there is no horizontal airflow to be deflected upslope, so the reduced equation is not the limit described. Along the Antarctic coastal margin, synoptic systems and cold-air outbreaks produce strong horizontal moisture gradients, and u·∇q can plausibly be comparable to w∂q_sat/∂z. Without a scale analysis or a quantitative evaluation using collocated reanalysis winds and moisture, Eq. (1) is not demonstrated to be the governing relation for the observed precipitation–temperature distributions. This is the load-bearing step for the abstract’s 0.02 m/s claim, and it needs either a defended derivation (for example, in terrain-following coordinates with an explicit slope term) or an explicit statement that Eq. (1) is a heuristic diagnostic rather than a retrieved physical vertical velocity.","section":"Appendix B, Eqs. (B3)–(B5)"},{"comment":"There is no quantitative measure of agreement between the observed scatter and the analytical curves. The manuscript states that the distribution “follows” or “evolves just over the line with triangle markers” (w = 0.01–0.02 m/s), but no skill score, correlation, likelihood, regression, or residual statistics are reported. Because Eq. (1) is integrated using the CloudSat-observed column temperatures, the fitted w absorbs all retrieval biases, unrepresented microphysical processes, and the neglected advection and pressure terms. The circularity also matters: a multi-curve family is drawn and w is selected by eye so that the curves align with the data, and the same w is then reported as the average vertical wind. I ask the authors to add a formal fit with uncertainty (for example, maximum-likelihood estimation of w per region and height with confidence intervals), or to reposition Eq. (1) as a purely descriptive scaling relation and remove the physical 0.02 m/s wording from the abstract.","section":"Section 4.2, Figs. 10–11"},{"comment":"The plausibility estimate w = u dz/dx uses a representative u = 5 m/s and a zonally averaged slope; this is not an observational determination of w, and dz/dx should be evaluated along the actual air trajectory rather than on the zonal-mean cross-section. The text also introduces a 0.2 m/s exceedance threshold without stating how it was chosen or what uncertainty it carries; Fig. 12 then uses it to classify extreme events. Finally, Eq. (B8) computes ∂q_sat/∂z from the Clausius–Clapeyron relation at constant pressure while the integral in Eq. (B9) is taken over a finite depth with varying pressure and temperature; the pressure dependence of q_sat is not accounted for. Please state the additional approximations explicitly and provide a sensitivity test of Eq. (1) to the pressure term. These issues do not affect the climatological maps, but they directly affect the magnitude of the inferred w.","section":"Section 4.2, Eq. (1) and the 0.2 m/s threshold"}],"minor_comments":[{"comment":"The heading “demonstation” should be “demonstration”.","section":"Appendix B heading"},{"comment":"In the caption of Fig. 4, “Antartic plateau” should be “Antarctic plateau”.","section":"Figure 4 caption"},{"comment":"The 2160 m row lists the temperature interval as “-17.1429 – -36.8571”, the reverse order of the other rows; correct it for readability.","section":"Table A2, 2160 m row"},{"comment":"The abstract and Section 3.1 use “relative seasonal variation” of ±11% and ±143%; please define this statistic explicitly (for example, half of the annual range divided by the annual mean).","section":"Abstract and Section 3.1"},{"comment":"The lower limit of the integral in Eq. (1) is not specified, and the text states that Γ_sat is the moist adiabatic lapse rate while using a constant value of −6.5 K/km; clarify the integration bounds and note the constant-lapse-rate approximation.","section":"Eq. (1) and Section 4.2"},{"comment":"The white contours labelled “σ standard deviation of the distributions” are not defined; state whether σ is computed on precipitation rate, on log precipitation, or on observation counts.","section":"Figures 10–11 captions"}],"recommendation":"major_revision","confidential_remarks":"The descriptive parts of the paper are solid and fit the scope of JGR: Atmospheres; the main risk is that the abstract overstates the physical retrieval. I would advise against accepting the mechanism claim in its current form unless the authors add a quantitative budget-closure analysis or explicitly downgrade Eq. (1) to a heuristic scaling relation. The 0.02 m/s value should not be presented as an observational estimate without an associated uncertainty and without a demonstration that the neglected horizontal-advection terms are small."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — this is a paper about a useful new dataset with an interpretation that outruns the data. The CloudSat 3D precipitation climatology over Antarctica (2007–2010, 1°×2°, 240 m vertical, north of 82°S) is real work: it extends Palerme et al. to all vertical levels, deals explicitly with ground clutter, and produces the kind of regional and vertical-profile diagnostics that climate modelers need. The maps, profiles, and seasonal contrasts (peripheral vs plateau, sea-ice dependence over the western ice shelves) are worth having on their own.\n\nThe soft spot is the mechanism claim. Section 4 and the abstract assert that precipitation is largely dependent on advection of air masses along topographic slopes with an average vertical wind of 0.02 m/s. That number comes from comparing observed precipitation–temperature scatter with a family of curves generated by Eq. 1, derived in Appendix B. The derivation is transparent, but it drops the horizontal advection terms u*dq/dx + v*dq/dy in going from Eq. B3 to B5. Over the Antarctic coast, where precipitation is highest, horizontal moisture gradients are large and the text offers no justification for assuming the vertical term dominates. The fitted w therefore absorbs everything the simplified budget doesn't capture. The paper also gives no quantitative skill score for how well the analytical curves match the binned data; it is a visual match, and the best-fit w is then reported as an average vertical wind. That is a mild circularity: the curves are parameterized by w, w is tuned to the data, and the tuned value is promoted to a physical quantity. The appendix does say “assuming a purely vertical motion,” but the abstract doesn't carry the caveat.\n\nNone of this kills the dataset. But the headline 0.02 m/s should be treated as an empirical curve-fitting parameter with a plausible orographic interpretation, not an independently estimated vertical velocity. A strong revision would (i) test the horizontal-advection neglect with reanalysis winds and moisture gradients, (ii) add a proper statistical comparison of Eq. 1 to the distributions, and (iii) release the gridded dataset so others can use it.\n\nWho it's for: anyone evaluating Antarctic snowfall in climate models, and anyone working on the vertical structure of polar precipitation. It deserves a serious referee — the dataset alone justifies that. I'd send it out, but I'd tell the authors to either back the mechanism claim or soften it.","headline":"A genuinely useful 3D Antarctic precipitation climatology whose headline 0.02 m/s vertical wind claim is a visually fitted parameter, not an independently estimated physical quantity.","tokens_in":21121,"tokens_out":3272,"would_cite":true,"duration_ms":34657,"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":"Antarctic precipitation is largely orographic: the new 3D CloudSat climatology shows air lifted up the ice-sheet slopes at about 0.02 m/s.","keywords":["CloudSat","Antarctic precipitation","snowfall vertical structure","orographic precipitation","precipitation-temperature relationship","ice sheet mass balance","satellite radar","polar meteorology"],"falsifier":"Compute a full column moisture budget over the Antarctic margin from high-resolution reanalysis or a regional model, retaining the horizontal advection terms $u\\,\\partial q/\\partial x+v\\,\\partial q/\\partial y$, and compare the diagnosed condensation profile to CloudSat's measured precipitation. If the full-budget condensation does not reproduce the observed precipitation-temperature scatter, or if direct wind measurements show vertical velocities far from 0.01 to 0.02 m/s over the slopes, then Equation 1 is not the controlling mechanism.","tokens_in":20032,"feed_emoji":"❄️","tokens_out":8670,"duration_ms":83932,"temperature":0.7,"pith_summary":"CloudSat's cloud-profiling radar lets scientists see, for the first time, where Antarctic snowfall forms and how it changes with height, not just how much reaches the surface. This paper builds the first continent-scale three-dimensional precipitation climatology from those measurements and uses it to argue that most Antarctic precipitation is orographic: moist air from the Southern Ocean is pushed up the ice-sheet slopes, cools, and condenses along the moist adiabat. The observations show a sharp split between the periphery, which receives about 275 mm/yr at 1200 m above ground with a ±11% seasonal swing, and the high plateau, which receives about 34 mm/yr with a ±143% swing. At every altitude, the spread of precipitation with temperature follows a simple analytical lifting curve with a vertical wind near 0.02 m/s. If this explanation is right, a climate model's three-dimensional snowfall field can be checked against one compact diagnostic rather than against scattered surface measurements.","feed_headline":"Antarctic snowfall follows one simple slope-lifting rule","feed_subtitle":"New 3D CloudSat view traces most Antarctic precipitation to air rising up the ice sheet at ~0.02 m/s.","key_machinery":"The load-bearing object is Equation 1, the saturation-forced-lifting relation\n$$\n\\mathrm{Pr}=-\\frac{w}{\\rho_{\\mathrm{water}}}\\int_z \\rho_{\\mathrm{atm}}\\frac{L\\,q_{\\mathrm{sat}}(T,p)}{R_{\\mathrm{vap}}$T^{2}$}\\Gamma_{\\mathrm{sat}}\\,dz,\n$$\nderived in Appendix B from a moisture budget by assuming steady, saturated air, zero evaporation, and purely vertical motion, so that precipitation equals the column-integrated condensation produced by lifting at vertical speed $w$. The paper overlays curves for $w=0.0001$ to $1\\ \\mathrm{m\\,s^{-1}}$ on the observed precipitation-temperature scatter plots; the dense part of the observations tracks the $0.01\\ \\mathrm{m\\,s^{-1}}$ curve, with the 0.02 m/s average cited in the abstract, and the deviations are attributed to steep local topography.","core_discovery":"The paper's central claim is that the vertical structure of Antarctic precipitation is governed by forced orographic lifting of near-saturated marine air over the ice-sheet margins, with an average vertical velocity of about $0.02\\ \\mathrm{m\\,s^{-1}}$. To establish this, the authors regrid the CloudSat 2C-SNOW-PROFILE snowfall retrievals onto a $1^{\\circ}\\times2^{\\circ}$ horizontal grid with 240 m vertical bins from roughly 1200 m to 10 km above ground, and compare the precipitation-temperature histograms at each level with an analytical relation for condensation by steady saturated ascent. The observed distributions sit between the curves for $w=0.01$ and $w=0.1\\ \\mathrm{m\\,s^{-1}}$, close to the $0.01\\ \\mathrm{m\\,s^{-1}}$ curve over the periphery and slightly above it over the plateau, and the paper interprets the spread at fixed temperature as the variety of slopes and horizontal wind strengths that set the local vertical wind. The same dataset yields the regional contrasts: high, weakly seasonal snowfall over the coasts and peninsula, low and rare snowfall on the plateau, and a strong dependence of ice-shelf precipitation on sea-ice coverage.","pith_inferences":["Not pursued in the paper: a direct computation of the same precipitation-temperature scatter from a full-physics model that retains horizontal moisture advection would test whether the $w\\approx0.02\\ \\mathrm{m\\,s^{-1}}$ relation is a causal mechanism or a curve fit.","A testable extension would be to predict interannual snowfall from the product of near-surface wind speed and slope, $u\\,dz/dx$, times saturation humidity; if the relation is causal, years with stronger onshore flow should show proportionally more peripheral precipitation.","Because ground clutter hides the lowest roughly 1200 m, the inferred vertical wind applies aloft; an observational follow-up using ground-based micro rain radars at coastal East Antarctic stations could show where the lifting curve breaks as sublimation and katabatic winds take over."],"forward_implications":["Climate-model evaluation can move from surface snowfall totals to three-dimensional precipitation-temperature distributions: a model that places its snowfall off the observed $w\\approx0.01$ to $0.02\\ \\mathrm{m\\,s^{-1}}$ envelope has a transport, thermodynamics, or microphysics bias.","The periphery/plateau contrast implies that model skill must be assessed separately for the two regimes, since on the plateau, where the seasonal swing is ±143%, a few large events dominate and mean-state comparisons over short periods can mislead.","Over the ice shelves, summer versus winter precipitation differences and their link to sea-ice coverage give a direct test of coupled sea-ice-atmosphere behavior in models.","The analytical relation turns the average Antarctic slope and horizontal wind into a predicted snowfall rate, so the paper's diagnostic can be applied without retuning."],"supporting_citations":[{"why":"It supplies the first multi-year CloudSat Antarctic snowfall climatology, the 2250 m topographic division, and the baseline 171 mm/yr estimate.","marker":"Palerme et al. (2014)"},{"why":"It provides the ground-clutter-aware revision (160 mm/yr) whose vertical extension forms the dataset used here.","marker":"Palerme et al. (2019)"},{"why":"It defines the 2C-SNOW-PROFILE retrieval, its snowfall-rate uncertainty, and the Snow Retrieval Status used to filter the data.","marker":"Wood (2011)"},{"why":"It is the ground-radar comparison that justifies the 1 degree by 2 degree grid and brackets CloudSat omissions and commissions.","marker":"Souverijns et al. (2018)"},{"why":"It gives the profile-wide uncertainty bounds used for the climatology's error bars.","marker":"Lemonnier et al. (2019)"},{"why":"It shows the low-level sublimation layer below CloudSat's clutter that explains why the observed profile has not yet turned over near the surface.","marker":"Grazioli, Madeleine, et al. (2017)"},{"why":"It documents that about a third of precipitation events at two East Antarctic sites are virga, framing why the vertical structure matters.","marker":"Durán-Alarcón et al. (2019)"}],"fun_headline_variants":[],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that horizontal moisture advection can be neglected, so the column condensation is simply $w\\,\\partial q_{\\mathrm{sat}}/\\partial z=-\\mathrm{Pr}$; if horizontal transport of moisture is as important as vertical lifting over Antarctic slopes, the inferred vertical wind of 0.01 to 0.02 m/s is not a real physical wind.","fun_headline_variants_meta":{"error":"Client error '402 Payment Required' for url 'https://api.deepseek.com/chat/completions'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/402"},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:54:58.820524+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute a full column moisture budget over the Antarctic margin from high-resolution reanalysis or a regional model, retaining the horizontal advection terms $u\\,\\partial q/\\partial x+v\\,\\partial q/\\partial y$, and compare the diagnosed condensation profile to CloudSat's measured precipitation. If the full-budget condensation does not reproduce the observed precipitation-temperature scatter, or if direct wind measurements show vertical velocities far from 0.01 to 0.02 m/s over the slopes, then Equation 1 is not the controlling mechanism.","supporting_citations":[{"cited_title":", Kay, J","cited_arxiv_id":null,"evidence_quote":"It supplies the first multi-year CloudSat Antarctic snowfall climatology, the 2250 m topographic division, and the baseline 171 mm/yr estimate."},{"cited_title":", Claud, C","cited_arxiv_id":null,"evidence_quote":"It provides the ground-clutter-aware revision (160 mm/yr) whose vertical extension forms the dataset used here."},{"cited_title":"APACrefauthors \\ 2011","cited_arxiv_id":null,"evidence_quote":"It defines the 2C-SNOW-PROFILE retrieval, its snowfall-rate uncertainty, and the Snow Retrieval Status used to filter the data."},{"cited_title":", Gossart, A","cited_arxiv_id":null,"evidence_quote":"It is the ground-radar comparison that justifies the 1 degree by 2 degree grid and brackets CloudSat omissions and commissions."},{"cited_title":", Madeleine, J","cited_arxiv_id":null,"evidence_quote":"It gives the profile-wide uncertainty bounds used for the climatology's error bars."}],"review_version":1}