{"id":"7f119a97-fec4-4f38-8c11-59ad941fb544","arxiv_id":"2412.01081","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"During the 2009 major SSW, mesospheric CO2 density and CO2 15 um infrared cooling were anti-correlated, with temperature and atomic oxygen variations, not CO2 abundance, driving the cooling changes.","lead":"This study reports that during the 2009 sudden stratospheric warming, carbon dioxide density in the polar mesosphere increased while its infrared cooling decreased, an apparent contradiction. The authors argue that temperature and atomic oxygen changes, not CO2 abundance, controlled the cooling, which matters for understanding upper atmospheric energy balance.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The anti-correlation in Fig. 4a may be an artifact of converting ACE-FTS CO2 VMR to density using SABER temperature; the conversion fields are unspecified and the printed CO2 density values at 0.003 hPa are physically implausible.","rationale":"The paper's central claim is the event-specific anti-correlation between CO2 density and CO2 15 um cooling, not the climatological CO2-greenhouse-cooling relationship. For that claim to be true, the CO2 density anomalies must be a faithful measure of actual CO2 abundance changes at 0.003 hPa, not an ideal-gas transform of the same temperature field that drives the SABER cooling product. The manuscript does not describe the conversion. The reader's weakest_assumption identifies exactly this gap; my stress-test strengthens it with two concrete observations: (1) at fixed pressure, using SABER T in the conversion forces n_CO2 to be anti-correlated with the cooling even for constant VMR, because the cooling correlates strongly with T; and (2) the Fig. 4a axes imply density values roughly 10^3 times the physically expected VMR * air density at 0.003 hPa, which suggests a unit/conversion error. I do not base the objection on the simplified two-level population model in Section 3.4: that model is a secondary attribution tool, and the density conversion must be settled first. The WACCM-X comparisons in Figs. 3b/d provide partial independent support for the qualitative pattern, but they are model output, not an observational confirmation of the ACE-FTS density conversion. A conditional verdict is appropriate: the authors should be required to state the T/P fields, recompute Fig. 4a using VMR and independent temperatures, and verify absolute units. The reader's verdict already requires this, so no change is needed.","tokens_in":18912,"tokens_out":9037,"duration_ms":82864,"concrete_test":"Recompute Figure 4a using ACE-FTS CO2 VMR directly (mixing-ratio anomalies) and using density anomalies computed with three temperature fields at 0.003 hPa: SABER T, SD-WACCM-X/MERRA-2 T, and a fixed reference T. If the anti-correlation weakens or vanishes in the VMR-based or independent-T versions, the reported R = -0.865 is an artifact of the conversion. Separately, verify the absolute magnitude: 370 ppm at 0.003 hPa and 200 K should give about 4.0e10 cm^-3; if the actual Fig. 4a axis is 10^13, the conversion factor needs correction before any causal claim based on density can be accepted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step for the headline claim (R = -0.865 in Fig. 4a) is the conversion of ACE-FTS CO2 VMR to number density in Section 2. The manuscript cites Finlayson-Pitts and Pitts Jr (1999) but does not specify which temperature and pressure fields enter the conversion. At the fixed 0.003 hPa surface used for Fig. 4, any conversion of the form n_CO2 = VMR * P/(k_B T) makes the density anomalies proportional to 1/T when SABER temperature is used. SABER cooling is strongly correlated with SABER temperature (R = 0.927, Fig. 4b), and both are products of the same 15 um radiance, so the cooling/T correlation is not an independent test. A 1/T density proxy would therefore be anti-correlated with the cooling even if the CO2 VMR were constant, which would produce the counter-intuitive relationship without any real change in CO2 abundance. The concern is sharpened by Figure 4a: at ~85 km (0.003 hPa, ~200 K), air number density is ~1.1e14 cm^-3, so 370 ppm CO2 should correspond to ~4e10 cm^-3, not the printed values around 2.5-4.5e13 cm^-3. The conversion appears to be off by roughly 10^3. Section 3.4 then builds the O/CO2 and O2/CO2 ratios and the population-ratio attribution on this CO2 density, so the unspecified conversion propagates into the conclusion that temperature and atomic oxygen, not CO2, control the cooling. The verdict should remain conditional until the authors state the T and P fields and re-derive Fig. 4a with an independent temperature.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript studies the 2009 major sudden stratospheric warming (SSW) and its effect on mesospheric CO2 15 µm radiative cooling. Using SABER temperature, CO2 cooling, and O density; ACE-FTS CO2 volume mixing ratio (VMR) converted to number density; and SD-WACCM-X winds, O2, and CO2 fields, the authors report a strong anti-correlation (R = -0.865) between CO2 density and CO2 IR cooling at 0.003 hPa (~85 km) in the 60-70°N zonal mean during January-February 2009. They interpret this in terms of circulation-driven upwelling/downwelling: during the SSW main phase, reduced temperature and atomic oxygen depress CO2 IR cooling despite increased CO2 density, while the recovery phase shows enhanced cooling despite reduced CO2 density. The paper concludes that temperature changes dominate the CO2 IR cooling variability, followed by atomic oxygen, with the CO2 density change itself insignificant.","tokens_in":19323,"tokens_out":6375,"duration_ms":59190,"significance":"The science question is timely and relevant: short-term, event-driven variability in mesospheric CO2 15 µm cooling bears directly on the MLT energy budget and on interpretations of long-term upper-atmosphere contraction. The study combines multiple observational data sets (ACE-FTS, SABER) with a specified-dynamics model run, uses no fitted parameters, and adopts rate coefficients from prior literature. If the central anti-correlation and attribution survive scrutiny, the paper would be a useful observational case study of dynamical control of mesospheric CO2 cooling during an SSW. However, the headline result as presented is not yet trustworthy: the VMR-to-density conversion is underspecified, the printed CO2 densities are unphysical, and the temperature-cooling correlation is partly non-independent. These issues affect the quantitative claims in Figures 4, 6, and 7 and the attribution in Section 3.4.","major_comments":[{"comment":"The VMR-to-density conversion is not specified beyond a citation to Finlayson-Pitts and Pitts Jr (1999). The authors must state exactly which temperature and pressure fields are used. If SABER temperature is used at the fixed 0.003 hPa surface, then n_CO2 = VMR * P / (k_B T), so the density anomalies become a 1/T transform of the temperature anomalies. Since SABER temperature and SABER CO2 cooling are both retrieved from the same 15 µm radiance, an anti-correlation between this 1/T proxy and the cooling would arise even for a perfectly constant CO2 VMR. Please redo the conversion with an independent temperature field (e.g., ACE-FTS or WACCM-X) and state the fields explicitly.","section":"Section 2 and Figure 4a"},{"comment":"The printed CO2 density values, approximately 2.5-4.5 × 10^13 cm^-3 at 0.003 hPa (~85 km), are physically implausible. At ~200 K and 0.003 hPa the air number density is about 1.1 × 10^14 cm^-3, so a VMR of 370 ppm corresponds to about 4 × 10^10 cm^-3, not 10^13 cm^-3. This implies a conversion or unit error of roughly three orders of magnitude. Because Section 3.2 and Section 3.4 use these densities to compute percentage anomalies and the O/CO2 and O2/CO2 ratios, the quantitative results and the attribution are directly affected. The authors must correct the conversion and recompute Figures 3, 4, 6, and 7.","section":"Section 3.2 and Figure 4a"},{"comment":"The attribution model is a simplified two-level collisional excitation formula based on reactions R1 and R2 only. It omits radiative absorption, spontaneous emission, and the non-LTE exchange terms that determine the actual 15 µm cooling rate. The decomposition of the cooling variability into temperature, O, and O2 contributions in Figure 8 is therefore not a demonstrated decomposition of the cooling rate itself. Please justify this simplified population-ratio model against a non-LTE cooling calculation, or explicitly reframe the attribution as qualitative rather than quantitative.","section":"Section 3.4 and Figure 8"},{"comment":"The correlation coefficients (-0.865, 0.927, 0.905) are quoted without confidence intervals, significance tests, or a statement of effective sample size. The daily-mean time series are strongly autocorrelated, so the number of independent samples is much smaller than the number of days. In addition, the R = 0.927 between SABER temperature and SABER CO2 cooling is not an independent check, because both quantities are derived from the same 15 µm broadband radiance. Please provide uncertainty estimates and clearly identify which correlations are physically independent.","section":"Section 3.2 and Figures 3-4"}],"minor_comments":[{"comment":"There are typographical and grammatical errors: 'mesosphere and lower thermospher' should be 'mesosphere and lower thermosphere', and 'such a dramatic events' should be 'such dramatic events'.","section":"Abstract"},{"comment":"The axis labels use '#1013' and '#10-8' instead of proper superscript notation. They should be rendered as '×10^13 cm^-3' and '×10^-8 W m^-3' for clarity.","section":"Figure 4"},{"comment":"The sentence defining anomalies, 'Δ CO2 = CO2 − CO2', uses the same symbol for the daily value and the pre-SSW mean. Please use an overbar or a distinct symbol for the mean value to avoid ambiguity.","section":"Section 3.2"},{"comment":"The text states that CO2 density increased during the SSW main phase and decreased during recovery, while Figure 3 shows cooling anomalies of opposite sign. A short table of phase-mean anomalies for CO2 density, temperature, O density, and CO2 IR cooling would make the anti-correlation easier to follow and would help the reader judge the magnitudes.","section":"Section 3.2"}],"recommendation":"major_revision","confidential_remarks":"The central problem is the VMR-to-density conversion: it is underspecified, and the printed densities are unphysical by about three orders of magnitude. This is a load-bearing error, but it is potentially correctable within the manuscript's scope by specifying the temperature and pressure fields, using an independent temperature, and recomputing the affected figures. I am not recommending rejection because the scientific question is interesting, the multi-dataset approach is valuable, and the error appears identifiable and fixable. The authors should also be asked to clarify that SABER temperature and cooling are not independent, and to provide a sensitivity test in which CO2 VMR is held constant while temperature varies."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper claims a counter-intuitive anti-correlation between CO2 density and CO2 15 µm cooling during the 2009 SSW, but I don't think that claim survives contact with the methodology. The VMR-to-density conversion is never specified, and if it uses SABER temperature, the anti-correlation is built in rather than discovered. The printed densities in Fig. 4a are also off by roughly three orders of magnitude, which makes me suspect a unit error or a substitution of the wrong pressure field.\n\nThat said, the paper does useful things. It puts together ACE-FTS VMR, SABER cooling/temperature/O, and WACCM-X in a single event study, and the circulation narrative—upwelling during the warming peak, downwelling in the recovery—is consistent with prior work on tracer transport. The population-ratio analysis is a legitimate way to separate temperature, O, and O2 contributions, even if it is simplified. The topic matters: SSW effects on MLT energy balance are understudied.\n\nThe soft spots are real. The conversion scheme is the load-bearing wall. Without knowing which T and P fields go into n = VMR·P/kT, the R = -0.865 correlation is not interpretable. If the T field is SABER's, then the density anomalies are essentially 1/T, and since cooling tracks T strongly (R=0.927 in Fig. 4b), the anti-correlation is likely an artifact. There are no error bars, no significance tests, and no detrending; the correlations could be inflated by the common seasonal or event-scale trend. The paper's own conclusion that CO2 density changes are \"insignificant\" is also not independent, because the same converted density is used in the population-ratio model. Finally, it's a single event; even if the analysis were clean, it would be a case study, not a general result.\n\nThe reader's take is close to mine, but I'd be even firmer on the central claim. The stress-test concern about circularity and the order-of-magnitude density error is not a nitpick; it changes what the paper can claim.\n\nWho should read it? People studying SSW impacts on the MLT energy budget, and anyone who wants to see how ACE-FTS and SABER can be combined. It deserves a serious referee, because the question is real and the data are public—but the referee should ask for a rederived Fig. 4a using a temperature-independent density (e.g., from WACCM-X or a specified T field), explicit error bars, and a statement of the conversion. I would not cite it in its current form.\n\nRecommendation: send to peer review, expect heavy revision.","headline":"The paper's headline anti-correlation is likely a conversion artifact, not a real geophysical finding, but the case study is worth referee attention.","tokens_in":19791,"tokens_out":4863,"would_cite":false,"duration_ms":40661,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"During the 2009 sudden stratospheric warming, mesospheric carbon dioxide density and its infrared cooling moved in opposite directions, with temperature and atomic oxygen, not CO2 abundance, controlling the cooling.","keywords":["sudden stratospheric warming","mesosphere","CO2 infrared cooling","15 µm emission","atomic oxygen","ACE-FTS","SABER","middle atmosphere circulation"],"falsifier":"Recompute the CO2 number-density anomaly at 0.003 hPa using ACE-FTS mixing ratios with an independent temperature–pressure profile (for example, from a reanalysis or a model run that does not assimilate SABER) and repeat the correlation with SABER cooling; if the correlation coefficient falls well below 0.865 in magnitude, the anti-correlation is an artifact of the shared temperature field. A second check is to rerun the population-ratio calculation with CO2 density held at pre-SSW values and see whether the temperature- and O-driven cooling pattern still reproduces the observed anomaly.","tokens_in":18768,"feed_emoji":"🌡️","tokens_out":9406,"duration_ms":72610,"temperature":0.7,"pith_summary":"This paper tries to establish that the 2009 major sudden stratospheric warming (SSW) reversed the usual relationship between carbon dioxide abundance and carbon dioxide infrared cooling in the polar mesosphere. Using satellite measurements near 85 km (0.003 hPa), the authors find that CO2 density increased during the warming peak and dropped during recovery, while CO2 15 µm radiative cooling moved the other way, giving a strong anti-correlation ($R = -0.865$). They argue that the cooling changes were driven mainly by temperature swings of about 20 K in either direction, followed by large atomic oxygen density changes, and that the CO2 density change itself was insignificant. If correct, this means short-term dynamical events can decouple CO2 cooling from CO2 concentration, which matters for interpreting satellite cooling records and upper-atmosphere energy budgets.","feed_headline":"CO2 density rose while its cooling fell during 2009 SSW","feed_subtitle":"Temperature, not CO2 abundance, drove the mesospheric 15 µm cooling swings at 85 km.","key_machinery":"The $\\mathrm{CO_2}(0110)/\\mathrm{CO_2}(0000)$ population ratio—the fraction of CO2 molecules in the first excited bending vibrational state that emits at 15 µm—carries the argument. The paper evaluates this ratio with a two-level collisional excitation model using rate coefficients $k_1$ for O2/N2 and $k_2$ for O taken from the non-LTE literature. During the SSW, mesospheric temperature and O density fall at peak warming and rise at recovery, and those changes lower and raise the population ratio even though CO2 density changes in the opposite direction; the population ratio tracks the observed cooling pattern. The same calculation is then used to isolate the separate contributions of temperature, O, and O2 changes, showing temperature first and O second.","core_discovery":"During the 2008–2009 Arctic winter, a major SSW with an elevated stratopause reversed the residual mean meridional circulation over 60–70°N: upwelling during the main phase lifted CO2-rich air upward, increasing mesospheric CO2 density by roughly 10–12%, while downwelling during recovery brought CO2-poor air down, reducing density by about 15%. Despite the larger CO2 density, SABER observed roughly 20% weaker 15 µm cooling at 0.003 hPa during the warming peak, and despite the lower CO2 density, roughly 120% stronger cooling during recovery. The correlation analysis gives $R = -0.865$ between CO2 density and cooling, $R = 0.927$ with temperature, and $R = 0.905$ with atomic oxygen density. The paper concludes that the temperature decrease during upwelling and the temperature increase during downwelling, amplified by O density changes through collisional excitation of the CO2 ν2 mode, control the cooling, making CO2 abundance changes dynamically secondary.","pith_inferences":["One testable extension is to check other major SSWs (for example, 2013 and 2018) with the same SABER and ACE-FTS datasets to see whether the anti-correlation is a generic SSW signature or unique to the 2009 elevated-stratopause event.","Because SABER temperature is retrieved from the same 15 µm emission that defines the cooling, the high temperature–cooling correlation is partly self-referential; an independent temperature dataset would strengthen the causal attribution.","The large O density enhancement during recovery implies SSW downwelling delivers atomic oxygen from the thermosphere into the mesosphere, which could affect ozone chemistry and nightglow emissions in addition to CO2 cooling."],"forward_implications":["During a major SSW, mesospheric CO2 density is a poor proxy for CO2 infrared cooling on sub-seasonal timescales.","The roughly 20 K mesospheric cooling at the warming peak is the dominant driver of the reduced cooling, and the comparable warming during recovery drives the enhanced cooling.","Atomic oxygen variations ranging from about −97% to +500% act as a secondary control by changing the collisional excitation of CO2 into the emitting state.","The same qualitative anti-correlation appears in the specified-dynamics model output, suggesting the upwelling/downwelling mechanism is robust, not a single-satellite artifact."],"supporting_citations":[{"why":"supplies the non-LTE CO2 radiative transfer framework and the collisional excitation rate coefficients k1 and k2 used in the population-ratio analysis.","marker":"(López-Puertas & Taylor, 2001)"},{"why":"sets the long-term baseline that increasing CO2 causes more cooling and contraction, against which the short-term anti-correlation is contrasted.","marker":"(Mlynczak et al., 2022)"},{"why":"provides the SABER CO2 15 µm infrared cooling observations used as the primary cooling dataset.","marker":"(Mlynczak et al., 2010)"},{"why":"gives the mechanism by which SSW westward wave forcing and gravity-wave filtering reverse the meridional circulation, driving the upwelling and downwelling.","marker":"(Liu & Roble, 2002)"},{"why":"models SSW-induced CO2 changes in the polar MLT and supports focusing on 60–70°N for transport-related density anomalies.","marker":"(Orsolini et al., 2022)"},{"why":"describes the composite MLT response and elevated-stratopause recovery used to interpret the recovery-phase circulation.","marker":"(Limpasuvan et al., 2016)"},{"why":"documents the record-breaking 2009 Arctic SSW and its dynamics and transport, providing the event context.","marker":"(Manney et al., 2009)"},{"why":"provides the ideal-gas scheme used to convert ACE-FTS CO2 mixing ratios to number densities, the step on which the anti-correlation result depends.","marker":"(Finlayson-Pitts & Pitts Jr, 1999)"}],"fun_headline_variants":["CO2 density up but cooling down during 2009 SSW","More CO2, less cooling: 2009 SSW's counterintuitive twist","Temperature and atomic oxygen, not CO2, drive 2009 SSW cooling","Mesospheric CO2 cooling decouples from density in major SSW","2009 SSW: CO2 abundance secondary to temperature for cooling"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper converts ACE-FTS CO2 mixing ratios to density but does not state which temperature and pressure data it uses for that conversion; if those come from the same SABER 15-µm retrieval that defines the cooling, the density and cooling anomalies share a common source and the reported anti-correlation could be inflated.","fun_headline_variants_meta":{"raw":{"variants":["CO2 density up but cooling down during 2009 SSW","More CO2, less cooling: 2009 SSW's counterintuitive twist","Temperature and atomic oxygen, not CO2, drive 2009 SSW cooling","Mesospheric CO2 cooling decouples from density in major SSW","2009 SSW: CO2 abundance secondary to temperature for cooling"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000667,"raw_usage":{"total_tokens":3037,"prompt_tokens":932,"completion_tokens":2105,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":548,"completion_tokens_details":{"reasoning_tokens":2011}},"tokens_in":548,"tokens_out":2105,"duration_ms":15839,"temperature":1.0,"reasoning_tokens":2011,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T04:42:07.738460+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the CO2 number-density anomaly at 0.003 hPa using ACE-FTS mixing ratios with an independent temperature–pressure profile (for example, from a reanalysis or a model run that does not assimilate SABER) and repeat the correlation with SABER cooling; if the correlation coefficient falls well below 0.865 in magnitude, the anti-correlation is an artifact of the shared temperature field. A second check is to rerun the population-ratio calculation with CO2 density held at pre-SSW values and see whether the temperature- and O-driven cooling pattern still reproduces the observed anomaly.","supporting_citations":[{"cited_title":"\\ Pitts Jr, J N","cited_arxiv_id":null,"evidence_quote":"provides the ideal-gas scheme used to convert ACE-FTS CO2 mixing ratios to number densities, the step on which the anti-correlation result depends."}],"review_version":1}