{"id":"89c80c33-10eb-430e-be02-433fdc52cd01","arxiv_id":"2411.14183","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A local energy balance model shows that exponential forcing, linear feedbacks, a fixed forcing pattern, and linear dynamics make the warming pattern invariant, reconciling pattern scaling with the pattern effect.","lead":"Climate model warming patterns, once divided by the global average, stay nearly constant across scenarios and time, and this paper explains why with a simple energy balance theory. The same theory shows why the pattern is not constant in abrupt CO2 experiments, reconciling pattern scaling with the pattern effect.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The exponential-forcing proof is sound, but the paper never independently tests the linear/time-invariant dynamics assumption that bridges the idealized model to CMIP6 SSPs.","rationale":"The reader's weakest-assumption identification is the same one I find most load-bearing: the representation of ocean heat uptake and heat-transport changes by local constant C(r) and negligible or linear-diffusive dynamics. I agree with the conditional verdict. The theoretical core is sound; the issue is empirical support. The paper's own Section 4d flags the weakness, and the available evidence (linear local-global regressions for the multi-model mean) is consistent with the theory but not independent of it. A quantitative energy-budget residual test would directly decide whether the dynamics condition holds in the models that exhibit pattern invariance. If the residual is large, the conclusion that 'these conditions are roughly met' would need to be weakened to 'the SSP warming patterns happen to be invariant for reasons not yet established.' If the residual is small, the proposed mechanism is confirmed as the actual cause. Either way, the paper's contribution—the clean exponential-eigenfunction explanation and the contrast with abrupt forcing—remains valuable, so the necessary change is to strengthen or qualify the empirical claim, not to reject the work.","tokens_in":24160,"tokens_out":9053,"duration_ms":93014,"concrete_test":"Diagnose the local energy-budget residual in a CMIP6 SSP with available diagnosed forcing (e.g., ssp245): Q(r,t)=C_eff(r)∂ΔT/∂t − [R(r,t)+λ(r)ΔT(r,t)+D∇²ΔT(r,t)], where C_eff(r) is fixed from the first 20 years as column heat-content tendency divided by ∂ΔT/∂t, λ(r) is the local Gregory slope from abrupt-4xCO2, and D is a Sellers diffusivity fit to the same model. Compute Q from CMIP6 output over 2015–2100. If the root-mean-square of Q is comparable to R(r,t) or grows relative to R over time, the linear time-invariant dynamics assumption is not met and the SSP pattern invariance needs a different explanation. If Q stays below about 10–20% of R with no secular trend, the dynamics assumption is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The mathematical derivation in Section 3b is correct: a linear, time-invariant system driven by exponential forcing with a fixed spatial pattern has a time-independent response pattern. The load-bearing step is the empirical claim that these conditions are approximately met in CMIP6 SSP projections. Section 4d explicitly concedes that the treatment of meridional heat transport and ocean heat uptake is 'highly simplified', that compensating atmospheric and oceanic heat transport anomalies would require a more sophisticated model, and that advective/eddy transport is not addressed. Yet the only quantitative evidence offered for the dynamics condition is the linearity of the local-versus-global regressions in Figure 7 (Section 4a). That linearity is precisely the pattern-scaling property the paper aims to explain, so it does not independently validate the proposed mechanism. The Gregory regressions in Section 4b test whether TOA radiative response ΔN is linear in local ΔT, but they do not test the key assumption that the local heating term can be written as C(r)∂ΔT/∂t with constant C(r), nor that MHT/OHU anomalies can be represented as a linear time-independent operator rather than nonlinear, state-dependent advection. If real ocean heat uptake is nonlocal, lagged, and time-varying, the eigenfunction argument still applies to any linear time-invariant operator, but the paper does not establish that the relevant dynamics are linear and time-invariant; it asserts qualitative consistency.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript develops a local energy-balance theory for the time and scenario invariance of the normalized warming pattern ΔT(r,t)/ΔT(t). It shows analytically that under exponential forcing, linear stabilizing feedbacks, a separable constant forcing pattern, and linear time-invariant (or diffusive) dynamics, the pattern is exactly time- and magnitude-invariant (Eq. 9); under abrupt or overshoot forcing, regional timescales τ(r)=−C(r)/λ(r) make the pattern evolve. The authors then compare the idealized cases with CMIP6 multi-model-mean data: ssp585 (exponential), ssp119 (overshoot), and abrupt-4xCO2, and use RFMIP forcing diagnostics and Gregory-style regressions to argue that the conditions are approximately met except in the Arctic and in regions with strong aerosol forcing. They conclude that pattern scaling is robust in SSPs and that the pattern effect arises from non-exponential forcing.","tokens_in":24517,"tokens_out":5063,"duration_ms":48549,"significance":"If the theory holds, it provides a simple mechanistic explanation for a widely used empirical procedure, reconciles pattern scaling with the pattern effect, and explains why nonlinear emulators give only marginal improvements. The analytical derivations are transparent, the key result Eq. (9) is exact under stated assumptions, model parameters come from prior published work rather than being tuned to the CMIP6 patterns, and Appendix A extends the result to nonlocal radiative feedbacks. The main weakness is that the empirical validation of the dynamics assumption is incomplete, so the significance is conditional on further testing.","major_comments":[{"comment":"The load-bearing assumption that ocean heat uptake and heat transport anomalies are representable by a constant local heat capacity C(r) plus a linear time-independent operator is not independently tested. The only quantitative evidence offered for the dynamics condition is the linearity of the local-to-global regressions in Figure 7, but that linearity is the pattern-scaling property the paper aims to explain, so it cannot validate the proposed mechanism. The Gregory regressions in Figure 8 test whether ΔN is linear in local ΔT, not whether H(r,t)=C(r)∂ΔT/∂t with constant C or whether MHT/OHU anomalies are a linear time-independent function of ΔT. Section 4d concedes that the treatment of MHT is 'highly simplified', that compensating atmospheric and oceanic heat transport anomalies would require a more sophisticated model, and that advective/eddy transport is not addressed. Without an independent test of this assumption, the central claim that the CMIP6 SSP behavior is explained by this specific mechanism is not established.","section":"Section 4a and 4d"},{"comment":"The ssp119 case does not exhibit the hysteresis predicted by the Overshoot idealization, and the explanation that ssp119 temperatures decline too slowly is post hoc. The theory says pattern invariance requires exponential forcing; ssp119 forcing is not exponential, yet Figure 7a shows near-linear relationships for Land, TO and SO. If the mechanism is correct, deviations from linearity should be predictable from the forcing shape and regional time scales; the paper does not quantify when 'too slowly' means. This weakens the empirical support for the exponential-forcing condition as the cause of SSP pattern invariance.","section":"Section 4a, Figure 7"},{"comment":"All CMIP6 comparisons use multi-model averages without showing intermodel spread. Pattern-scaling linearity in the mean can be much stronger than in individual models, especially if models have different internal variability or different regional responses. Since the paper claims the conditions are 'approximately met in most CMIP6 SSP projections', it should show the distribution across models (e.g., regression R² or pattern error per model) for ssp119, ssp585, and abrupt-4xCO2.","section":"Section 2 and Figure 7"}],"minor_comments":[{"comment":"The text refers to 'ssp126, ssp585 and abrupt-4xCO2' but the scenarios studied are ssp119, ssp585, and abrupt-4xCO2; this appears to be a typo.","section":"Section 2"},{"comment":"The phrase 'An extendend pattern scaling approach' should read 'An extended pattern scaling approach'.","section":"Section 4b"},{"comment":"The sentence 'the difference between between the individual patterns' contains a duplicated 'between'.","section":"Section 4c"},{"comment":"The caption uses 'DiffAbrupt' while Table 1 uses 'DiffusionAbrupt'; the labels should be consistent.","section":"Figure 6 caption"}],"recommendation":"major_revision","confidential_remarks":"The mathematical core is sound and the paper is likely within scope for a climate journal. The main risk is that the empirical bridge between the idealized theory and CMIP6 behavior is incomplete, particularly for the dynamics assumption and the ssp119 mismatch. I would not reject: the limitations are openly acknowledged and the framework is useful, but the authors should either provide an independent test of the linear time-invariant dynamics assumption or substantially soften the claim that the conditions are approximately met."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is a genuinely useful paper. It takes the old observation that pattern scaling works in SSPs and the pattern-effect literature that says patterns evolve, and shows they are consistent once you recognize that exponential forcing makes a linear system respond with a fixed spatial pattern. The derivation in Section 3 is clean: Eq. 9 is correct, and the three-region example nicely illustrates why abrupt forcing and overshoot break pattern scaling while exponential forcing does not. I also like that they do not tune their idealized parameters to CMIP6; they use Armour's and Merlis's numbers, and then test predictions qualitatively. That is honest work.\n\nThe main soft spot is exactly the one in the stress-test note. The bridge from the ODE to CMIP6 rests on the assumption that local ocean heat uptake and heat transport changes can be represented by a constant effective heat capacity and a linear, time-invariant operator. The paper never independently tests that. Figure 7 shows local-global regressions are linear, but that linearity is the pattern scaling property itself—not a test of the proposed mechanism. The Gregory regressions in Figure 8 show feedbacks are roughly linear outside the Arctic, which supports assumption (ii), but they say nothing about C(r) being constant or about heat transport being representable as a diffusive linear term. Section 4d acknowledges the MHT treatment is 'highly simplified' and that advective/eddy processes are not addressed. That is a real gap, and it is load-bearing.\n\nTwo smaller things. The ssp119 comparison is weak: the predicted hysteresis does not appear in the multi-model mean and the authors explain it post hoc by saying temperatures decline too slowly. And the CMIP6 plots are multi-model means without spread, so we do not know how robust the qualitative agreement is across models.\n\nThat said, the central argument holds up as a reconciliation. Even if real ocean dynamics are nonlocal and time-varying, the eigenfunction logic applies to any linear time-invariant operator; the open empirical question is whether the real system is close enough to LTI. The authors are upfront about this. This paper is worth engaging seriously. I would circulate it in the reading group—it clarifies a conceptual confusion that has been floating around for a decade. A referee should ask for a quantitative test of the dynamics assumption and model spread, but this is not a desk reject.","headline":"A clean analytical reconciliation of pattern scaling and the pattern effect; the empirical bridge to CMIP6 is qualitative but the paper is honest about it.","tokens_in":25001,"tokens_out":2073,"would_cite":true,"duration_ms":19970,"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":"A local energy-balance theory explains why the climate's warming pattern stays fixed in common future scenarios and only changes when forcing is abrupt or nonlinear, reconciling pattern scaling with the pattern effect.","keywords":["pattern scaling","warming pattern","pattern effect","climate sensitivity","local energy balance","exponential forcing","CMIP6","effective heat capacity"],"falsifier":"In an overshoot experiment where global mean temperature crosses the same threshold twice, the theory predicts a slow region (e.g., the Southern Ocean) will show two different local temperatures at the two crossings. A coupled-model overshoot run in which regional temperatures are identical at both crossings would refute the claim that pattern evolution is driven by regional time scales under nonexponential forcing.","tokens_in":23987,"feed_emoji":"🌍","tokens_out":6850,"duration_ms":56788,"temperature":0.7,"pith_summary":"Pattern scaling, the widely used method that multiplies the global average temperature anomaly by a fixed local pattern, works because of a specific mathematical property: exponential forcing, linear feedbacks, a constant forcing pattern, and linear diffusive dynamics force the ratio $\\Delta T(\\mathbf{r},t)/\\Delta T(t)$ to be exactly time-independent. The authors show this with a local energy balance where each region has an effective heat capacity and a stabilizing feedback, and they demonstrate that the conditions are approximately met in most CMIP6 Shared Socioeconomic Pathway projections. The same theory explains why abrupt-4xCO2 and overshoot experiments show evolving patterns: non-exponential forcing activates different regional time scales, producing the 'pattern effect' without requiring nonlinear feedbacks. If correct, the result explains why simple linear emulators have been hard to beat for end-of-century temperatures and clarifies where they should fail.","feed_headline":"Exponential CO2 growth keeps warming maps stable in climate models","feed_subtitle":"A local energy-balance theory explains why pattern scaling works in end-of-century projections but not under abrupt CO2 forcing.","key_machinery":"The key object is the local top-of-atmosphere energy balance for an atmosphere-ocean column, written as $C(\\mathbf{r})\\, \\partial \\Delta T(\\mathbf{r},t)/\\partial t = P_R(\\mathbf{r})\\, R(t) + \\lambda(\\mathbf{r})\\, \\Delta T(\\mathbf{r},t)$, optionally augmented with Sellers diffusion. $C(\\mathbf{r})$ is the column's effective heat capacity, $\\lambda(\\mathbf{r})$ is the local time-invariant feedback parameter, and $P_R(\\mathbf{r})$ is the time-invariant forcing pattern. This linear system has exponential eigenfunctions: with forcing $R(t) = R_0 e^{t/\\tau_0}$, the solution factorizes as $\\Delta T(\\mathbf{r},t) = A(\\mathbf{r}) e^{t/\\tau_0}$, so the ratio $\\Delta T(\\mathbf{r},t)/\\Delta T(t) = A(\\mathbf{r})/A$ is exactly constant. The regional time scale $\\tau(\\mathbf{r}) = -C(\\mathbf{r})/\\lambda(\\mathbf{r})$ is what determines whether a scenario keeps that constancy: exponential forcing locks all regions to the forcing time scale, while abrupt or peaking forcing lets the spread of $\\tau(\\mathbf{r})$ emerge, making the pattern time-dependent.","core_discovery":"The paper's central claim is that the warming pattern $\\Delta T(\\mathbf{r},t)/\\Delta T(t)$ is exactly independent of time and of forcing magnitude whenever the climate response is linear, the forcing pattern is fixed in space, the forcing grows exponentially, and heat transport changes can be represented as a linear time-independent operator. Under these conditions every location warms at the same exponential rate, so the pattern is set only by the spatial distribution of effective heat capacity, local feedback, and forcing, not by the calendar time or scenario. Relaxing the exponential-forcing condition (constant or peaking forcing) makes regional time scales $\\tau(\\mathbf{r}) = -C/\\lambda$ emergent, and the pattern evolves as fast and slow regions adjust at different rates; this is the mechanism the paper proposes for the pattern effect in abrupt-4xCO2 experiments. The Arctic breaks the linear-feedback condition through a temperature-dependent albedo feedback, and regions with large scenario-dependent aerosol forcing break the fixed forcing-pattern condition; the paper uses CMIP6 data to show these are exactly the places where pattern scaling fails.","pith_inferences":["Deviations from pattern scaling in CMIP6 output could be used diagnostically: regions where local-to-global regressions curve are precisely where the linearity assumptions (fixed feedbacks, separable forcing, diffusive dynamics) break, which is a more direct test than comparing emulator skill.","The theory implies historical reconstructions, with their regionally shifting aerosol forcing, should exhibit pattern evolution similar to the paper's AerosolForcing case; this is a quantitative prediction that could be checked against observed and modeled twentieth-century warming maps.","The pattern ratio in exponential-forcing scenarios is an invertible function of local feedback, heat capacity, and forcing pattern, suggesting that pattern scaling regressions could be used as an inverse method to estimate regional climate response parameters from existing CMIP6 runs.","The reconciliation suggests that impact assessments relying on pattern scaling should state the forcing regime explicitly; transferring patterns across scenarios is only justified when the scenarios share the near-exponential, fixed-pattern conditions under which the pattern is mathematically invariant."],"forward_implications":["Pattern scaling is a sound approximation for the SSP1-2.6 and SSP5-8.5 projections in CMIP6 over land, the tropical ocean, and the Southern Ocean, because their forcing remains near-exponential.","The pattern effect in idealized abrupt-4xCO2 experiments is explained by spatially varying regional time scales under non-exponential forcing, and does not by itself require nonlinear feedbacks.","In SSP projections, deviations from pattern invariance are concentrated in the Arctic (nonlinear albedo feedback) and in regions with strong scenario-dependent aerosol forcing such as East Asia and eastern North America.","For overshoot scenarios, where global forcing peaks and declines, a single time-invariant pattern is insufficient, so pattern scaling errors will be largest for slow regions such as the Southern Ocean.","The same mechanism underlies the near-constancy of the Regional Transient Climate Response to cumulative emissions (RTCRE) and its global counterpart (TCRE), tying local scaling to carbon budget calculations."],"supporting_citations":[{"why":"Supplies the three-region energy-balance model and the time-varying regional feedback framework that the paper's transient solutions build on.","marker":"Armour et al. 2013"},{"why":"Provides the exponential-eigenfunction result for linear carbon-climate systems that underpins the pattern-invariance derivation.","marker":"Raupach 2013"},{"why":"Defines the diffusive heat-transport parameterization used to show that linear time-independent dynamics preserve pattern invariance.","marker":"Sellers 1969"},{"why":"Establishes the local regression method the paper uses to test linear feedbacks in CMIP6 abrupt-4xCO2 output.","marker":"Gregory et al. 2004"},{"why":"Supplies the CO2 concentration and radiative-forcing formula showing why SSP forcing is approximately exponential.","marker":"Meinshausen et al. 2020"},{"why":"Documents the pattern scaling method whose success this paper aims to explain.","marker":"Tebaldi and Arblaster 2014"},{"why":"Defines the Regional Transient Climate Response to cumulative emissions that the paper's mechanism also explains.","marker":"Leduc et al. 2016"},{"why":"Identifies fast and slow warming components used to interpret the difference between SSP and abrupt experiments.","marker":"Held et al. 2010"}],"fun_headline_variants":["Why climate warming maps stay stable in projections","New theory reconciles stable warming patterns and pattern effect","Pattern scaling works when forcing grows exponentially","Arctic and aerosols break invariant warming pattern","Climate model warming pattern invariance explained by energy balance"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on representing ocean heat uptake and heat transport changes as a local linear response with fixed effective heat capacity and at most diffusive transport; if real ocean circulation changes are nonlocal and time-varying enough to matter, the mechanism would not explain the invariance.","fun_headline_variants_meta":{"raw":{"variants":["Why climate warming maps stay stable in projections","New theory reconciles stable warming patterns and pattern effect","Pattern scaling works when forcing grows exponentially","Arctic and aerosols break invariant warming pattern","Climate model warming pattern invariance explained by energy balance"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000172,"raw_usage":{"total_tokens":1304,"prompt_tokens":1004,"completion_tokens":300,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":620,"completion_tokens_details":{"reasoning_tokens":232}},"tokens_in":620,"tokens_out":300,"duration_ms":3339,"temperature":1.0,"reasoning_tokens":232,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:26:56.428988+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"In an overshoot experiment where global mean temperature crosses the same threshold twice, the theory predicts a slow region (e.g., the Southern Ocean) will show two different local temperatures at the two crossings. A coupled-model overshoot run in which regional temperatures are identical at both crossings would refute the claim that pattern evolution is driven by regional time scales under nonexponential forcing.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the three-region energy-balance model and the time-varying regional feedback framework that the paper's transient solutions build on."},{"cited_title":"R., 2013: The exponential eigenmodes of the carbon-climate system, and their implications for ratios of responses to forcings","cited_arxiv_id":null,"evidence_quote":"Provides the exponential-eigenfunction result for linear carbon-climate systems that underpins the pattern-invariance derivation."},{"cited_title":"M., and Coauthors, 2004: A new method for diagnosing radiative forcing and climate sensitivity","cited_arxiv_id":null,"evidence_quote":"Establishes the local regression method the paper uses to test linear feedbacks in CMIP6 abrupt-4xCO2 output."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the Regional Transient Climate Response to cumulative emissions that the paper's mechanism also explains."}],"review_version":1}