{"id":"7fe91f5b-dac6-4532-859a-8ba7d68d4388","arxiv_id":"2507.03830","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Spatially localised doubly diffusive convection states exist in an axisymmetric spherical shell, forming equatorial-convectons, anticonvectons and pole-convectons via imperfect bifurcations.","lead":"This paper uses numerical simulation to find families of spatially localised convection states in a spherical shell where heat and salt diffuse at different rates. It is the first study of these 'convectons' in spherical geometry, and it shows how curvature forces localisation through imperfect bifurcations.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No resolution or convergence data is reported for the continuation; the existence of the slanted snaking branches, not just their quantitative shape, rests on an unverified spectral truncation.","rationale":"I read the central claim as existence of localised doubly diffusive states in the axisymmetric spherical shell and the associated imperfect-bifurcation mechanism. For this claim to hold, the numerical branches must be genuine converged solutions of the discretised PDE. The paper gives no truncation or convergence data anywhere in the continuation results, so the existence claim depends on an unverified numerical assumption. The reader's concern about 3D and zonal-flow perturbations is legitimate for physical relevance, but it would not falsify the stated existence claim, since axisymmetric solutions are invariant solutions of the full 3D equations and the paper explicitly frames the study as axisymmetric. I therefore see numerical resolution as the most load-bearing gap, with the 3D-stability issue as an important secondary limitation. The deposited code, the Dedalus validation, and the documented data availability are genuine supporting evidence, so the appropriate outcome remains the reader's CONDITIONAL verdict rather than REJECT; the proposed resolution test can settle whether the central claim is accepted.","tokens_in":18154,"tokens_out":9575,"duration_ms":117450,"concrete_test":"Using the deposited Zenodo code, recompute the LA-10 branch at RaS=350 (Figure 10) and the LC-10 branch at RaS=450 (Figure 9) at the published resolution and at doubled N_theta and N_r (with the 3/2-dealiased latitudinal basis). Compare the RaT coordinates of at least five successive saddle-nodes and the domain-filling connection point. If any saddle-node shifts by more than about 0.1% in RaT, or if any saddle-node appears or disappears, the branches are underresolved and the existence claim is not established. If all shifts are below tolerance, the central numerical claim is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central existence claim rests entirely on pseudo-spectral continuation (Section 5), yet the manuscript reports no truncation parameters (N_r, N_theta) and no convergence test for any computed branch. The branches in Figures 7-15 are delicate objects: saddle-node locations, the extent of slanted snaking, and the inferred imperfect bifurcations can all change if the Chebyshev or latitudinal truncation is insufficient. This is especially relevant for the ell=10 and ell=11 states, whose polar/equatorial amplitude variation is strong and whose branch structure involves repeated near-degenerate saddle-nodes. A spurious saddle-node or a shifted imperfect bifurcation would directly change the qualitative claim that localised states are forced to arise via imperfect bifurcations. Validation against Dedalus and the deposited Zenodo code are real positive evidence, but they establish the time-stepper and not the convergence of the arc-length continuation of unstable steady branches. The axisymmetric restriction is a scope limitation rather than a correctness flaw, because axisymmetric solutions are genuine invariant-subspace solutions of the full 3D equations; the more load-bearing assumption is that the reported branches are numerically resolved. Without a quantitative convergence check, the primary claim remains conditional.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies doubly diffusive convection in an axisymmetric spherical shell, imposing longitudinal invariance and no zonal flow (Section 2). Using a pseudo-spectral method with Chebyshev radial collocation and sine/cosine latitudinal bases, the authors perform numerical continuation of steady axisymmetric states and compute bifurcation diagrams. They report three families of spatially localised solutions: equatorial-convectons (C±), anticonvectons (A±), and pole-convectons (L±_11). The primary even-mode bifurcation is interpreted as a transcritical-saddle-node (TSN) bifurcation, the equatorial-convecton branches are disconnected from the conduction state, and the localised branches exhibit slanted snaking attributed to imperfect bifurcations induced by the spatial modulation of curved convection rolls. The results are discussed through an analogy with the Swift–Hohenberg equation on a disk.","tokens_in":18329,"tokens_out":4593,"duration_ms":49335,"significance":"If the reported branches are numerically converged, this is the first demonstration of spatially localised doubly diffusive convection in a spherical shell, and it offers a concrete scenario in which curvature replaces translational invariance with spatial modulation, forcing localised states via imperfect bifurcations. The paper is commendable for depositing code and data, validating its time-stepper against Dedalus, and using direct numerical continuation rather than fitting to target states, so the circularity burden is low. The axisymmetric restriction is a genuine scope limitation but not a correctness flaw, since axisymmetric solutions are genuine invariant-subspace solutions of the full equations. The central existence claim remains conditional on (i) a missing resolution/convergence study and (ii) unverified support for the TSN classification.","major_comments":[{"comment":"The manuscript does not report the spectral truncation (N_r, N_theta) used for the continuation runs, nor any spatial-resolution or convergence test for the computed branches. Saddle-node locations, the extent of slanted snaking, and the inferred imperfect bifurcation points (RaS ≈ 404, RaS ≈ 198, RaT ≈ 2674) are all quantities that can shift with insufficient truncation, and the existence of near-degenerate saddle-nodes on the ℓ = 10 and ℓ = 11 branches is exactly the kind of result that requires a convergence check. The deposited code and the Dedalus validation establish the time-stepper, but not the convergence of the arc-length continuation of unstable steady branches. Please add N_r, N_theta for representative states and a resolution study (e.g., comparing E or saddle-node RaT at two or more resolutions) for at least one branch per family.","section":"§5 (pseudo-spectral method), Figures 7–15"},{"comment":"The conclusion that the even-parity primary bifurcation is a codimension-two transcritical-saddle-node (TSN) bifurcation is invoked from [BC15] and is load-bearing for the claim that anticonvectons bifurcate directly from the conduction state and that the equatorial-convecton branches are disconnected by imperfect bifurcations. However, the problem solved here has no-slip boundary conditions (Eq. 25), a different streamfunction formulation (Eq. 20), and finite aspect ratio Γ = 8.9224; none of these are checked against the hypotheses of the cited hidden-symmetry result. Since the unfolding is explicitly left out of scope, the manuscript should at least verify the TSN structure numerically for this system, for example by checking the quadratic tangency of the two LA± branches at the bifurcation point or by comparing with a direct normal-form computation.","section":"§4 (Eq. 33) and §5.1"}],"minor_comments":[{"comment":"The phrase \"Other work investing the role of temporal symmetry\" should read \"investigating the role\".","section":"§1"},{"comment":"The word \"dealiase\" should be \"dealias\" in the description of the nonlinear term treatment.","section":"§5"},{"comment":"The paper would be strengthened by reporting a stability analysis in the axisymmetric subspace: currently only one stable large-amplitude equatorial-convecton is identified by time-stepping, which leaves unclear whether the other computed localised branches are realisable or transient.","section":"Figures 7–15"},{"comment":"The admission that symmetry-breaking equatorial-convectons could not be computed is an honest statement of limitation, but the statement \"We are convinced that these states exist\" is not supported by any numerical or analytical evidence and should be phrased as a conjecture.","section":"§5.2"},{"comment":"There are minor typographical issues in the references, e.g., \"Swift––Hohenberge quation\" in [LS09] should read \"Swift–Hohenberg equation\".","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely acceptable after a focused revision. The essential additions are a resolution/convergence study for the continuation and a numerical check of the TSN bifurcation claim; both are feasible within the manuscript's scope. The axisymmetric restriction is a defensible scope choice for a first study, and the code/data deposit is exemplary."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead arXiv:2507.03830. Bottom line: this is the first computation of spatially localised doubly diffusive convection in a spherical shell, and it is a credible piece of numerical work. The new content is real: the geometry change from planar to spherical removes translation invariance, the primary eigenmodes are spatially modulated, and the authors show that localised states arise directly from imperfect bifurcations rather than from secondary snaking instabilities. The branch families (equatorial-convectons, anticonvectons, pole-convectons) and the slanted snaking are new states, not repackaged planar results. The Swift-Hohenberg-on-a-disc analogy is used as interpretation, not as a fitted model, and the connection is drawn carefully.\n\nWhat the paper does well: the numerical setup follows established methods (pseudo-spectral continuation, Stokes preconditioning), the code and data are on Zenodo, and the code is validated against Dedalus. That is real evidence, and it is more than many papers in this area provide. The authors also state plainly that one family (symmetry-breaking equatorial convectons) could not be computed; that honesty does not undermine the main claim, since those states are conjectural.\n\nThe soft spots, in proportion. First, no resolution or convergence study is reported for the continuation. The branches in Figures 7-15 are delicate objects with near-degenerate saddle-nodes, and the claim that localised states are forced via imperfect bifurcations depends on the qualitative shape of these branches. The Dedalus validation checks the time-stepper, not the arc-length continuation of unstable steady states. This is the one gap that could change the qualitative conclusions, and it is fixable: report N_r, N_theta and show a convergence test for representative branches. Second, the TSN bifurcation classification rests on a cited hidden-symmetry result ([BC15]) rather than on a local normal-form computation here; that is acceptable but should be stated as inherited. Third, the axisymmetric restriction and absence of zonal flow is a real scope limitation. It is not a correctness flaw, because axisymmetric solutions are solutions of the full 3D equations in an invariant subspace, but the paper's broader relevance to spherical-shell dynamics depends on whether these states survive 3D perturbations. The authors say this explicitly; I do not fault them for it, but I would not overstate the astrophysical implications yet.\n\nWho benefits: pattern-formation theorists working on localised states, spherical convection, and Swift-Hohenberg analogies. For them, this deserves a serious referee. For me, the missing convergence data is a conditional, not a rejection.\n\nRecommendation: yes, send to peer review. Ask for a resolution/convergence statement before acceptance.","headline":"First numerical evidence for localised double-diffusive convection in a spherical shell, with a clear mechanism and deposited code; the main gap is missing resolution/convergence data for the continuation.","tokens_in":18859,"tokens_out":1994,"would_cite":true,"duration_ms":21609,"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":"In an axisymmetric spherical shell, curved convection rolls force localised convection states to arise through imperfect bifurcations.","keywords":["doubly diffusive convection","spatial localisation","convectons","spherical shell","snaking bifurcation","imperfect bifurcation","pattern formation"],"falsifier":"Run a direct three-dimensional simulation, or a linear stability analysis against azimuthal modes of order $m\\neq 0$, of the axisymmetric convectons computed at $Ra_S=400$, $\\Gamma=8.9224$ and at $Ra_S=150$, $\\Gamma=10.029$; if $m\\neq 0$ perturbations grow and destroy these states before they can be sustained, the reported bifurcation diagrams describe solutions that are not realisable in the full spherical shell.","tokens_in":17928,"feed_emoji":"🌐","tokens_out":8409,"duration_ms":90117,"temperature":0.7,"pith_summary":"Doubly diffusive convection in a flat fluid layer usually fills the domain, but this paper shows that in a spherical shell the rolls are curved and therefore vary in strength with latitude. That spatial modulation removes the possibility of forming spatially periodic convection, and localised convection states are instead forced to arise directly through imperfect bifurcations. The authors compute these states numerically in an axisymmetric shell and identify three families — equatorial-convectons, anticonvectons at both poles, and pole-convectons at one pole — together with their slanted snaking bifurcation branches. The result matters because it takes a phenomenon previously studied almost exclusively in planar layers into the spherical geometry relevant to planetary and astrophysical interiors, and because the mechanism is generic enough to carry over to other pattern-forming systems in curved geometries.","feed_headline":"Curved convection rolls localise in a spherical shell","feed_subtitle":"Three families of localised convection states arise through imperfect bifurcations, with snaking slanted by curvature.","key_machinery":"The mechanism that carries the argument is the latitudinal modulation of the eigenmodes of the conduction state: the Legendre modes that destabilise the shell have their largest amplitude at the poles because convection rolls there cover a smaller surface area, so they must be stronger to transport the same heat or solute. This modulation breaks the translational invariance of the planar problem and, together with a hidden symmetry that makes the even-parity primary bifurcation a codimension-two transcritical–saddle-node (TSN) event, causes localised states rather than periodic states to emerge from the primary bifurcation. The bifurcation diagrams are obtained by spectral numerical continuation in a streamfunction formulation, using the solutal Rayleigh number as a homotopy parameter to expose the imperfect bifurcations and the slanted snaking.","core_discovery":"The central claim is that spatially localised doubly diffusive convection states exist in an axisymmetric spherical shell and that their origin is geometrical: convection rolls wrapping around the inner sphere are stronger near the poles, where the contact surface is smaller, so the linear eigenmodes are latitudinally modulated rather than periodic. Consequently the primary bifurcation does not create a spatially periodic pattern, and polar localised states bifurcate directly from the conduction state, while equatorial-convectons form branches disconnected from it. Continuing these branches numerically reveals slanted snaking — repeated saddle-node bifurcations with a leftward drift in the thermal Rayleigh number — organised by imperfect bifurcations that connect or disconnect the families as the solutal Rayleigh number is varied. The paper reports six branch types grouped in three families: C± equatorial-convectons, A± anticonvectons, and L± pole-convectons.","pith_inferences":["The axisymmetric restriction is an idealisation: a linear stability check of the computed convectons against azimuthal modes of order $m\\neq 0$ would show which of the reported families survive in a truly three-dimensional spherical shell.","Tracking the slanted snaking as the shell aspect ratio increases should recover the planar aligned-snaking limit, giving a quantitative test of how much of the spherical scenario is due to curvature.","If the curvature mechanism is generic, similar localised states should appear in spherical-shell magnetoconvection or rapid rotation problems, where subcritical transitions have been suggested; that connection is not explored in this paper.","The paper's inability to compute symmetry-breaking equatorial-convectons leaves an explicit gap; the predicted states could be searched for by starting from the symmetry-breaking pole-convecton branches and continuing through the parameter space the paper did not cover."],"forward_implications":["In an axisymmetric spherical shell, localised convection states are the natural primary products of the instability rather than secondary structures that require a separate amplitude-modulating bifurcation.","The solutal-to-thermal buoyancy balance selects the pattern: polar localisation is favoured at low solutal Rayleigh numbers and equatorial localisation at higher ones.","Spherical curvature and the absence of top-down symmetry make the snaking slanted and sometimes destroy saddle-nodes, so the spherical bifurcation diagrams differ quantitatively from the aligned snaking of planar doubly diffusive convection.","The same curvature-induced localisation mechanism is expected to operate in other axisymmetric pattern-forming systems, by analogy with the disk pattern-forming equation.","The computed axisymmetric families provide the starting bifurcation structure for fully three-dimensional spherical-shell studies, where additional localised states breaking longitudinal invariance are anticipated."],"supporting_citations":[{"why":"Supplies the planar doubly diffusive convection setup and parameter values whose snaking bifurcation diagram is the baseline for the spherical-shell comparison.","marker":"[BBK11]"},{"why":"Provides the antipodal symmetry RAP and the hidden-symmetry argument that turn the even-parity primary bifurcation into a transcritical–saddle-node bifurcation.","marker":"[BC15]"},{"why":"Establishes localised radial solutions of the axisymmetric disk pattern-forming equation, the analogue used for pole-convectons and anticonvectons.","marker":"[LS09]"},{"why":"Computes spot and target patterns of the disk pattern-forming equation, providing the bifurcation structure expected for spherical localised states.","marker":"[VKU21]"},{"why":"Introduces the convecton/anticonvecton terminology and planar localised states in binary fluid convection with non-standard boundary conditions, the planar analogues of the families found here.","marker":"[MBAK11]"},{"why":"Describes homoclinic snaking in bounded domains and the imperfect bifurcations that appear when translational invariance is absent, the mechanism invoked for the spherical-shell branches.","marker":"[HK09]"},{"why":"Supplies the streamfunction formulation and spectral projection method used for the numerical simulations and continuation.","marker":"[MT87]"}],"fun_headline_variants":["Curvature forces localised convection in spherical shell","Spherical shell yields localised double-diffusive convection","Slanted snaking: localised convectons in a sphere","Polar and equatorial convectons emerge from curvature"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The computed families all assume the flow is independent of longitude and has no zonal component ($\\partial/\\partial\\phi=0$ and $u_\\phi=0$), so the axisymmetric bifurcation diagrams may not survive three-dimensional perturbations.","fun_headline_variants_meta":{"raw":{"variants":["Curvature forces localised convection in spherical shell","Spherical shell yields localised double-diffusive convection","Slanted snaking: localised convectons in a sphere","Polar and equatorial convectons emerge from curvature"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000161,"raw_usage":{"total_tokens":1244,"prompt_tokens":962,"completion_tokens":282,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":578,"completion_tokens_details":{"reasoning_tokens":224}},"tokens_in":578,"tokens_out":282,"duration_ms":3593,"temperature":1.0,"reasoning_tokens":224,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:00:54.505708+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a direct three-dimensional simulation, or a linear stability analysis against azimuthal modes of order $m\\neq 0$, of the axisymmetric convectons computed at $Ra_S=400$, $\\Gamma=8.9224$ and at $Ra_S=150$, $\\Gamma=10.029$; if $m\\neq 0$ perturbations grow and destroy these states before they can be sustained, the reported bifurcation diagrams describe solutions that are not realisable in the full spherical shell.","supporting_citations":[],"review_version":1}