{"id":"24ccbf32-09a0-4878-a697-2c83f6c6c048","arxiv_id":"2502.04982","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Rough continuity and transport equations are well-posed for Osgood and DiPerna-Lions drifts, yielding a rough Yudovich theorem for 2D Euler and a continuous random dynamical system.","lead":"This paper proves that rough, non-smooth noise can be added to fluid transport equations without destroying unique solvability, even when the fluid velocity is only mildly regular (Osgood or DiPerna-Lions class). It establishes a rough-path analogue of Yudovich's theorem for 2D Euler equations in vorticity form and builds a continuous random dynamical system for fractional Brownian noise.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the central Yudovich-type theorem is internally coherent under the explicitly stated geometric-rough-path and divergence-free assumptions.","rationale":"The reader's acceptance is justified. The central claim, Theorem 1.5, depends on the geometric rough path and divergence-free conditions, but these are explicitly imposed and are natural for Stratonovich-type transport noise. The proof chain is long but each step is supported: the RDE flow theory gives the homeomorphism flow, Corollary 3.12 gives quasi-incompressibility via the geometric rough path and divergence-free assumption, the linear theory provides uniqueness and renormalization, and Proposition 5.8 closes the nonlinear Yudovich argument with an Osgood modulus. My review found no internal inconsistency or unstated regularity assumption that would invalidate the main theorem. The minor display issue in Proposition 5.8 is not load-bearing because the correct weighted estimate follows from the stated pushforward relations and Assumption 5.1. Therefore the verdict should remain unchanged.","tokens_in":78656,"tokens_out":35247,"duration_ms":384381,"concrete_test":"Re-derive the uniqueness estimate of Proposition 5.8 keeping all pushforward weights explicit: verify that I_s^2 is bounded by a constant times ||rho^2_s||_{L1 cap L^infty} times the integral of |rho0(y)| h(|Phi^1_s(y)-Phi^2_s(y)|) dy, and check that Jensen's inequality with the normalized measure |rho0|/||rho0||_1 gives the Osgood--Gronwall closure that yields I_t identically zero.","verdict_should_be":"UNCHANGED","load_bearing_attack":"After tracing the chain from the Osgood RDE flow (Theorem 3.9), quasi-incompressibility of the flow (Corollary 3.12), linear DiPerna--Lions theory (Theorem 4.29 and Proposition 4.33), and the nonlinear comparison argument (Proposition 5.8), the proof of Theorem 1.5 closes without an internal gap. The genuinely load-bearing hypothesis is that Z is a geometric p-rough path and the transport vector fields are divergence-free; these enter the Jacobian determinant formula (40), the conservativity relation (59), and the quasi-incompressibility estimate used for the L^p preservation (131). This restriction is stated explicitly in the introduction and Section 4.1 and is standard for Stratonovich-type noise such as fractional Brownian motion with H in (1/3,1). It is a scope limitation, not a hidden inconsistency. The only soft spots encountered are typographical/sketch-level, such as a display in Proposition 5.8 that temporarily drops the |rho0| weight in the estimate for I_s^2; the required weighted inequalities and pushforward identities are available, so the argument can be closed without changing the theorem.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a rough-path theory for (nonlinear) continuity and transport equations with non-Lipschitz drifts. It first proves well-posedness and the existence of a homeomorphic flow for RDEs with Osgood drifts (Theorem 3.9), then establishes a DiPerna–Lions-type well-posedness theory for linear rough continuity and transport equations on R^d with geometric p-rough-path noise for p in [2,3), including a product formula and duality (Theorem 4.20). These linear results are combined with the RDE flow theory to obtain flow representations and quasi-incompressibility. For nonlinear continuity equations with convolutional kernels, the paper proves existence, uniqueness, stability, and flow representation (Theorems 1.4, 5.9), and then specializes to the 2D Euler equations in vorticity form, obtaining a rough Yudovich theorem (Theorem 1.5) and a random dynamical system when the noise is a geometric rough path cocycle such as fractional Brownian motion with H in (1/3,1). A weak-existence result for L^1 ∩ L^p vorticities is also given (Theorem 1.6).","tokens_in":78851,"tokens_out":18379,"duration_ms":162662,"significance":"If the results are correct, this is a substantial contribution to the rough-PDE literature. It extends classical DiPerna–Lions and Yudovich theory to geometric rough-path noise, handles the full space, and constructs a continuous random dynamical system for rough 2D Euler, going beyond the partial overlap in [RT24]. The proofs are detailed and largely self-contained, and the main theorems are stated with explicit regularity assumptions. The paper also provides useful new tools, notably the RDE flow theory for Osgood drifts on p-variation spaces and the product-formula-based uniqueness argument. The geometric rough-path and divergence-free assumptions are clearly stated and are standard for Stratonovich-type noise, so they represent a scope limitation rather than a hidden inconsistency. The main issue I found is a load-bearing gap in the uniqueness proof for the nonlinear equation, which appears repairable.","major_comments":[{"comment":"The displayed computation of I^2_s is incorrect as written. The first equality should read I^2_s = ∫ | ∫ [K_s(x,Φ^1_s(y)) - K_s(x,Φ^2_s(y))] ρ0(y) dy | |ρ2_s(x)| dx; the factor ρ0(y) in the inner integral is missing. Consequently, the next bound by ∫ h(|Φ^1_s(y)-Φ^2_s(y)|) dy is unjustified, and in fact cannot hold with the unweighted Lebesgue measure: for the Biot-Savart kernel the double integral with respect to Lebesgue measure is not controlled by the displayed expression, and the y-integral is not controlled by I_s unless the measure |ρ0|dy is carried through. This step is load-bearing because it produces the Osgood inequality I_t ≲ ∫ ... h(I_s) ds that yields uniqueness in Proposition 5.8 and hence in Theorems 1.4 and 1.5. The gap is repairable: one should write I^2_s ≤ ∫ |ρ0(y)| [∫ |K_s(x,Φ^1_s(y))-K_s(x,Φ^2_s(y))| |ρ2_s(x)| dx] dy, apply Assumption 5.1 with f = ρ2_s in the K̃ form to bound the inner integral by C h(|Φ^1_s(y)-Φ^2_s(y)|) ||ρ2_s||_{L1∩L∞}, and then use Jensen's inequality with the finite measure |ρ0|dy; the proof should be amended accordingly.","section":"§5.2, Proposition 5.8 (estimate of I^2_s)"}],"minor_comments":[{"comment":"In the estimate for I^1_s, the constant depending on ||ρ0||_{L1} is suppressed; please write it explicitly, since Jensen's inequality with the finite measure |ρ0|dy introduces such a constant and h is only subadditive/concave, not homogeneous.","section":"§5.2, Proposition 5.8 (estimate of I^1_s)"},{"comment":"In the final condition of (77), \"\\tilde R ≥ κ‖h‖_{L^1_x}\" should read \"\\tilde R ≥ κ‖h‖_{L^1_t}\" because h is a time-dependent function introduced in (72).","section":"§4.2, Lemma 4.17, display (77)"},{"comment":"The word \"beloging\" should be \"belonging\".","section":"Definition 5.5, item ii"},{"comment":"The proof invokes Corollary 4.35, which is stated for the continuity equation; since the statement concerns the transport equation, either the transport analogue should be stated explicitly or the proof should mention that the same argument applies verbatim using Corollary 4.32 and condition (109).","section":"§4.4, Corollary 4.36"},{"comment":"The notation \"for all q∈{p}∪(1,∞)\" is redundant when p>1 and slightly confusing when p=1; consider rewording to \"for all q∈[1,∞)\" in the strong-convergence statements.","section":"§1.1, Theorems 1.4 and 1.5"}],"recommendation":"major_revision","confidential_remarks":"The paper is a strong contribution and the identified gap in Proposition 5.8 is local and repairable. I do not see circularity or hidden inconsistency; the geometric-rough-path and divergence-free restrictions are explicit and standard. I recommend revision rather than acceptance because the uniqueness proof of the central nonlinear theorem contains an incorrect displayed estimate that must be corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [Colleague],\n\nPunchline: this is the real thing. It proves a rough-path analogue of Yudovich's theorem for 2D Euler on the whole space, and along the way builds a DiPerna–Lions theory for linear rough continuity equations. If I had to point a skeptic at one paper to convince them the rough-programme works for low-regularity fluid noise, this would be it.\n\nWhat is actually new is exactly what the abstract promises. The main theorems are as advertised: RDEs with Osgood drifts (even non-geometric rough paths, via a time-reversal lemma that fills a real gap), linear RPDEs under DiPerna–Lions conditions, and the nonlinear framework that covers the 2D Euler vorticity equation. The overlap with [RT24] is real but partial—they do torus Euler and log-Lipschitz RDEs; this paper does full space, general kernels, linear theory, and weak existence for L1∩Lp. The authors are honest about the relation.\n\nThe proofs are long and technical, but the architecture is sound. The product formula plus duality is used in place of renormalization as the central tool, which is a smart move and works. The stress test I ran—tracing from the Osgood flow to quasi-incompressibility to linear uniqueness to the nonlinear comparison—closes without an internal gap. The only soft spot of note is a display in Prop 5.8 that temporarily drops a |rho0| weight; the needed weighted estimate is available, so it's a typo not a gap.\n\nThe main scope limitation is real: everything requires geometric rough paths (Stratonovich-type noise) and divergence-free ξ. That's not a flaw—it's explicit and standard for the setting—but it does mean Itô-type noise is out. Similarly, ξ ∈ C^3_b is stronger than in the Brownian SPDE literature, a price of rough path theory. The paper leans heavily on the authors' own unbounded rough drivers framework, but that prior work is established and independently cited; self-citation here is not a red flag.\n\nWho this is for: anyone working on rough PDEs, stochastic transport, or 2D Euler with transport noise. It deserves a serious referee; the length will make refereeing painful, but the content justifies it. My recommendation: send to peer review. I'd accept it after the usual minor fixes.","headline":"A dense but solid paper that delivers the first rough-path Yudovich theorem on R^2 and a DiPerna–Lions theory for rough linear PDEs; deserves serious refereeing.","tokens_in":79421,"tokens_out":2893,"would_cite":true,"duration_ms":28298,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60L20","60L50","60H15","35R60","35Q31"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that the classical Yudovich theorem for 2D Euler extends to rough transport noise: initial vorticities in $L^1_x\\cap L^\\infty_x$ produce unique global solutions to the rough vorticity equation, with flow representation…","keywords":["rough partial differential equations","rough 2D Euler","Yudovich theorem","DiPerna-Lions theory","unbounded rough drivers","flow representation","fractional Brownian motion","geometric rough paths"],"falsifier":"Compute the Jacobian determinant of the rough flow $\\Phi_t$ generated by $\\mathrm{d}y_t=(K*\\omega_0)(y_t)\\,\\mathrm{d}t+\\xi(y_t)\\,\\mathrm{d}Z_t$ for a smooth vortex $\\omega_0$ and a fixed geometric rough path $Z$. The paper's quasi-incompressibility formula says the determinant equals $\\exp(\\int_0^t \\nabla\\cdot b_s\\,\\mathrm{d}s)$ with no noise term; if any numerical rough-path integrator shows dependence of $\\det D\\Phi_t$ on the rough area $Z_{st}$, the quasi-incompressibility identity fails and the theory collapses.","tokens_in":78457,"feed_emoji":"🌀","tokens_out":7579,"duration_ms":73519,"temperature":0.7,"pith_summary":"The paper proves that the classical Yudovich theorem for 2D Euler equations extends to rough transport noise: initial vorticities in $L^1_x\\cap L^\\infty_x$ produce unique global solutions to the rough vorticity equation, with no Sobolev regularity required. The solutions are renormalized, propagate by the rough flow $\\omega_t(x)=\\omega_0(\\Phi_t^{-1}(x))$, and conserve every $L^p_x$ norm. Well-posedness holds for geometric rough paths of finite $p$-variation with $p\\in[2,3)$, which covers fractional Brownian motion with Hurst parameter $H\\in(1/3,1)$ and yields a continuous random dynamical system. This matters because it gives a solution theory for stochastic fluid equations driven by non-Markovian, rough signals at the same level of regularity as in the deterministic theory.","feed_headline":"Yudovich's 2D Euler theorem survives rough transport noise","feed_subtitle":"Unique global vortex solutions with conserved norms now exist for rough paths and fractional Brownian noise.","key_machinery":"The argument runs on two coupled structures. On the Lagrangian side, a new solution theory for rough differential equations with Osgood non-Lipschitz drifts produces a continuous flow of homeomorphisms $\\Phi_t$ with explicit modulus estimates, built in finite $p$-variation spaces rather than H\\\"older spaces. On the Eulerian side, the unbounded rough drivers framework gives an intrinsic distributional meaning to rough continuity and transport equations; the load-bearing identity is the product formula and conservativity $A^*_{st}=-A_{st}$ of the rough driver, which follows from $\\nabla\\cdot\\xi=0$ and the geometricity of $Z$, yielding duality, uniqueness, and the flow representation. The quasi-incompressibility estimate for $\\Phi_t$, expressed through the Jacobian determinant formula, is what converts the Lagrangian flow into $L^p$ conservation for the vorticity.","core_discovery":"The central claim is that the Yudovich theory for 2D Euler survives rough transport noise. For a geometric rough path $Z$ of finite $p$-variation with $p\\in[2,3)$, divergence-free $\\xi\\in C^3_b$, and the Biot\\textendash Savart kernel $K$, every $\\omega_0\\in L^1_x\\cap L^\\infty_x$ gives a unique global solution to the rough vorticity equation, renormalized and of the form $\\omega_t(x)=\\omega_0(\\Phi_t^{-1}(x))$, with $\\|\\omega_t\\|_{L^p_x}=\\|\\omega_0\\|_{L^p_x}$ for all $p\\in[1,\\infty]$. The solution map depends continuously on initial data and on the rough path in both weak and strong topologies, and if $Z$ is a random geometric rough path cocycle the dynamics generate a continuous random dynamical system on bounded subsets of $L^1_x\\cap L^\\infty_x$. The same framework also yields a DiPerna\\textendash Lions type well-posedness theory for linear rough continuity and transport equations and weak existence for initial vorticities in $L^1_x\\cap L^p_x$.","pith_inferences":["Beyond the paper, the geometric requirement indicates that the theory targets Stratonovich-type or Wong\\textendash Zakai limits; It\\^o noise is explicitly excluded and would need a different duality argument.","The flow-representation and kernel estimates likely transfer to the torus and to bounded domains with slip boundary conditions, and to vanishing-viscosity limits of rough 2D Navier\\textendash Stokes, as the paper lists as future perspectives.","The Osgood-drift RDE theorem may serve as a building block for rough Vlasov\\textendash Poisson equations, where uniqueness relies on similar logarithmic moduli of continuity."],"forward_implications":["The rough 2D Euler vorticity equation is globally well-posed in $L^1_x\\cap L^\\infty_x$ for geometric rough paths, with unique renormalized solutions and flow representation.","The solution map generates a continuous random dynamical system when the noise is a fractional Brownian motion with Hurst parameter $H\\in(1/3,1)$, providing a pathwise framework for Wong\\textendash Zakai and large-deviation results.","Linear rough continuity and transport equations are well-posed under DiPerna\\textendash Lions regularity conditions, with product, duality, and renormalization formulas available.","For initial vorticities in $L^1_x\\cap L^p_x$, weak solutions to the rough 2D Euler equations exist for every $p\\in[1,\\infty)$, and they are renormalized when $p\\ge 2$.","The nonlinear continuity equation framework covers general convolutional kernels satisfying Osgood and integrability assumptions, going beyond the Biot\\textendash Savart case."],"supporting_citations":[{"why":"Supplies the deterministic 2D Euler well-posedness theorem, the Yudovich class, that the rough analogue extends.","marker":"[Yud63]"},{"why":"Supplies the DiPerna\\textendash Lions regularity conditions and the product/duality strategy adapted to rough PDEs.","marker":"[DL89]"},{"why":"Supplies the unbounded rough drivers framework used to define and solve rough continuity and transport equations.","marker":"[BG17]"},{"why":"Provides prior local well-posedness for rough Euler equations with Sobolev-regular vorticity, the result this paper improves to the Yudovich class.","marker":"[CHLN22a]"},{"why":"Establishes rough path theory, the basis for giving meaning to non-smooth driving signals $Z$.","marker":"[Lyo98]"},{"why":"Provides the rough path cocycle construction for fractional Brownian motion used to build the random dynamical system.","marker":"[BRS17]"},{"why":"Supplies standard rough path flow results for smooth drifts used in the approximation argument.","marker":"[FV10]"}],"fun_headline_variants":["Rough noise preserves 2D Euler uniqueness","Yudovich's result holds for rough paths","2D Euler well-posed with rough transport noise","Nonlinear rough continuity equations solved","Rough analogue of Yudovich theorem proven"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole construction requires the noise to be a geometric rough path, meaning a Stratonovich-type lift; if the same signal is given an It\\^o-type lift, the flow is no longer volume-preserving and the uniqueness argument collapses.","fun_headline_variants_meta":{"raw":{"variants":["Rough noise preserves 2D Euler uniqueness","Yudovich's result holds for rough paths","2D Euler well-posed with rough transport noise","Nonlinear rough continuity equations solved","Rough analogue of Yudovich theorem proven"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001055,"raw_usage":{"total_tokens":4465,"prompt_tokens":1019,"completion_tokens":3446,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":635,"completion_tokens_details":{"reasoning_tokens":3377}},"tokens_in":635,"tokens_out":3446,"duration_ms":24032,"temperature":1.0,"reasoning_tokens":3377,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T20:43:55.986183+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the Jacobian determinant of the rough flow $\\Phi_t$ generated by $\\mathrm{d}y_t=(K*\\omega_0)(y_t)\\,\\mathrm{d}t+\\xi(y_t)\\,\\mathrm{d}Z_t$ for a smooth vortex $\\omega_0$ and a fixed geometric rough path $Z$. The paper's quasi-incompressibility formula says the determinant equals $\\exp(\\int_0^t \\nabla\\cdot b_s\\,\\mathrm{d}s)$ with no noise term; if any numerical rough-path integrator shows dependence of $\\det D\\Phi_t$ on the rough area $Z_{st}$, the quasi-incompressibility identity fails and the theory collapses.","supporting_citations":[],"review_version":1}