Manifolds of vortex loops as coadjoint orbits

Ioana Ciuclea, Cornelia Vizman · 2026

Read Manifolds of vortex loops as coadjoint orbits by Ioana Ciuclea, Cornelia Vizman (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-17861.txt.

Introduction

Ioana Ciuclea and Cornelia Vizman study vortex loops through the lens of group geometry. Their 2026 preprint treats manifolds of such loops as coadjoint orbits of the group of area-preserving maps of the plane. A vortex loop is a closed curve carrying a vorticity density that may drop to zero at points. The result ties fluid shapes to the algebra of the group that moves them.

Catalog record

CREATORIoana Ciuclea, Cornelia Vizman
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE29 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-17861

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Manifolds of vortex loops as coadjoint orbits Ioana Ciuclea, Cornelia Vizman Abstract We study a class of coadjoint orbits of the area preserving diffeomorphism group of the plane consisting of vortex loops, namely closed curves in the plane decorated with one-forms (vorticity densities) allowed to have zeros. 2 2 footnotetext: West University of Timişoara, 300223-Timişoara, Romania. [email protected] 3 3 footnotetext: West University of Timişoara, 300223-Timişoara, Romania. [email protected] Keywords: coadjoint orbit, vortex loop, Hamiltonian group 1 Introduction Euler’s equations for an ideal fluid, \dot{\operatorname{v}}+\nabla_{\operatorname{v}}\operatorname{v}=-\nabla p , \operatorname{div}\operatorname{v}=0 , with \operatorname{v} denoting the fluid velocity and p the pressure, are geodesic equations on the volume preserving diffeomorphism group [ 1 , 4 ] . In particular, the vorticity \operatorname{curl}\operatorname{v} is confined to a coadjoint orbit of this group. In this article we focus on singular vorticities with 1-dimensional support in the two dimensional case. More precisely, we study a class of coadjoint orbits of the area preserving diffeomorphism group of the plane consisting of vortex loops, with the vorticity density allowed to have zeros. Here the area preserving diffeomorphism group is identified with \operatorname{Ham}_{c}(\mathbb{R}^{2}) , the group of compactly supported Hamiltonian diffeomorphisms of \mathbb{R}^{2} . Its Lie algebra can be seen as the space of compactly supported smooth functions on \mathbb{R}^{2} , whose dual is the space of distributions. Vortex loops are the 2D version of vortex sheets (see [ 2 , 6 ] ). These are pairs (C,\beta_{C}) with C denoting a closed plane curve and \beta_{C} a volume form (its vorticity density): \langle(C,\beta_{C}),h\rangle=\int_{C}h\beta_{C},\quad h\in C^{\infty}_{c}(\mathbb{R}^{2}). We denote by \mathcal{O}_{\textsf{w}}^{a} the set of vortex loops with the same enclosed area a and the same total vorticity \textsf{w}=\int_{C}\beta_{C} . This data forms a complete set of invariants of the coadjoint \operatorname{Ham}_{c}(\mathbb{R}^{2}) action on vortex loops. These coadjoint orbits are described in [ 5 ] (see also [ 3 ] ). In this paper, we expand the study of vortex loops to include the case of a vorticity density with a finite number of nondegenerate zeros. These are pairs (C,\beta_{C}) with C an oriented closed plane curve and \beta_{C} a one-form on C with finitely many nondegenerate zeros (i.e. a Morse form). Choosing one zero as “starting point” allows us to regard them as a k -tuple (x_{1},\dots,x_{k}) and to introduce k partial vorticities along the loop, \textsf{w}_{i}=\int_{x_{i}}^{x_{i+1}}\beta_{C} , regarded as a k -tuple \bar{\textsf{w}}=(\textsf{w}_{1},\dots,\textsf{w}_{k}) . We denote by \mathcal{O}_{\bar{\textsf{w}}}^{a} the set of vortex loops with the same enclosed area a and the same \bar{\textsf{w}} , as this data forms a complete set of invariants for the coadjoint \operatorname{Ham}_{c}(\mathbb{R}^{2}) action on vortex loops. Let (S^{1},\beta) be a model for the vortex loop (C,\beta_{C}) . It defines, together with the canonical area form \omega , a closed 2-form on the manifold of embeddings enclosing a fixed area \operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2}) , \Omega_{f}(u_{f},v_{f})=\int_{S^{1}}\omega(u_{f},v_{f})\beta, which fails to be symplectic if \beta is a volume form. In this case, the coadjoint orbit \mathcal{O}_{\bar{\textsf{w}}}^{a} is identified with the quotient manifold \operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2})/\operatorname{Rot}(S^{1}) , which can be obtained via symplectic reduction in the Marsden-Weinstein ideal fluid dual pair [ 13 ] . In this paper, we show that \Omega is symplectic when \beta is a Morse one-form (Proposition 3.2 ). When the partial vorticities have no rotational symmetry, the coadjoint orbit \mathcal{O}_{\bar{\textsf{w}}}^{a} is identified with \operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2}) and \Omega is its Kirillov-Kostant-Souriau (KKS) symplectic form (Remark 4.6 ). When there is a (smallest) step \ell rotational symmetry of \bar{\textsf{w}} , the coadjoint orbit \mathcal{O}_{\bar{\textsf{w}}}^{a} is identified with the quotient manifold \operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2})/\mathbb{Z}_{k/\ell} (Corollary 4.5 ), with \mathbb{Z}_{k/\ell} denoting the (discrete) cyclic group of degree k/\ell and the KKS symplectic form descending from \Omega . The vortex loops considered here resemble the pointed vortex loops in [ 3 ] . These are singular vorticities that combine features of point vortices and vortex loops. More precisely, they are triples \left(C,\beta_{C},(x_{1},\dots,x_{k})\right) with C a closed plane curve, \beta_{C} a nowhere zero vorticity density and (x_{1},\dots,x_{k}) point vortices attached to the loop: \left\langle\left(C,\beta_{C},(x_{1},\dots,x_{k})),h\right)\right\rangle=\int_{C}h\beta_{C}+\sum_{i=1}^{k}\Gamma_{i}h(x_{i}),\quad h\in C_{c}^{\infty}(\mathbb{R}^{2}) where \Gamma_{i} denotes the vorticity of the point x_{i} . The point vortices introduce partial vorticities along the loop, namely \textsf{w}_{i}=\int_{x_{i}}^{x_{i+1}}\beta_{C} . We denote by \mathcal{O}_{\bar{\textsf{w}}}^{a} the set of pointed vortex loops with the same enclosed area a and the same partial vorticities. The model for a pointed vortex loops is a triple \left(S^{1},\beta,(t_{1},\dots,t_{k})\right) , with (t_{1},\dots,t_{k}) denoting points on the circle corresponding to the point vortices on the loop. The closed 2-form \Omega^{\Gamma}=\Omega+\sum_{i=1}^{k}\Gamma_{i}\operatorname{ev}_{t_{i}}^{*}\omega where \operatorname{ev}_{t_{i}}:\operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2})\to\mathbb{R}^{2} is the evaluation map at the points t_{i} of the circle is a symplectic form on \operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2}) . If the vorticities (\Gamma_{i}) and (\textsf{w}_{i}) have no joint rotational symmetry, the coadjoint orbit is identified with \operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2}) and \Omega^{\Gamma} is its KKS symplectic form. If the vorticities (\Gamma_{i}) and (\textsf{w}_{i}) have joint rotational symmetry of step \ell , the coadjoint orbit \mathcal{O}_{\bar{\textsf{w}}}^{a} is identified with the quotient \operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2})/\mathbb{Z}_{k/\ell} , with KKS symplectic form descending from \Omega^{\Gamma} . We remark thus that the manifold \mathcal{O}_{\bar{\textsf{w}}}^{a} of vortex loops enclosing a fixed area and featuring partial vorticites can be realized as two different coadjoint orbits of \operatorname{Ham}_{c}(\mathbb{R}^{2}) , via identification with the quotient manifold \operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2})/\mathbb{Z}_{k/\ell} . If the partial vorticities arise from the existence of point vortices attached to the loop, the KKS symplectic form descends from \Omega^{\Gamma} . If the partial voticities arise because the vorticity density has isolated zeros, the KKS symplectic form descents from \Omega . In other words, the quotient manifold \operatorname{Emb}_{a}(S^{1},\mathbb{R}^{2})/\mathbb{Z}_{k/\ell} parameterizes two different coadjoint orbits of \operatorname{Ham}_{c}(\mathbb{R}^{2}) consisting of different types of vortex loops, with two different KKS symplectic forms. 2 Manifolds of vortex loops We consider the pair (S^{1},\beta) , where S^{1} denotes the oriented unit circle and \beta\in\Omega^{1}(S^{1}) is a one-form with nondegenerate zeros (i.e a Morse form). We denote by Z(\beta) its zero set. As a consequence of the Morse lemma for a compact space, the zeros of \beta are isolated and finitely many. By the Poincaré-Hopf lemma we deduce that the number of zeros of \beta must be even. Identifying S^{1} with the quotient \mathbb{R}/2\pi\mathbb{Z} we obtain a natural ordering of the zeros: we may regard them as a k -tuple \bar{t}=(t_{1},\dots,t_{k}) , with 0\leq t_{1}\leq\dots\leq t_{k}<2\pi . We make the following notations: \textsf{w}:=\int_{S^{1}}\beta,\quad\textsf{w}_{i}:=\int_{t_{i}}^{t_{i+1}}\beta, with the convention t_{k+1}=t_{1} . The \textsf{w}_{i} ’s form a k -tuple denoted by \bar{\textsf{w}}:=(\textsf{w}_{1},\dots,\textsf{w}_{k}) , which inherits an ordering from \bar{t} . We remark that, because of the nondegeneracy of the zeros, the signs of the \textsf{w}_{i} ’s alternate. We consider vortex loops based on this model, namely pairs (C,\beta_{C}) consisting of an oriented closed plane curve C , with no self intersections, endowed with a Morse vorticity density \beta_{C}\in\Omega^{1}(C) (i.e. a one-form with nondegenerate zeros). We denote its zero set by Z(\beta_{C}) . Choosing a “starting point” x_{1}\in Z(\beta_{C}) allows us to denote the zeros of \beta_{C} as a k -tuple \bar{x}:=(x_{1},\dots,x_{k}) ordered with respect to the orientation of the curve. We denote the total vorticity by \textsf{w}:=\int_{C}\beta_{C} and the partial vorticities induced by the zeros by \textsf{w}_{i}:=\int_{x_{i}}^{x_{i+1}}\beta_{C} , with the convention that x_{k+1}=x_{1} . Let f\in\operatorname{Emb}(S^{1},\mathbb{R}^{2}) be an embedding and \beta\in\Omega^{1}(S^{1}) a Morse form with \bar{t}=(t_{1},\dots,t_{k}) and \bar{\textsf{w}}=(\textsf{w}_{1},\dots,\textsf{w}_{k}) . The diffeomorphism f:S^{1}\to f(S^{1}) induces a Morse form f_{*}\beta on its image, with the same number of zeros as \beta and the same set

Read the complete text