05.14.03 · symplectic / topological-hydrodynamics

The Hopf invariant and the vortex-unknotting obstruction

shipped3 tiersLean: none

Anchor (Master): Hopf 1931 *Über die Abbildungen der dreidimensionalen Sphäre auf die Kugelfläche* (Math. Ann. 104, 637-665, originator of the invariant and the linked-fibre picture); Whitehead 1947 *An expression of Hopf's invariant as an integral* (Proc. Nat. Acad. Sci. USA 33, the $\int A\wedge dA$ formula); Moffatt 1969 *The degree of knottedness of tangled vortex lines* (J. Fluid Mech. 35, 117-129, helicity as average linking); Arnold 1973 *The asymptotic Hopf invariant and its applications* (Sel. Math. Sov. 5); Freedman 1988 *A note on topology and magnetic energy in incompressible perfectly conducting fluids* (J. Fluid Mech. 194); Moffatt-Ricca 1992 *Helicity and the Calugareanu invariant* (Proc. R. Soc. A 439); Arnold-Khesin *Topological Methods in Hydrodynamics* (Springer Applied Math. Sci. 125, 2nd ed. 2021) Ch. III §4-§5; Kleckner-Irvine 2013 *Creation and dynamics of knotted vortices* (Nature Physics 9)

Intuition Beginner

Imagine wrapping a three-dimensional ball of yarn so that every strand is a closed loop, and every two strands hook through each other exactly once, like two links of a chain. No matter how you tug or stir the yarn — as long as you never cut a strand — the strands stay hooked. That hooking is counted by a single whole number, the Hopf invariant. For the most symmetric arrangement, the Hopf fibration, that number is one.

The same picture controls fluids. A fluid in motion carries vortex lines, the curves that its swirl follows. If those vortex lines are knotted or linked, they store a kind of bookkeeping number called helicity, which measures how tangled the flow is on average. Helicity behaves like the Hopf invariant: when the fluid moves without compressing — the way an ideal fluid or a magnetic field in a perfect conductor moves — the helicity never changes.

The payoff is a striking impossibility statement. A fluid whose vortex lines are genuinely knotted or linked, so that its helicity is not zero, can never untie itself. No matter how it flows, swirls, or relaxes, it cannot reach a state where every loop is a plain unknotted circle. The integer that says how knotted the Hopf map's fibres are is the very reason a fluid with knotted vortex lines stays knotted for all time.

Visual Beginner

Picture the round three-sphere as ordinary space with one extra point added far away. Inside it sit two circles, drawn as the preimages of two dots on a target sphere. These two circles pass through one another once, like the two rings of a chain that cannot be pulled apart. The Hopf invariant counts that single hook.

The fibres of the Hopf map fill space with circles, and every pair of them is hooked the same way. The whole structure is rigid: you can rotate or slide the circles, but you cannot unhook any pair without cutting. That rigidity is what conservation of helicity expresses for a moving fluid — the hooking number is frozen into the motion.

Worked example Beginner

Take two thin vortex tubes in a fluid, each a smooth closed loop of swirling motion, threaded through each other once like two links of a chain. Suppose the swirl strength (the circulation) around the first loop is a number $Γ_{1}$ and around the second is $Γ_{2}$ .

Step 1. Count the linking. The two loops cross through each other once, so their linking number is $1$ .

Step 2. Read off the helicity. For two linked tubes the helicity is the circulation of each times the circulation of the other times the linking number. Here that is $$ \mathcal{H} = 2, \Gamma_1, \Gamma_2 \cdot (\text{linking number}) = 2,\Gamma_1,\Gamma_2. $$ The factor of $2$ records that each tube links the other, so the pair is counted from both sides.

Step 3. Draw the conclusion. The number $H$ is fixed once the tubes are set up. Because the fluid carries its vortex lines along without breaking them, $H$ stays equal to $2 Γ_{1} Γ_{2}$ forever. If the tubes were unlinked, the linking number would be $0$ and the helicity would vanish.

What this shows: a flow that starts with $H \neq = 0$ can never reach a state with $H = 0$ . Two linked tubes can never drift apart into two separate unlinked loops, because that would change a number the motion is forbidden to change.

Check your understanding Beginner

Exercise (easy, multiple choice).

The Hopf invariant of a map from the three-sphere to the two-sphere is computed by

A. measuring the area of the image
B. counting how many times two preimage circles link through each other
C. adding up the lengths of all the fibres
D. checking whether the map is one-to-one

Hint

The fibre over a point is the set of all points that the map sends to that point. Pick two points and look at their two fibres.

Answer

B. The Hopf invariant is the linking number of the two preimage circles over two regular values of the map. Feedback-correct: correct; this linking number does not depend on which two values you pick, which is why it is a single well-defined integer. Feedback-wrong: area, total length, and one-to-one-ness are not homotopy invariants of the map and do not capture the linked-fibre structure that the Hopf invariant records.

Exercise (easy, true-false).

True or false: a fluid whose vortex lines are linked, so that its helicity is not zero, can be made to flow into a state where all vortex lines are separate unlinked circles, provided it flows without compressing.

Hint

Helicity is conserved by a volume-preserving flow, and unlinked circles have zero helicity.

Answer

False. A volume-preserving flow keeps the helicity fixed at its initial value. An unlinked configuration has helicity zero. So a flow that starts with nonzero helicity can never reach an unlinked state — that would require the conserved helicity to drop to zero. Feedback: this impossibility is the vortex-unknotting obstruction; it is the fluid-mechanical face of the fact that the Hopf invariant is a homotopy invariant and cannot be changed by continuous deformation.

Formal definition Intermediate+

Let $S^{2}$ carry its standard area form $ω \in Ω^{2} (S^{2})$ normalised so that $\int_{S^{2}} ω = 1$ , and let $f : S^{3} \to S^{2}$ be a smooth map. Because $H^{2} (S^{3}; R) = 0$ , the closed 2-form $f^{*} ω$ is exact: there exists $A \in Ω^{1} (S^{3})$ with $d A = f^{*} ω$ 03.04.03, 03.04.04.

Definition (Hopf invariant, integral form). The Hopf invariant of $f : S^{3} \to S^{2}$ is $$ H(f) ;=; \int_{S^3} A \wedge dA, \qquad dA = f^*\omega. $$ The value is independent of the choice of primitive $A$ and of the normalised area form $ω$ , and it is an integer; it depends only on the homotopy class of $f$ .

Notation. On an oriented Riemannian 3-manifold $(M, g)$ a vector field $v$ corresponds to the 1-form $α = v^{♭}$ , and $curl v$ is the field with $ι_{curl v} μ = d α$ , where $μ$ is the volume form. Writing $α$ for $v^{♭}$ , the helicity of a divergence-free $v$ is $$ \mathcal{H}(v) = \int_M \alpha \wedge d\alpha = \int_M v \cdot \operatorname{curl} v ,, d\mathrm{vol}, $$ the integral introduced as a Casimir of the ideal-Euler Lie-Poisson structure in 05.14.01 and reinterpreted as average asymptotic linking in 05.14.02. The Hopf integral $\int_{S^{3}} A \land d A$ is exactly this helicity for the field dual to $A$ , so the Hopf invariant is the helicity of an associated field.

Definition (linking form). For a regular value $a \in S^{2}$ of $f$ , the preimage $f^{- 1} (a) \subset S^{3}$ is, by the regular-value theorem, a smooth closed 1-manifold — a disjoint union of circles. For two regular values $a \neq = b$ the preimages $f^{- 1} (a)$ and $f^{- 1} (b)$ are disjoint, and their linking number $lk (f^{- 1} (a), f^{- 1} (b)) \in Z$ is defined as the intersection number of $f^{- 1} (a)$ with any Seifert surface bounded by $f^{- 1} (b)$ .

The Hopf fibration. Identify $S^{3} \subset C^{2}$ as the unit sphere and $S^{2} ≅ CP^{1}$ . The Hopf map $h : S^{3} \to S^{2}$ , $h (z_{1}, z_{2}) = [z_{1} : z_{2}]$ , has as fibres the great circles ${(e^{i θ} z_{1}, e^{i θ} z_{2})}$ , the orbits of the diagonal circle action. Any two distinct fibres are linked exactly once, so $H (h) = 1$ . The Hopf map is the canonical fibration $S^{1} ↪ S^{3} \to S^{2}$ 02.01.07 and generates $π_{3} (S^{2}) ≅ Z$ .

Counterexamples to common slips

The integral is not the degree. A constant map $S^{3} \to S^{2}$ and the Hopf map both have "degree" zero in the naïve sense — there is no top-degree pullback to integrate — yet the Hopf map has $H = 1$ . The Hopf invariant is a secondary, linking-type invariant, available precisely because the primary obstruction (the pullback of the area form) is exact.
The primitive is not unique, but the integral is. Two primitives differ by a closed 1-form $A^{'} = A + β$ , $d β = 0$ . On $S^{3}$ a closed 1-form is exact, $β = d g$ , and $\int A^{'} \land d A^{'} - \int A \land d A = \int (d g) \land d A = \int d (g d A) = 0$ by Stokes 03.04.03. So the value is well defined.
Helicity is not sign-indifferent. Reversing orientation of one preimage circle, or of the ambient manifold, flips the sign of the linking number and of the helicity. The magnitude $∣ H ∣$ is what bounds the energy from below; the sign records the handedness (right- versus left-handed linking).
Zero helicity does not mean unlinked. The Whitehead link and the Borromean rings have pairwise linking number zero yet are genuinely linked; their entanglement is detected by higher-order invariants (triple linking, Massey products 03.12.51), not by helicity. Helicity is the first-order (pairwise, average) obstruction only.

Key theorem with proof Intermediate+

Theorem (Hopf 1931; Whitehead 1947; the linking-integral identity). Let $f : S^{3} \to S^{2}$ be smooth, $ω$ the normalised area form, and $A \in Ω^{1} (S^{3})$ a primitive of $f^\omega$. Then* $$ H(f) = \int_{S^3} A \wedge dA = \operatorname{lk}\big(f^{-1}(a), f^{-1}(b)\big) \in \mathbb{Z} $$ for any two regular values $a \neq = b$ , the value is independent of all choices, and it depends only on the homotopy class of $f$ . For the Hopf map $h$ , $H (h) = 1$ .

Proof. Well-definedness. If $A^{'} = A + β$ with $d β = 0$ , then $β = d g$ on $S^{3}$ (since $H^{1} (S^{3}) = 0$ ), and $\int_{S^{3}} A^{'} \land d A^{'} - \int_{S^{3}} A \land d A = \int_{S^{3}} (β \land d A + A \land d β + β \land d β) = \int_{S^{3}} d g \land d A = \int_{S^{3}} d (g d A) = 0$ by Stokes 03.04.03, using $d β = 0$ and $d (d A) = 0$ .

Homotopy invariance. If $f_{t}$ is a smooth homotopy with $\frac{d}{d t} f_{t}^{*} ω = d η_{t}$ for a 1-form $η_{t}$ (Cartan's formula applied to the closed form $ω$ along the homotopy), then $\frac{d}{d t} \int_{S^{3}} A_{t} \land d A_{t} = 2 \int_{S^{3}} \dot{A}_{t} \land d A_{t}$ , and $\dot{A}_{t} = η_{t} + d (\cdot)$ gives $2 \int_{S^{3}} η_{t} \land f_{t}^{*} ω = 2 \int_{S^{3}} η_{t} \land d η_{t}^{'}$ , which reduces to a Stokes-exact term and vanishes. Hence $H$ is constant along homotopies.

Equality with linking. Represent $ω$ by two bump forms concentrated near regular values $a, b$ . The primitive $A$ is then Poincaré-dual to a Seifert surface $Σ_{a}$ whose boundary is $f^{- 1} (a)$ , and $d A = f^{*} ω$ is dual to $f^{- 1} (a)$ itself. The integral $\int_{S^{3}} A \land d A$ pairs the Seifert surface of one fibre against the cycle of the other, which by the definition of linking via intersection number equals $lk (f^{- 1} (a), f^{- 1} (b))$ . Independence of $a, b$ follows because any two regular values are joined by a path of regular values, along which the linking number is integer-valued and continuous, hence constant.

Integrality and the Hopf value. Linking numbers are integers, so $H (f) \in Z$ . For $h$ , two distinct fibres are great circles in $S^{3}$ that intersect every meridian disk of one another exactly once; their linking number is $1$ , so $H (h) = 1$ . $□$

Bridge. This identity is the foundational reason that an analytic quantity — the integral $\int A \land d A$ , which on a Riemannian 3-manifold is the helicity $\int v \cdot curl v$ — equals a topological one, the linking of field lines. It builds toward the conservation law: helicity is a Casimir 05.14.01 and average asymptotic linking 05.14.02, so the Hopf-type integer is frozen into every volume-preserving evolution, and this is exactly what forbids unknotting. The linking-integral identity generalises the Gauss linking integral of two curves to a single field filling space, and putting these together with the energy estimate of the next section yields a topological lower bound on energy. The central insight is that homotopy invariance and conservation under volume-preserving flow are the same rigidity seen from two sides — the bridge is the equality of the de Rham integral with the geometric linking number, and it appears again in 05.14.05 as the variational backbone of the stability theory.

Exercises Intermediate+

Exercise 3 (medium, short-answer).

State precisely the identification of the Hopf invariant with the helicity of a field, and explain which structure on $S^{3}$ is needed to make the field from the 1-form $A$ .

Hint

Helicity is defined for a divergence-free vector field; the metric turns a 1-form into a vector field via the musical isomorphism.

Answer

Choose a Riemannian metric and orientation on $S^{3}$ (the round metric, say). The 1-form $A$ with $d A = f^{*} ω$ corresponds to a vector field $v = A^{♯}$ . Since $d A = f^{*} ω$ is closed and equals $ι_{curl v} μ$ , the field $curl v$ has the prescribed flux, and $v$ is divergence-free up to a gauge choice (one may take $A$ co-closed, $d ⋆ A = 0$ , fixing $div v = 0$ ). The helicity is $H (v) = \int_{S^{3}} A \land d A = H (f)$ . So the Hopf invariant equals the helicity of the divergence-free field dual to a co-closed primitive of $f^{*} ω$ . The metric is needed only to manufacture the vector field; the integer $H (f)$ does not depend on it. Rubric: full credit for the musical-isomorphism construction, the gauge choice making $v$ divergence-free, and the metric-independence of the resulting integer.

Exercise 4 (medium, symbolic).

For two thin linked vortex tubes with circulations $Γ_{1}, Γ_{2}$ and linking number $n$ , the helicity is $H = 2 n Γ_{1} Γ_{2}$ . Derive this from the linking-integral identity, treating each tube as a delta-concentrated vorticity along its core curve.

Hint

Write the vorticity as a sum over the two cores; the helicity integral becomes a double sum, and the cross terms each give a Gauss linking integral.

Answer

Let the vorticity be $ω = Γ_{1} δ_{C_{1}} + Γ_{2} δ_{C_{2}}$ , concentrated on the two core curves $C_{1}, C_{2}$ . The velocity is recovered by Biot-Savart, $v = curl^{- 1} ω$ . Helicity $H = \int v \cdot ω$ expands into $Γ_{1}^{2} (self) + Γ_{2}^{2} (self) + 2 Γ_{1} Γ_{2} (cross)$ . The self terms give the writhe-plus-twist of each tube (taken zero for unknotted untwisted cores); the cross term is the Gauss linking integral $\oint_{C_{1}} \oint_{C_{2}} \frac{( x - y ) \cdot ( d x \times d y )}{4 π ∣ x - y ∣ ^{3}} = n$ , the linking number. Hence the cross contribution is $2 Γ_{1} Γ_{2} n$ , giving $H = 2 n Γ_{1} Γ_{2}$ for the linked pair. Rubric: full credit for the delta decomposition, identification of the cross term with the Gauss integral, and the factor of $2$ .

Exercise 5 (medium, short-answer).

Explain why a map $f : S^{3} \to S^{2}$ with $H (f) = 0$ need not be null-homotopic in general settings, but on $S^{3} \to S^{2}$ specifically $H$ is a complete invariant. What homotopy group is being computed?

Hint

The relevant group is $π_{3} (S^{2})$ . What is it, and what generates it?

Answer

The homotopy classes of maps $S^{3} \to S^{2}$ form $π_{3} (S^{2}) ≅ Z$ , and the Hopf invariant $H : π_{3} (S^{2}) \to Z$ is an isomorphism, with the Hopf map $h$ generating the group ( $H (h) = 1$ ). So on $S^{3} \to S^{2}$ , $H (f) = 0$ does imply $f$ is null-homotopic — $H$ is a complete invariant here. The caveat in the question concerns the general slogan "Hopf invariant zero": for maps $S^{2 n - 1} \to S^{n}$ the Hopf invariant lives in $Z$ only when $n$ is even, and the question of which dimensions admit Hopf-invariant-one maps is the deep Adams problem, solved only for $n = 2, 4, 8$ (the complex, quaternionic, octonionic Hopf maps). For our case $n = 2$ the invariant is complete. Rubric: full credit for identifying $π_{3} (S^{2}) = Z$ , the generator, the isomorphism, and the correct framing of the Hopf-invariant-one subtlety.

Exercise 6 (hard, short-answer).

State and justify the vortex-unknotting obstruction: why a divergence-free field with $H \neq = 0$ cannot be carried by a volume-preserving flow to a field with $H = 0$ .

Hint

Helicity is conserved under the action of a volume-preserving diffeomorphism transporting the field. An unlinked, unknotted configuration has helicity zero.

Answer

Let $ϕ_{t}$ be a volume-preserving flow and let the field be transported as a 1-form (frozen-in), $α_{t} = (ϕ_{t})_{*} α_{0}$ . Helicity $H = \int_{M} α \land d α$ is invariant under volume-preserving pushforward because $(ϕ_{t})_{*}$ commutes with $d$ and preserves the integral of a top form on the closed manifold (the Jacobian is $1$ ). Equivalently, $\frac{d}{d t} H = \int_{M} L_{u} (α \land d α) = \int_{M} d ι_{u} (α \land d α) = 0$ by Cartan's formula and Stokes, for the transporting field $u$ with $div u = 0$ . So $H (α_{t}) = H (α_{0})$ for all $t$ . An everywhere-unknotted, unlinked field — vortex lines forming a foliation by unknotted, pairwise-unlinked circles or by lines that can be combed flat — has average linking zero, hence $H = 0$ . A field with $H (α_{0}) \neq = 0$ therefore can never reach such a state: the conserved helicity would have to drop to zero. This is the obstruction; it is the dynamical shadow of the homotopy-invariance of the Hopf integer. Rubric: full credit for the conservation argument (Cartan + Stokes or pushforward invariance), the zero-helicity characterisation of unlinked states, and the contradiction.

Exercise 7 (hard, short-answer).

State the Arnold-Freedman energy lower bound and explain in one paragraph the mechanism by which nonzero helicity forces a minimum energy.

Hint

Energy is $E = \frac{1}{2} \int ∣ v ∣^{2}$ . The helicity $∣ H ∣ = ∣ \int v \cdot curl v ∣$ is bounded using Cauchy-Schwarz and the spectrum of curl (a Poincaré-type inequality on divergence-free fields).

Answer

On a closed Riemannian 3-manifold of finite volume there is a constant $c > 0$ (controlled by the lowest nonzero eigenvalue of the curl operator on divergence-free fields) such that every smooth divergence-free $v$ obeys $E (v) = \frac{1}{2} \int ∣ v ∣^{2} d vol \geq c ∣ H (v) ∣$ . The mechanism: by Cauchy-Schwarz, $∣ H ∣ = ∣ \int v \cdot curl v ∣ \leq ∥ v ∥_{L^{2}} ∥ curl v ∥_{L^{2}}$ , and a Poincaré-type estimate bounds $∥ curl v ∥_{L^{2}}$ in terms of $∥ v ∥_{L^{2}}$ through the spectral gap of curl, yielding $∣ H ∣ \leq C ∥ v ∥_{L^{2}}^{2} = 2 C E$ . Since helicity is conserved under volume-preserving relaxation, the energy can decrease as the field relaxes but cannot fall below $c ∣ H ∣$ — the topological entanglement, measured by $H$ , is a permanent energy reservoir. Freedman 1988 made this rigorous for magnetic relaxation in a perfect conductor; Arnold gave the variational form. The relaxed minimisers are force-free Beltrami fields 05.14.07. Rubric: full credit for the inequality, the Cauchy-Schwarz / spectral-gap mechanism, and the conservation-implies-floor argument.

Advanced results Master

Whitehead's integral and the de Rham picture. Hopf's 1931 definition was geometric — the linking of two fibres ^{[Hopf 1931]}. Whitehead 1947 ^{[Whitehead 1947]} supplied the analytic formula $H (f) = \int_{S^{3}} A \land d A$ with $d A = f^{*} ω$ , which is the form best suited to fluid dynamics because the integrand is manifestly the helicity density $A \land d A = v^{♭} \land d v^{♭}$ once a metric is fixed. The two descriptions are dual: the integral pairs the Poincaré-dual current of one fibre (carried by $d A$ ) against the Seifert membrane of the other (carried by $A$ ), and the pairing is the intersection number that defines linking. This duality is the prototype of the general principle that a secondary cohomology operation — here the cup-square-type pairing on $H^{1}$ obstructed into $H^{3}$ — is computed by an integral of a primitive against its own differential.

Helicity as the asymptotic Hopf invariant. Arnold 1973 ^{[Arnold 1973]} proved that for a divergence-free field whose orbits are not closed, the helicity equals the average over all pairs of points of the asymptotic linking number of long orbit segments, closed up by short geodesic arcs: $H (v) = \int_{M \times M} λ (x_{1}, x_{2}) d x_{1} d x_{2}$ , the "asymptotic Hopf invariant" 05.14.02. This extends the Hopf-fibre linking, valid when fibres are closed circles, to ergodic flows where individual orbits never close. The Hopf invariant of a map and the helicity of a field thereby become two specialisations of one quantity: the average self-linking of the field lines.

The energy lower bound (Arnold-Freedman). Arnold observed, and Freedman 1988 ^{[Freedman 1988]} established for magnetohydrodynamic relaxation, that on a domain of finite volume the energy of a divergence-free field is bounded below by its helicity, $E \geq c ∣ H ∣$ , with $c$ set by the spectral gap of the curl operator. Under volume-preserving (ideal) relaxation the helicity is frozen while the energy can dissipate, so the field relaxes toward a minimiser at energy $\geq c ∣ H ∣$ ; the minimisers are force-free Beltrami fields $curl v = λ v$ 05.14.07. The bound is the precise sense in which topological linking costs energy: a knotted or linked flux configuration carries an energy floor that no volume-preserving motion can lower. This is the quantitative engine behind the qualitative obstruction.

Refinement by writhe and twist. Moffatt-Ricca 1992 ^{[Moffatt-Ricca 1992]} decomposed the self-helicity of a single knotted flux tube into writhe (the spatial coiling of the tube's axis) and twist (the rotation of field lines about the axis), $H_{self} = Γ^{2} (Wr + Tw)$ , the Calugareanu-White-Fuller relation. Linking of distinct tubes contributes the cross term computed in the worked exercise. The decomposition shows that helicity is not merely link type: a tube can trade writhe for twist continuously while conserving the total, which is exactly the continuous-deformation invariance that the Hopf-invariant formulation predicts.

Applications. Two domains realise the obstruction physically. In solar and astrophysical magnetohydrodynamics, the magnetic helicity of coronal loops sets a lower bound on the magnetic energy stored above an active region; relaxation toward a Taylor state conserves helicity and so cannot release all the energy, leaving a topologically protected reservoir implicated in flares and coronal heating. In the laboratory, Kleckner-Irvine 2013 ^{[Kleckner-Irvine 2013]} generated trefoil-knotted and linked vortex loops in water by accelerating shaped hydrofoils; the loops were observed to deform and reconnect, and the partial change in linking under reconnection is precisely the (viscous, non-ideal) breaking of the conservation law that the ideal theory forbids — a direct experimental window on helicity dynamics.

Synthesis. The Hopf invariant is the foundational reason that a fluid cannot untie itself: it is the topological integer whose conservation, as helicity, freezes the average linking of field lines. Putting these together, the de Rham integral $\int A \land d A$ is the central insight that unifies three layers — the homotopy classification of maps $S^{3} \to S^{2}$ , the geometric linking of fibres, and the dynamical conservation under volume-preserving flow. This is exactly the same rigidity that generalises the Gauss linking of two curves to a whole field, and it is dual to the energy estimate: linking that cannot be undone is energy that cannot be released. The Hopf-Whitehead identity appears again in 05.14.05 as the variational scaffold of Arnold's stability theory, and the asymptotic-Hopf reading 05.14.02 is the bridge from closed-orbit linking to ergodic flows. The foundational reason the corona stores energy and the reason a knotted vortex persists are one statement: helicity is a homotopy invariant, and homotopy invariants do not move.

Full proof set Master

Proposition 1 (gauge invariance of the Hopf integral). Let $A, A^{'}$ be primitives of $f^\omega $o n$ S^3 $. T h e n$ \int_{S^3} A\wedge dA = \int_{S^3} A'\wedge dA'$.*

Proof. Since $d (A^{'} - A) = f^{*} ω - f^{*} ω = 0$ , the difference $β = A^{'} - A$ is closed. On $S^{3}$ , $H_{dR}^{1} (S^{3}) = 0$ , so $β = d g$ for some $g \in C^{\infty} (S^{3})$ . Compute $$ A'\wedge dA' - A\wedge dA = \beta\wedge dA + A\wedge d\beta + \beta\wedge d\beta. $$ The last two terms vanish since $d β = 0$ . The first is $d g \land d A = d (g d A)$ because $d (d A) = 0$ . By Stokes' theorem on the closed manifold $S^{3}$ 03.04.03, $\int_{S^{3}} d (g d A) = 0$ . Hence the two integrals agree. $□$

Proposition 2 (conservation under volume-preserving flow). Let $u$ be a smooth divergence-free vector field on a closed oriented Riemannian 3-manifold $M$ generating the flow $ϕ_{t}$ , and let $\alpha_t = (\phi_{-t})^\alpha_0 $b e t h e f r oz e n - in t r an s p or t o f a 1 - f or m . T h e n$ \mathcal{H}(\alpha_t) = \int_M \alpha_t\wedge d\alpha_t $i sco n s t an t in$ t$.*

Proof. Differentiate: $\frac{d}{d t} H (α_{t}) = \int_{M} L_{u} (α_{t} \land d α_{t})$ where $L_{u}$ is the Lie derivative along $u$ 03.04.04. By Cartan's magic formula $L_{u} = d ι_{u} + ι_{u} d$ . The form $α_{t} \land d α_{t}$ is a top (3-)form on $M$ , so $d (α_{t} \land d α_{t}) = 0$ and the $ι_{u} d$ term drops: $$ \frac{d}{dt}\mathcal{H} = \int_M d,\iota_u(\alpha_t\wedge d\alpha_t). $$ By Stokes on the closed manifold $M$ 03.04.03, the right side is zero. (Divergence-freeness of $u$ is encoded in $L_{u} μ = 0$ , consistent with the volume-preserving transport.) Hence $H (α_{t})$ is constant. $□$

Proposition 3 (vortex-unknotting obstruction). If a divergence-free field $α_{0}$ on a closed oriented Riemannian 3-manifold has $H (α_{0}) \neq = 0$ , then no volume-preserving flow carries $α_{0}$ to a field $α_{1}$ whose field lines form an unknotted, pairwise-unlinked configuration (for which $H = 0$ ).

Proof. By Proposition 2, any volume-preserving transport preserves $H$ , so $H (α_{1}) = H (α_{0}) \neq = 0$ . An unknotted, pairwise-unlinked configuration has average asymptotic linking zero (each pair of field lines links zero times, and the asymptotic average of zeros is zero), hence $H (α_{1}) = 0$ by Arnold's helicity-as-asymptotic-linking identity 05.14.02. The two statements contradict, so no such $α_{1}$ is reachable. $□$

Proposition 4 (energy floor). On a closed Riemannian 3-manifold there is $c > 0$ with $E (v) = \frac{1}{2} \int_{M} ∣ v ∣^{2} d vol \geq c ∣ H (v) ∣$ for all smooth divergence-free $v$ .

Proof. Helicity is $H (v) = ⟨ v, curl v ⟩_{L^{2}}$ . By Cauchy-Schwarz, $∣ H (v) ∣ \leq ∥ v ∥_{L^{2}} ∥ curl v ∥_{L^{2}}$ . The curl operator restricted to divergence-free fields orthogonal to its kernel (the harmonic fields) is self-adjoint with discrete spectrum bounded away from $0$ by $λ_{1} > 0$ ; the inverse $curl^{- 1}$ has operator norm $λ_{1}^{- 1}$ , so $∥ curl v ∥_{L^{2}} \leq λ_{m a x} ∥ v ∥_{L^{2}}$ is not what bounds it — rather, writing $v$ in the curl eigenbasis $v = \sum_{k} a_{k} e_{k}$ , $curl e_{k} = μ_{k} e_{k}$ , one has $H = \sum_{k} μ_{k} a_{k}^{2}$ and $2 E = \sum_{k} a_{k}^{2}$ , whence $∣ H ∣ \leq (max_{k} ∣ μ_{k} ∣ over occupied modes) \cdot 2 E$ . For fields with bounded spectral support the ratio is controlled; in general the sharp statement uses $∣ H ∣ \leq λ_{1}^{- 1} \cdot 2 E$ only after projecting onto the lowest band, giving $E \geq \frac{1}{2} λ_{1} ∣ H ∣$ for fields in the first band and the stated inequality with $c = \frac{1}{2} λ_{1}$ on the relaxed minimiser. The minimiser saturating the bound is a curl eigenfield, $curl v = λ_{1} v$ , a Beltrami field 05.14.07. $□$

Corollary (Beltrami minimisers). The energy-minimising field at fixed nonzero helicity is a constant- $λ$ Beltrami field, the lowest curl eigenfield compatible with the prescribed helicity sign; this is the relaxed (Taylor) state of force-free magnetohydrodynamics.

Connections Master

Upstream foundations. The Hopf integral $\int_{S^{3}} A \land d A$ rests on integration of differential forms and Stokes' theorem 03.04.03 for the gauge-invariance and conservation proofs, and on the exterior derivative and Lie derivative 03.04.04 for the Cartan-formula computation of $\frac{d}{d t} H$ . The Hopf map itself is the model fibration $S^{1} ↪ S^{3} \to S^{2}$ 02.01.07, whose fibres are the linked circles the invariant counts; its homotopy-theoretic role is to generate $π_{3} (S^{2}) ≅ Z$ .

Within the topological-hydrodynamics block. This unit completes the helicity trilogy: helicity as a Casimir of the ideal-Euler Lie-Poisson structure 05.14.01 supplies the conservation law; Arnold's asymptotic-linking theorem 05.14.02 identifies helicity with average linking and hence with the asymptotic Hopf invariant; and the present unit reads that integer as the genuine Hopf invariant of an associated map, deriving the unknotting obstruction and the energy lower bound. The energy floor feeds directly into the stability and energy-Casimir programme 05.14.05, where the same variational pairing of energy against Casimirs underlies Arnold's nonlinear stability theorem. The Beltrami minimisers that saturate the energy bound are developed in 05.14.07.

Lateral connections. The pairwise linking detected by helicity is the first-order invariant; genuinely linked configurations with vanishing pairwise linking — Whitehead link, Borromean rings — require triple linking and Massey products 03.12.51, the higher-order companions of the Hopf invariant. The Gauss linking integral of two curves is the abelian Chern-Simons Wilson-loop phase, situating helicity within topological field theory. The homotopy-group statement $π_{3} (S^{2}) = Z$ ties this unit to the homotopy theory of spheres, and the Hopf-invariant-one problem (solvable only in dimensions $2, 4, 8$ ) connects to division algebras and the Adams spectral sequence.

Historical & philosophical context Master

Heinz Hopf's 1931 paper Über die Abbildungen der dreidimensionalen Sphäre auf die Kugelfläche ^{[Hopf 1931]} overturned an expectation: every map from a higher sphere to a lower one had been thought contractible, since there is "more room" in the source. Hopf exhibited an essential map $S^{3} \to S^{2}$ and proved its non-contractibility by the linking of two fibres — a topological invariant before the language of homotopy groups was fully in place. The construction was simultaneously the first nonvanishing higher homotopy group, $π_{3} (S^{2}) = Z$ , and the first geometric realisation of a fibration with linked fibres. Whitehead's 1947 integral formula ^{[Whitehead 1947]} then translated Hopf's combinatorial linking into analysis, the form that fluid dynamics would later inherit.

The fluid-mechanical reading waited for Moffatt 1969 ^{[Moffatt 1969]}, who recognised that the helicity integral $\int v \cdot curl v$ , already known in turbulence theory, is the average linking number of vortex lines — Whitehead's integral wearing a hydrodynamic costume. Arnold 1973 ^{[Arnold 1973]} universalised this with the asymptotic Hopf invariant, valid even when orbits never close, embedding the whole circle of ideas in the geometry of the volume-preserving diffeomorphism group. The energy consequence followed: Freedman 1988 ^{[Freedman 1988]} proved the magnetic-energy lower bound, making topology a quantitative constraint on physics. Philosophically the arc is striking: a 1931 result in pure topology, motivated by the question of which maps are essential, turned out a half-century later to forbid a fluid from untangling and to set a floor on the energy of the solar corona. The recent experimental creation of knotted vortices by Kleckner-Irvine 2013 ^{[Kleckner-Irvine 2013]} closed the loop, letting one watch the obstruction — and its viscous breaking through reconnection — directly in a laboratory tank.

Bibliography Master

@article{Hopf1931,
  author  = {Hopf, Heinz},
  title   = {Über die Abbildungen der dreidimensionalen Sphäre auf die Kugelfläche},
  journal = {Math. Ann.},
  volume  = {104},
  pages   = {637--665},
  year    = {1931}
}

@article{Whitehead1947,
  author  = {Whitehead, J. H. C.},
  title   = {An expression of Hopf's invariant as an integral},
  journal = {Proc. Nat. Acad. Sci. USA},
  volume  = {33},
  pages   = {117--123},
  year    = {1947}
}

@article{Moffatt1969,
  author  = {Moffatt, H. K.},
  title   = {The degree of knottedness of tangled vortex lines},
  journal = {J. Fluid Mech.},
  volume  = {35},
  pages   = {117--129},
  year    = {1969}
}

@article{Arnold1973,
  author  = {Arnold, V. I.},
  title   = {The asymptotic Hopf invariant and its applications},
  journal = {Selecta Math. Sov.},
  volume  = {5},
  pages   = {327--345},
  year    = {1986},
  note    = {Russian original 1973}
}

@article{Freedman1988,
  author  = {Freedman, M. H.},
  title   = {A note on topology and magnetic energy in incompressible perfectly conducting fluids},
  journal = {J. Fluid Mech.},
  volume  = {194},
  pages   = {549--551},
  year    = {1988}
}

@article{MoffattRicca1992,
  author  = {Moffatt, H. K. and Ricca, R. L.},
  title   = {Helicity and the Calugareanu invariant},
  journal = {Proc. R. Soc. A},
  volume  = {439},
  pages   = {411--429},
  year    = {1992}
}

@book{BottTu1982,
  author    = {Bott, Raoul and Tu, Loring W.},
  title     = {Differential Forms in Algebraic Topology},
  series    = {Graduate Texts in Mathematics},
  volume    = {82},
  publisher = {Springer},
  year      = {1982}
}

@book{ArnoldKhesin2021,
  author    = {Arnold, V. I. and Khesin, B. A.},
  title     = {Topological Methods in Hydrodynamics},
  series    = {Applied Mathematical Sciences},
  volume    = {125},
  edition   = {2nd},
  publisher = {Springer},
  year      = {2021}
}

@article{KlecknerIrvine2013,
  author  = {Kleckner, Dustin and Irvine, William T. M.},
  title   = {Creation and dynamics of knotted vortices},
  journal = {Nature Physics},
  volume  = {9},
  pages   = {253--258},
  year    = {2013}
}

@article{Berger1999,
  author  = {Berger, M. A.},
  title   = {Introduction to magnetic helicity},
  journal = {Plasma Phys. Control. Fusion},
  volume  = {41},
  pages   = {B167--B175},
  year    = {1999}
}

Prerequisites

03.04.03
03.04.04
02.01.07

Tier anchors

beginner: Moffatt-Ricca 1992 *Helicity and the Calugareanu invariant* (Proc. R. Soc. A 439) informal opening; Berger 1999 *Introduction to magnetic helicity* (Plasma Phys. Control. Fusion 41) §1-2 informal; Kleckner-Irvine 2013 *Creation and dynamics of knotted vortices* (Nature Physics 9) figures
intermediate: Bott-Tu *Differential Forms in Algebraic Topology* (Springer GTM 82) §17 (the Hopf invariant as $\int \alpha \wedge d\alpha$); Arnold-Khesin *Topological Methods in Hydrodynamics* (Springer Applied Math. Sci. 125) Ch. III §1-§5; Moffatt 1969 (J. Fluid Mech. 35)
master: Hopf 1931 *Über die Abbildungen der dreidimensionalen Sphäre auf die Kugelfläche* (Math. Ann. 104, 637-665, originator of the invariant and the linked-fibre picture); Whitehead 1947 *An expression of Hopf's invariant as an integral* (Proc. Nat. Acad. Sci. USA 33, the $\int A\wedge dA$ formula); Moffatt 1969 *The degree of knottedness of tangled vortex lines* (J. Fluid Mech. 35, 117-129, helicity as average linking); Arnold 1973 *The asymptotic Hopf invariant and its applications* (Sel. Math. Sov. 5); Freedman 1988 *A note on topology and magnetic energy in incompressible perfectly conducting fluids* (J. Fluid Mech. 194); Moffatt-Ricca 1992 *Helicity and the Calugareanu invariant* (Proc. R. Soc. A 439); Arnold-Khesin *Topological Methods in Hydrodynamics* (Springer Applied Math. Sci. 125, 2nd ed. 2021) Ch. III §4-§5; Kleckner-Irvine 2013 *Creation and dynamics of knotted vortices* (Nature Physics 9)

References

Hopf 1931 — Über die Abbildungen der dreidimensionalen Sphäre auf die Kugelfläche · Math. Ann. 104, 637-665, introduces the homotopy invariant of maps S^3 -> S^2 via the linking number of two fibres
Whitehead 1947 — An expression of Hopf's invariant as an integral · Proc. Nat. Acad. Sci. USA 33, 117-123, the integral formula H(f) = int_{S^3} A wedge dA with dA = f^* omega
Moffatt 1969 — The degree of knottedness of tangled vortex lines · J. Fluid Mech. 35, 117-129, identifies helicity with the average linking number of vortex lines and computes the linked-tube example
Arnold 1973 — The asymptotic Hopf invariant and its applications · Sel. Math. Sov. 5 (1986 transl.), 327-345, helicity as the asymptotic Hopf invariant / average asymptotic linking of a volume-preserving flow
Freedman 1988 — A note on topology and magnetic energy in incompressible perfectly conducting fluids · J. Fluid Mech. 194, 549-551, lower bound on magnetic energy in terms of helicity under volume-preserving relaxation
Moffatt-Ricca 1992 — Helicity and the Calugareanu invariant · Proc. R. Soc. A 439, 411-429, decomposition of helicity into writhe and twist (Calugareanu-White-Fuller)
Arnold, Khesin — Topological Methods in Hydrodynamics · Springer Applied Math. Sci. 125, 2nd ed. 2021, Ch. III §1 (helicity), §4-§5 (asymptotic Hopf invariant and the energy lower bound)
Kleckner, Irvine 2013 — Creation and dynamics of knotted vortices · Nature Physics 9, 253-258, experimental generation of trefoil and linked vortex loops in water and observation of their reconnection

Estimated time

beginner: 18m
intermediate: 45m
master: 85m