Papers
Topics
Authors
Recent
Search
2000 character limit reached

Voros Symbols in Exact WKB Analysis

Updated 4 July 2026
  • Voros symbols are formal exponentials of WKB period integrals that encode analytic and geometric data for Schrödinger-type equations.
  • They connect local asymptotic expansions with global structures by incorporating Stokes automorphisms, monodromy, and wall-crossing phenomena.
  • Borel summation transforms these divergent series into analytic functions, enabling precise exact quantization and cluster coordinate applications.

Voros symbols are formal or Borel-resummed exponentials of WKB period integrals associated with a linear differential equation, most commonly a Schrödinger-type equation in exact WKB analysis. They package the global analytic data carried by the odd WKB one-form into multiplicative objects attached to paths and cycles on the relevant spectral cover. In practice, they mediate between local asymptotic expansions and global structures such as Stokes automorphisms, monodromy, wall-crossing, and exact quantization conditions. In the modern literature they also appear as coordinates closely related to spectral-network and cluster-type structures (0907.3987).

1. Definition and basic construction

For a second-order scalar equation of Schrödinger type,

(2d2dz2Q(z,))ψ(z,)=0,\left(\hbar^2 \frac{d^2}{dz^2}-Q(z,\hbar)\right)\psi(z,\hbar)=0,

with

Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,

the standard WKB ansatz is

ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),

where SS solves the Riccati equation

S2+dSdz=Q(z,).S^2+\hbar\,\frac{dS}{dz}=Q(z,\hbar).

One chooses a formal solution with expansion

S(z,)=n1nSn(z),S(z,\hbar)=\sum_{n\ge -1} \hbar^n S_n(z),

and isolates the odd part SoddS_{\mathrm{odd}}, characterized by the change of sign under interchange of the two local branches of Q0\sqrt{Q_0}.

Voros symbols are then obtained by integrating SoddS_{\mathrm{odd}} along appropriate homology classes. For a closed cycle γ\gamma one considers

Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,0

and for an open path Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,1 one uses a regularized integral

Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,2

where the regularization subtracts the singular leading term required near turning points or poles. The multiplicative quantities

Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,3

are the cycle and path Voros symbols.

The logarithms Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,4 and Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,5 are often called Voros coefficients in specific one-parameter situations, especially when a distinguished path is fixed by boundary conditions or by a chosen normalization of WKB solutions. Terminology is not completely uniform across the literature: some authors reserve “Voros coefficient” for the logarithm attached to a particular path, while “Voros symbol” denotes its exponential.

2. Geometric setting

The natural geometric arena is the spectral double cover determined by the principal term Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,6,

Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,7

On Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,8, the leading WKB differential is Q(z,)=Q0(z)+Q1(z)+2Q2(z)+,Q(z,\hbar)=Q_0(z)+\hbar Q_1(z)+\hbar^2 Q_2(z)+\cdots,9, and the odd formal solution may be viewed as its quantum deformation. Paths and cycles relevant for Voros symbols therefore live not directly on the base curve but on the covering curve, with branch points at turning points of the equation.

This homological viewpoint is essential. Closed-cycle symbols encode period data, while open-path symbols encode connection data between distinguished sectors, singularities, or turning points. Their dependence on the choice of sheet, reference paths, and normalization is systematic rather than accidental. A change of basis in relative or absolute homology changes the symbols multiplicatively, so the meaningful content usually lies in their transformation laws and in invariant combinations.

The associated Stokes graph is built from trajectories of the quadratic differential

ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),0

Its topology controls the analytic continuation of WKB solutions. Turning points, simple poles, and saddle trajectories determine how Voros symbols are defined, where their Borel sums exist, and when they jump. In this sense, Voros symbols are not merely formal integrals; they are coordinates adapted to the Stokes geometry of the differential equation.

3. Borel summation and analytic meaning

As formal series, ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),1 and ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),2 are typically divergent. Their analytic content is extracted by Borel summation, when the relevant summability conditions hold. The Borel-summed Voros symbols then become genuine analytic functions of ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),3 in a sector, and they control exact connection formulas among canonical WKB solutions.

This passage from formal to resummed objects is central to exact WKB analysis. The formal series remembers the algebraic structure, while the resummed symbol captures the actual Stokes data. When the Borel direction crosses a Stokes ray associated with a saddle trajectory, the Borel sums can jump. The jump is governed by a Stokes automorphism acting multiplicatively on Voros symbols.

In a common schematic normalization, one writes the jump across a single active cycle ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),4 as

ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),5

and

ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),6

where ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),7 and ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),8 denote the relevant intersection pairings. The precise sign conventions and exponents depend on the setup, but the structural point is stable: Voros symbols transform by birational multiplicative rules governed by homological intersection.

This mechanism explains why Voros symbols are effective intermediaries between local asymptotics and global analytic continuation. They encode the obstruction to naive analytic continuation of WKB solutions and provide a compact language for wall-crossing.

4. Relation to spectral networks, wall-crossing, and cluster structures

The modern significance of Voros symbols extends beyond classical exact WKB analysis. In the spectral-network perspective of Gaiotto, Moore, and Neitzke, WKB trajectories associated with quadratic differentials define coordinate systems on moduli spaces, and the associated wall-crossing behavior is controlled by discontinuous automorphisms closely parallel to the Stokes jumps of Voros symbols (0907.3987).

This suggests a broad identification between exact-WKB period data and multiplicative coordinate systems on moduli spaces of flat connections. In that setting, cycle-type quantities behave like Darboux or Fock–Goncharov-type coordinates, while their discontinuities across walls are governed by intersection-theoretic formulas. A plausible implication is that Voros symbols furnish an exact-WKB realization of the same wall-crossing algebra that appears in spectral networks and cluster-like mutation formalisms.

In the exact WKB literature, this viewpoint is sharpened by interpreting saddle reductions of the Stokes graph as mutation-like operations. Under such a move, the distinguished basis of paths and cycles changes, and the corresponding Voros symbols transform by rules of the same general form as cluster mutations. This does not mean that every occurrence of a Voros symbol defines a cluster variable canonically; rather, it means that in a large and structured class of problems the Stokes geometry organizes the symbols into mutation-compatible coordinate systems.

The wall-crossing aspect is therefore not secondary. It is one of the main reasons Voros symbols remain prominent: they provide explicit analytic representatives of abstract automorphisms that, in other formalisms, are encoded more implicitly.

5. Voros coefficients and exact quantization

In many concrete problems one selects a distinguished path and studies the logarithm of the corresponding symbol,

ψ(z,)exp ⁣(zS(z,)dz),\psi(z,\hbar)\sim \exp\!\left(\int^z S(z,\hbar)\,dz\right),9

where SS0 denotes the chosen Voros symbol. This logarithm is usually called a Voros coefficient. It measures the discrepancy between different canonical normalizations of WKB solutions, and it enters exact connection formulas, monodromy calculations, and quantization conditions.

For one-dimensional quantum problems, exact quantization is often expressed in terms of cycle integrals. At leading order one recovers the Bohr–Sommerfeld period condition; beyond leading order the full resummed cycle symbol supplies the nonperturbative correction. In this role, Voros symbols unify perturbative WKB periods and exponentially small Stokes data. Their logarithms can also be expanded asymptotically, producing explicit series whose coefficients are often expressed through Bernoulli-type structures, special-function data, or topological-recursion invariants, depending on the model.

Voros coefficients are especially important near irregular singularities and turning-point configurations where normalization at infinity or at a pole is nontrivial. There they measure the change between local WKB solutions normalized in different sectors. For special equations, closed expressions or recursive constructions for these coefficients can be derived, and these formulas are frequently the starting point for concrete exact-WKB computations.

A recurrent misconception is that a Voros coefficient is an intrinsic scalar attached to the differential equation alone. In fact, it depends on auxiliary choices: reference normalization, branch selection, path class, subtraction convention in the regularization, and Borel direction. What is intrinsic are the relations among such quantities and the analytic invariants extracted from them.

6. Scope, variants, and interpretive issues

Voros symbols are most familiar for second-order equations, but the underlying idea is broader. Whenever a semiclassical problem yields a meromorphic WKB differential on a covering curve and the associated formal periods are exponentiated, an analogue of the Voros symbol emerges. This suggests natural extensions to higher-rank systems, quantum curves, isomonodromic problems, and settings where exact WKB interacts with topological recursion or with the geometry of flat connections.

Several distinctions are important.

First, formal versus resummed symbols: the formal symbol is an asymptotic object in SS1, while the Borel-summed symbol is an analytic function defined sectorially. Many structural statements are first established formally and then interpreted analytically through summability.

Second, cycle versus path symbols: cycle symbols encode absolute homology and are naturally adapted to monodromy and quantization; path symbols encode relative homology and are adapted to connection problems and normalization changes.

Third, geometric versus physical interpretation: in mathematical exact WKB, the symbol is primarily a period-like coordinate on a spectral cover. In semiclassical physics, it often appears as the exponential correction governing tunneling, connection matrices, or exact level conditions. These are complementary rather than competing interpretations.

Finally, Voros symbols should not be conflated with Stokes multipliers themselves. Stokes multipliers are entries of connection matrices between canonical solutions; Voros symbols are period-type exponentials from which those multipliers can often be reconstructed. Their utility lies precisely in converting connection problems into multiplicative period data.

In contemporary usage, Voros symbols occupy a junction point among asymptotic analysis, complex geometry, and wall-crossing formalisms. They provide a language in which the global analytic behavior of WKB solutions becomes computable, homological, and, in many important cases, birationally structured.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (1)

Topic to Video (Beta)

No one has generated a video about this topic yet.

Whiteboard

No one has generated a whiteboard explanation for this topic yet.

Follow Topic

Get notified by email when new papers are published related to Voros Symbols.