Skip to content

Quasicircle

A planar Jordan curve with uniform bounded turning, equivalently the image of a circle under an ambient quasiconformal homeomorphism.

Version
v1 · 2026-10-17 · History
Domain-specific #
14270
Domain group
Formal Sciences
Origin domain
Mathematics
Subdomains
Quasiconformal Geometry, Complex Dynamics → Mathematics
Aliases
Quasi-circle

Core Idea

A quasicircle is a planar simple closed curve that can be obtained as the image of a circle under a quasiconformal homeomorphism of the whole plane. The same class can be recognized from the curve without exhibiting a map: there is one finite constant that bounds how far an appropriate subarc can turn away from the straight-line distance between any two of its points. The constant must work for every point pair. This quantitative condition distinguishes the class from Jordan curves known only to be simple and closed.[1][2]

In the intrinsic formulation, choose the connecting subarc with smaller diameter, not the one with shorter arclength. A quasicircle may have infinite arclength. Brück writes a three-point inequality for points on that smaller-diameter arc; the connected-subset diameter formulation states the same bounded-turning class with a changed numerical constant. Ahlfors's earlier reflection theorem gives a related quantitative geometric characterization of curves admitting quasiconformal reflection.[1][2][3]

The class includes curves produced by unlike processes. Under a stated small-parameter bound, a quadratic Julia boundary is a quasicircle. The classical outward Koch snowflake is one too, despite its infinite length. Those are examples of the curve type, not parts of its definition.[1][2]

Structural Signature

  • Planar Jordan carrier: A simple closed curve supplies two connecting subarcs between each distinct pair of its points. A self-intersecting boundary fails before any turning estimate is relevant.[2]
  • Ambient or intrinsic witness: Either a quasiconformal homeomorphism maps a model circle onto the carrier, or the equivalent bounded-turning condition is proved directly on it.[1][2]
  • Point-pair comparison: The chord distance between two curve points is compared with the diameter of a connected curve piece joining them. This exposes an arbitrarily large detour between nearby endpoints.[2]
  • Uniform finite bound: One constant must control every such pair. A separate bound chosen after seeing each pair would not establish a quasicircle.[1][2]
  • Curve rather than construction: Julia iteration, snowflake recursion, or a particular choice of ambient map can produce or witness membership, but none is necessary to every quasicircle.[1][2]

Recognition test: Is the object a planar Jordan curve, and can a single ambient quasiconformal circle-image construction or one uniform bounded-turning estimate be supplied? A drawing or topological homeomorphism alone leaves membership unverified; a proven cusp or self-intersection can rule it out.

What It Is Not

  • Not an arbitrary Jordan curve. A simple closed boundary supplies topology, while quasicircle membership adds quantitative control. Brück reports the Julia set of \(z^2+1/4\) as a Jordan curve with cusps that is not a quasicircle.[1]
  • Not a quasisymmetric map. That live catalog identity is a homeomorphism controlled by ratios of metric triples. Here the subject is the planar curve produced or certified by a geometric condition; related boundary maps do not turn a curve into a map.
  • Not every Julia set. Brück's proved sequence case assumes \(|c_n|\leq\delta<1/4\). A Julia origin alone does not settle the Jordan and turning tests.[1]
  • Not every Koch-style snowflake. Van Golden, Kombrink, and Samuel prove bounded turning under injectivity of the closed parametrization and contrast the classical outward snowflake with a self-intersecting anti-snowflake.[2]
  • Not a smoothness or finite-length condition. Nonrectifiable quasicircles are possible, including the classical snowflake case used below.[1][2]

Scope of Application

The term applies literally to planar Jordan curves whose global geometry meets the map or turning criterion. In geometric function theory it identifies controlled quasiconformal circle images and supports geometric recognition without fixing one map. In complex dynamics it classifies a Julia boundary only when the needed parameter and curve conditions are proved. In self-similar fractal geometry it can classify a snowflake boundary when the construction yields an injective Jordan parametrization and bounded turning.[3][1][2]

These settings do not make all fractals, all Julia sets, or all topological circles quasicircles. Each new carrier needs its own global witness. The examples here cover Brück's \(\delta<1/4\) family and the snowflake conditions proved in the cited preprint; they do not silently extend either theorem beyond those hypotheses.[1][2]

Clarity

There are two equivalent ways to state membership and two distinct questions they answer. An ambient map shows how a circle is deformed into the candidate curve. Bounded turning shows what global constraint the resulting curve satisfies. The latter is not a claim about the shortest route along the curve: on a nonrectifiable quasicircle, arclength can be infinite. Diameter measures how far the connecting curve piece spreads in the plane.[1][2]

The quantifiers matter. A fixed candidate curve must admit one finite turning bound that works for all point pairs. A value of \(K\) associated with one witnessing map is not the identity of the curve; maps may be changed by precomposing a rotation of the model circle, and an already valid upper bound can be enlarged. Conversely, absence of a supplied proof is not evidence of failure. The \(z^2+1/4\) Julia boundary is a reported negative because it has cusps, whereas a newly drawn Jordan curve with no quantitative analysis is merely unclassified.[1]

Moving, rotating, or uniformly rescaling the whole plane preserves the intrinsic test. Both the connecting-subarc diameter and the endpoint distance scale by the same factor, so their ratio and any finite bound on it remain unchanged. This is a direct geometric inference from the bounded-turning formulation; it does not claim that an arbitrary quasiconformal change preserves the same numerical bound.[1][2]

Manages Complexity

The class compresses many different curve constructions into one checkable global restriction: the curve cannot make arbitrarily large detours relative to the chord of nearby points. That lets a result about quasicircles be recognized in a dynamical boundary or a recursive fractal even though those constructions share little besides their resulting geometry.[1][2]

The compression has a limit. Membership does not itself give the best turning constant, finite length, smoothness, or a particular Hausdorff dimension. Brück explicitly allows nonrectifiable quasicircles, and the snowflake example has infinite length. Any finer claim must be proved for the particular family rather than read back from the class label.[1][2]

Abstract Reasoning

  1. Identify the proposed carrier as a planar simple closed curve. If this fails, stop: bounded turning cannot rescue a non-Jordan object.[2]
  2. Choose a proof route. Exhibit an ambient quasiconformal image of a circle, or formulate a bound on the diameter of a connecting curve piece for all endpoint pairs.[1][2]
  3. In a three-point formulation, use the subarc of smaller diameter and keep the same finite constant throughout the curve. Do not use arclength as a surrogate.[1]
  4. Carry every source theorem's hypotheses into the conclusion: \(|c_n|\leq\delta<1/4\) for Brück's sequence result; injectivity for the generalized snowflake theorem.[1][2]
  5. Infer class membership only at that scope. A named construction may produce both qualifying and nonqualifying curves when its parameters or topology change.[1][2]
  6. Keep later measurements of length, dimension, or a sharp constant separate from the membership proof.[1][2]

Knowledge Transfer

The invariant that travels between the two source settings is the whole-curve membership test: Jordan topology plus a quasiconformal image or uniform turning bound. The dynamics of a Julia set do not transfer to a snowflake, and snowflake recursion does not transfer to a Julia set. What transfers is a way of recognizing the resulting planar curves as belonging to one quantitative geometric class.[1][2]

Outside planar quasiconformal geometry, “controlled distortion” may be a useful analogy, but it is not the Quasicircle identity unless the carrier and recognition condition survive. A generic notion of distortion does not by itself say that an output is a simple closed plane curve or that one bound controls every point pair. The curve's topology and uniform pairwise test are what give this label its precise reach.

Examples

Canonical: a Koch snowflake that is a quasicircle

Take the classical outward Koch snowflake, with \(p=1/3\) and the consistent outward replacement construction. Van Golden, Kombrink, and Samuel note that its closed parametrization is injective apart from the endpoint identification, so it is a Jordan curve. Their Theorem 3.3 then gives bounded turning under that injectivity hypothesis. The same paper establishes infinite length for the Koch-type curves. Thus this example separates quasicircle membership from finite arclength. Their inward anti-snowflake at the same \(p\) has repeated points in its parametrization and does not inherit the conclusion.[2]

Mapped back: carrier = injective closed snowflake boundary; witness = proved bounded turning; point-pair relation = a connecting continuum controlled by endpoint distance; uniformity = the theorem's one finite bound over the curve; excluded neighbor = the noninjective anti-snowflake.[2]

Applied: a small-parameter Julia boundary

Brück studies Julia sets for compositions of quadratic maps \(z^2+c_n\). Corollary 4.5 shows that if all parameters obey \(|c_n|\leq\delta<1/4\), the resulting Julia set is a quasicircle. Setting every \(c_n=c\) gives a fixed-polynomial instance for \(|c|<1/4\), within this sufficient range. The source proves membership; this entry does not claim a numerical optimal bound or extend the result to arbitrary quadratic parameters. Brück specifically reports that the \(c=1/4\) Julia set is still Jordan but has cusps and is not a quasicircle.[1]

Mapped back: carrier = Julia boundary of the qualified quadratic system; witness = Brück's quasicircle theorem, which entails a uniform curve condition; point-pair relation = the corresponding bounded-turning test; boundary = the strict parameter condition cannot be dropped merely because the set remains Jordan at \(c=1/4\).[1]

Structural Tensions

T1: Constructive map witness vs. intrinsic curve proof. One can try to build an ambient quasiconformal map from a circle, which makes the deformation explicit but may require a difficult construction for a recursive boundary. Or one can prove bounded turning directly on the curve, which can establish membership without exhibiting the map but requires one estimate covering all pairs, including small-scale configurations. The two methods are equivalent as membership tests, yet they demand different evidence and can expose different gaps in a proposed argument. This is a methodological inference from the equivalent characterizations and the proof routes used in the cited works, not a claim that those authors measured a general cost of proof.[3][1][2]

Diagnostic: For this new curve, do we actually have an ambient map, or can the construction deliver one uniform pairwise turning bound? Which step would fail near the tightest bends?

Structural–Framed Character

Evaluative weight. Membership is a mathematical classification, not praise of a curve's beauty or usefulness. Human-practice dependence. A proof chooses a model circle, map, parameter convention, or turning formulation; those choices affect how membership is established but do not create it. Institutional origin. No statute, standard-setting body, or organizational role defines the class; its operative condition is mathematical. Vocabulary travel. “Circle,” “Jordan curve,” “quasiconformal,” and “bounded turning” keep this entry in geometric analysis; loose uses of “quasi” or “distortion” elsewhere do not carry its full test. Import versus recognition. Once the curve and plane metric are fixed, the question is whether the curve satisfies the condition, not whether a reviewer assigns it a label.[3][1]

Its character: predominantly structural, with bounded mathematical framing in the choice of proof witness and quantitative convention. That framing can change the presentation of an argument, while the eligible carrier and uniform criterion remain the same.

Structural Core vs. Domain Accent

The portable skeleton is Boundedness: one finite threshold controls a specified magnitude over an entire carrier, and the bound survives the relevant changes of position and scale. Here the magnitude is a geometric turning ratio and the carrier is every distinct point pair on one planar Jordan curve. The model circle and ambient quasiconformal witness are equivalent ways to recognize that class. Remove the planar Jordan and turning constraints and one has a generic bound, not a quasicircle.[1][2]

A future Prime proposal would need unlike nongeometric carriers whose necessary roles and failure tests preserve the same whole pattern, not merely examples with finite distortion. The present evidence offers two unlike constructions inside planar geometry. If only a generic “bounded change” motif survives outside that domain, the full Quasicircle identity remains domain-specific. No broader Prime is asserted from these examples alone.

This entry presupposes Boundedness and presupposes Ratio.

Quasicircle presupposes two independent Primes in its intrinsic test. Ratio supplies the ordered, same-endpoint comparison of smaller connecting-subarc diameter with nonzero chord length; its Comparison ancestor is present when those lengths are aligned in one Euclidean frame. Boundedness requires one finite upper threshold on that quotient across every point pair, preserved by Euclidean similarities. A non-quasicircle Jordan curve can still have the quotient without its uniform finite bound, and a generic bounded quantity need not be a ratio. Curve is an ordinary geometric genus; the present entry requires the narrower Jordan and turning conditions.

The live Distortion Prime describes a mapping-induced deviation from a faithful input-to-output rendering. Quasicircles are related to controlled deformation, but a round circle itself qualifies and has no required deviation from a faithful rendering. Bounded distortion in a witnessing map therefore does not turn every quasicircle into an instance of Distortion. The live Transformation Prime similarly describes a mapping, while the present subject is a curve type; a map can witness membership without becoming the curve's genus. The map and the curve answer different structural questions.

The live Quasisymmetric Map entry is a nearby domain-specific neighbor: its triple-ratio condition controls a map, whereas this entry's intrinsic test controls a plane curve. The identities interact in quasiconformal geometry but answer different recognition questions. The bounded-turning condition speaks about the whole curve even when no particular boundary parametrization is chosen.

Relationships to Other Abstractions

Local relationship map for QuasicircleParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.QuasicircleDOMAINPrime abstraction: Boundedness — presupposesBoundednessPRIMEPrime abstraction: Ratio — presupposesRatioPRIME

Current abstraction Quasicircle Domain-specific

Parents (2) — more general patterns this builds on

  • Quasicircle presupposes Boundedness Prime

    A quasicircle presupposes one finite upper bound on its smaller-subarc-diameter to chord ratio over every point pair.

  • Quasicircle presupposes Ratio Prime

    The uniform turning condition presupposes a same-endpoint ratio of smaller connecting-subarc diameter to nonzero chord length.

Hierarchy paths (2) — routes to 2 parentless roots

Neighborhood in Abstraction Space

Quasicircle sits in a sparse region of the domain-specific corpus (82nd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Riemannian & Symplectic Geometry (17 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-17

Not to Be Confused With

An unverified Jordan curve has an open membership question; the cusped \(z^2+1/4\) Julia boundary is a reported negative. The smaller-diameter subarc is not the shorter-arclength subarc. A K-quasiconformal witness has a quantitative bound, but the chosen map or number K is not the curve's identity. The Koch snowflake's infinite length does not prevent membership, while the self-intersecting anti-snowflake fails the Jordan prerequisite. A Julia-set construction is not itself a guarantee: carry its parameter and topology hypotheses into every claim.[1][2]

References

[1] Brück, Rainer. “Geometric Properties of Julia Sets of the Composition of Polynomials of the Form \(z^2+c_n\).” Pacific Journal of Mathematics 198, no. 2 (2001): 347–372. https://doi.org/10.2140/pjm.2001.198.347. Original publisher PDF: https://msp.org/pjm/2001/198-2/pjm-v198-n2-p06-s.pdf. Definition and smaller-diameter three-point condition, p. 351; Corollary 4.5 and reported \(c=1/4\) cusped counterexample, p. 357. The latter paragraph reports prior results and cites earlier books; it is not presented as Brück's new proof. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w ↩x ↩y ↩z ↩27 ↩28

[2] van Golden, Sven, Sabrina Kombrink, and Tony Samuel. “On the Geometry of Generalised Koch Snowflakes.” arXiv:2601.07371v1 (12 January 2026). Original 15-page preprint: https://arxiv.org/pdf/2601.07371. Infinite-length setting in the abstract and §2; bounded-turning definition, Proposition 3.2, Theorem 3.3, and the classical-versus-anti-snowflake contrast in §3, PDF pp. 7–9. Preprint status is retained; no later journal edition was consulted. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w ↩x ↩y ↩z ↩27

[3] Ahlfors, Lars V. “Quasiconformal Reflections.” Acta Mathematica 109 (1963): 291–301, Part I, Theorem 1 and proof, pp. 295–297. Original article full text: https://archive.ymsc.tsinghua.edu.cn/pacm_download/117/5956-11511_2006_Article_BF02391816.pdf. The scanned text extraction damages mathematical typography, so this entry does not reproduce its displayed inequality. registry ↩a ↩b ↩c ↩d