Marked stable L'evy processes in Brownian local time
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Mathematics. The day in mathematics: new theorems, new proofs, and what they open.
Ada: We're looking at the paper: "Marked stable L'evy processes in Brownian local time".
Emmy: Here's the gist of the paper. We construct a random partition of R+ times R using two coupled stochastic squared Bessel flows, whose parameters differ by delta in (0, 2).
Ada: The cells of this partition correspond to squared Bessel excursions with a negative parameter-delta, which are embedded within the jumps of a spectrally positive (1 + delta 2) -stable process.
Emmy: This framework naturally gives rise to interval partition evolutions
FPRW20a: and stable shredded disks
BCS22: .
Ada: Specifically, we consider two coupled BESQ flows driven by the same white noise W.
Title and authors: Ada: We're talking about the paper "Marked stable L'evy processes in Brownian local time" and what it says about these coupled squared Bessel flows creating a random partition of space-time. We've touched on how this links to stable processes and the geometric structures they produce, like the spindles.
Emmy: Exactly. To recap, the core contribution is showing that the cells in this random partition correspond to squared Bessel excursions with a negative parameter, embedded in jumps of a spectrally positive stable process with index one plus delta squared. We also saw how this relates to interval partition evolutions and stable shredded disks arising from this framework.
Kolya: What I want to emphasize is the representation they found: they establish that the widths of these spindles as a function of vertical level are governed by a Poisson point process of BESQ excursions with parameter minus delta, which is in the range negative two to zero. That’s a very specific distribution we can use.
Ada: And then there's the second representation, where they show that those spindles themselves are naturally embedded in the jumps of this spectrally positive stable Lévy process, which has an index between one and two. This is really connecting the continuous geometry to discrete jumps again.
Emmy: The real structural insight here is that they establish a bijection between these spindles and the jumps of this stable process, which keeps track of their left-right order. That’s a very strong statement about how these two seemingly different objects are fundamentally related in this random setting.
Kolya: From my perspective on the probability side, it means we can analyze the behavior of these spindles by looking at the properties of that stable process. Specifically, conditional on knowing that process, the width processes simplify to being independent squared Bessel excursions with a duration dictated by that stable process's movement. That’s a simplification for analysis.
Ada: So they've managed to take this intricate construction involving two coupled flows and distill it down into these two key representations concerning the widths and the embedding of the spindles in jumps. It’s a lot to digest, but it gives us tools to study these complex partitions.
Emmy: And this connection is what makes it relevant for geometry too. They show that the gasket formed by these non-spindle regions is homeomorphic to something simpler, which is a half-plane analogue of the stable shredded sphere. That’s a nice piece of structural classification for this random set.
Kolya: It's interesting because they are using Ray–Knight theorems in their setup to give an alternative proof for those theorems using BESQ flows, which connects this new construction back to established probabilistic results in a different way.
The paper's summary: Ada: Moving on to the summary of "Marked stable L'evy processes in Brownian local time". They are essentially setting up a framework with two coupled squared Bessel flows driven by the same white noise, and they define regions delimited by blue and red lines.
Emmy: These regions, what they call spindles, are disjoint and bounded subsets of the half-plane that roughly partition it. The main point is that these spindles are not just random shapes; their properties have very specific statistical laws.
Kolya: They prove that the widths of these spindles when you look at them as a function of the vertical level follow a Poisson point process governed by BESQ excursions with parameter minus delta, where delta is between zero and two. That’s a concrete law for those widths.
Ada: And they also show that these spindles are embedded within the jumps of a spectrally positive stable Lévy process, which has an index of one plus delta squared, which is also between one and two. This links the geometry directly to this specific type of stable process.
Emmy: The main result they hammer home is the bijection between the spindles and these jumps of the stable process, ensuring that it respects their left-right ordering. This means if spindle A comes before spindle B, then jump A comes before jump B in some sense.
Kolya: And conditionally on knowing that stable process, the width processes of those spindles are independent BESQ excursions with a duration equal to the size of the corresponding jump in that stable process. That’s how they get those conditional simplifications.
Ada: So basically, they take this initial construction from coupled flows and distill it into these two main representations: one about widths governed by a Poisson process and one about the embedding in stable jumps, with a clear mapping between them.
Emmy: It's not just descriptive; it’s constructive. They are building something that explicitly shows how these two systems interact to produce this structured partition, which is really showing the mechanism behind the structure.
Kolya: The implication for me is that if you want to study these partitions, you don't need to look at every single spindle individually; you can focus on the aggregate behavior described by those point processes and stable processes.
The paper's improvements: Ada: Now let's talk about what improvements the authors suggest for "Marked stable L'evy processes in Brownian local time". They are looking at extending this framework beyond what they just proved.
Emmy: They suggest that the next step is to explore how these structures might relate to other types of partitions or perhaps how they could be used in different areas, like modeling something more dynamic.
Kolya: One improvement the paper points towards is studying how these spindle properties behave when you change some of those parameters, specifically by looking at what happens as delta changes. They want to see the sensitivity of these results to that parameter delta.
Ada: And another direction they suggest is looking at different types of partitions entirely, maybe trying to generalize this idea to partitions where the underlying flows aren't just coupled BESQ flows.
Emmy: They are suggesting a way to investigate how this whole system might fit into broader theories involving other stochastic objects, like perhaps looking at it through the lens of other algebraic structures.
Kolya: From a probabilistic viewpoint, they want to understand if these bounds they derived still hold when you move away from the specific range of parameters for delta or if there are ways to generalize those results. That’s checking the robustness of the findings.
Ada: So it seems like they are trying to see how far this specific framework can be pushed before it loses its structural integrity, which is a natural thing when you start pushing boundaries.
Emmy: It’s about finding that deeper connection between this specific construction and more general theories, rather than just proving the theorem for one fixed parameter set.
Conclusion: Ada: To wrap up on "Marked stable L'evy processes in Brownian local time". The paper shows us how to create a random partition using coupled squared Bessel flows, leading to two key representations: the widths of the spindles as governed by a Poisson process and their embedding in jumps of a stable process.
Emmy: And they solidified this through the bijection between spindles and stable jumps, showing that this structure is robust across different configurations. This gives us a solid mathematical model for analyzing these partitions.
Kolya: From my side, I just want to say that the results give us concrete probabilistic tools to analyze these structures without having to deal with every single spindle individually; it’s about focusing on the aggregate behavior governed by those point processes and stable processes.
Ada: And they also pointed out that they can look at how changing the parameter delta affects those results, which is important for seeing the limits of this specific construction.
Emmy: It’s a nice piece of work because it takes these complex flows and gives us a clear way to visualize and quantify their statistical behavior through these stable processes.
Kolya: So that's what we have here: concrete probabilistic tools to analyze random partitions, focusing on the aggregate behavior described by those point processes and stable processes.
Ada: We've looked at the paper "Marked stable L'evy processes in Brownian local time" and seen how this construction connects complex flows to fundamental stable processes. Thanks for tuning in with us today.
Elie A¡d®kon, Quan Shi, Chengshi Wang
Fudan University · Chinese Academy of Sciences
math.PR
Submitted: 2025-12-05
Updated: 2026-11-10
Comments: 58 pages, 11 figures. Revised title and exposition. Added a theorem confirming the conjecture of [Forman et. al. 2021] concerning the independence between the de-Poissonized interval partition evolution and the total mass process (Section 5.3)
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 95/100
The gist: We construct a random partition of R+ times R using two coupled stochastic squared Bessel flows, whose parameters differ by delta in (0, 2).
Key concepts
- Coupled BESQ Flows
- These are two stochastic flows driven by the same white noise, modeled as squared Bessel processes. They are used to generate a random partition of space-time. The parameters of these flows differ by $\delta$, which dictates the properties of the resulting structures, like interval partitions.
- Stable Lévy Process
- This is a type of stochastic process whose jumps have heavy tails, meaning extreme events are more likely than in a normal distribution. The paper shows that the underlying structure generating the partition is related to a spectrally positive stable process with index $1 + \delta^2$.
- Spindles and Jumps
- The random partition is characterized by 'spindles,' which are bounded subsets of the half-plane. The key finding is that these spindles are naturally embedded within the jumps of a spectrally positive stable Lévy process, creating a direct correspondence between geometric features and discrete jump events.
- Marked Lévy Processes
- This framework represents the complex random partition as a marked Lévy process. Here, the jump times of the underlying stable process correspond to spindle boundaries, and the sizes of these jumps relate to the widths of the corresponding spindles.
Terminology
Summary
We construct a random partition of R+ times R using two coupled stochastic squared Bessel flows, whose parameters differ by delta in (0, 2). p. 0 The cells of this partition correspond to squared Bessel excursions with a negative parameter-delta, which are embedded within the jumps of a spectrally positive (1 + delta 2) -stable process. p. 0 This framework naturally gives rise to interval partition evolutions [FPRW20a] and stable shredded disks [BCS22]. p. 0
Specifically, we consider two coupled BESQ flows driven by the same white noise W. p. 1 The region delimited by the blue line and the red line at its left is denoted by T(a,r) in (3. p. 16). Spindles are disjoint, bounded subsets of the half-plane that roughly form a partition. p. 2 The main results establish two key representations: 1. p. 2 The widths of spindles, viewed as a function of the vertical level, arise as a Poisson point process of BESQ- delta excursions with parameter-delta in (-2, 0); 2. p. 2 Spindles are naturally embedded in the jumps of a spectrally positive stable Lévy process xi of index 1 + delta squared in (1, 2). p. 2
The central result is a bijection between the spindles generated by (B, R) and the jumps of a stable process xi, such that the collection of pairs (r, z) over all spindles is given by the collection of jumps of xi. p. 2 This bijection preserves the left-right order of the spindles. p. 2 Conditionally on xi, the width processes of the spindles are independent BESQ- delta excursions of duration delta xi(t). p. 2
The random partition is reminiscent of studies in [PY82], where Pitman and Yor construct a BESQc process as a sum of a Poisson point process of BESQ0-excursions. p. 2 In view of the Ray–Knight theorem, this corresponds to decomposing the Brownian path into excursions. p. 2 A limiting object is shown to be related to stable shredded spheres [BCS22]. p. 3 The gasket K is shown to be homeomorphic to the half-plane analogue of the stable shredded sphere (Theorem 5. p. 39).
The analysis shows that xi:= X kappa eta is a (1 + delta 2) -stable Lévy process, spectrally positive starting from 0. p. 0 Its Lévy measure is nu(-delta) BESQ (zeta(f) in dy), and its Laplace exponent is 2 - delta squared q 1+ delta squared, for q at least 0. p. 2 This implies that the point process N = (e, at least 0) marking each jump of the subordinator eta is a Poisson point process whose intensity is a multiple of the BESQ- delta excursion measure nu(-delta) BESQ. p. 1 The range of i:= kappa eta is exactly the gasket K, and its jump times coincide with those of eta. p. 2 The vertical coordinate xi:= X kappa eta is a stable Lévy process. p. 2 The jumps of xi are further marked by excursions (e: at least 0) given by the width of the spindle T, and the lifetime of each excursion e equals the size of the corresponding jump delta xi. p. 4
The set A = A t T for a spindle T is characterized as A = A t T, t E. p. 1 The process A(c) is associated with (c). p. 4 The set A has Lebesgue measure zero and no isolated points. p. 2 The process L(u):= [at least 0: eta > u] is (Gu) -adapted and a. p. 0s. continuous, where Gu is a filtration defined in terms of the spindles discovered before time kappa u. p. 8 The joint distributions of (W c T, b T, B b T, R b T) and (W,, B, R) are equal. p. 3 The process X bT is continuous and L bT is bicontinuous. p. 1 The set A coincides with the set of A t 's for t E or t in the right part of a spindle.
The collection of spindles discovered before time T are measurable with respect to G b,s, where G b,s is a sigma-field defined in (3. p. 029). The bottom points (a, r) and the processes (R r, x(a-), x at least r) are G b,s-measurable. p. 4 The first exit time t T is also G b,s-measurable. p. 4 The process X lambda t - H T is a BESQ (2 s 0) flow driven by W cT on (I T, infinity), and the flow R - L(T, times) is a BESQ (delta H T 0) flow driven by theta H T W cT on (-infinity, I T). p. 1 The process Y:= (Y x, x at least I T) is a non-killed BESQ (2 - delta s - delta) process. p. 3
The collection of spindles T discovered before time T, their entrance times sT and their first exit times tT are measurable with respect to G b,s. p. 1 The set A is regenerative and progressive with respect to (Gu) u at least 0. p. 0 The set A theta D u is independent of G Du and has the same law as A. p. 2
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Enhanced understanding of complex stochastic partitions: The system can model random partitions of space-time using coupled squared Bessel flows to identify underlying stable Lévy processes, enabling it to recognize structures like interval partition evolutions and stable shredded disks in data.
-
Advanced analysis of local time exploration: The AI can perform detailed analysis of Brownian local time by interpreting the process as an explorer of the space-time plane, specifically identifying how Brownian motion explores spindles through left and right parts using concepts like hitting times and exit times.
-
Connection between geometric structures and probability: The system can establish a bijection between spindles generated by coupled flows and jumps of a spectrally positive stable Lévy process, allowing it to connect continuous spatial structures with discrete jump processes.
-
Characterization of random geometric objects: The AI can represent the gasket K, which is shown to be homeomorphic to the half-plane analogue of the stable shredded sphere, providing a mathematical framework for analyzing random planar maps.
-
Modeling exploration dynamics: The system can simulate and analyze the exploration process ϱ using time-changed measures and martingales derived from coupled flows to understand how an explorer moves through discovered and undiscovered regions of a spindle over time.
-
Identifying regenerative properties in random structures: The AI can recognize when a set of events related to spindle discovery has a regenerative property, allowing for the analysis of the set A = At where t is not in E, which is shown to be independent of certain filtrations.
-
Representation as marked Lévy processes: The system can represent complex random partitions as marked Lévy processes where jump times correspond to spindle boundaries and excursion sizes correspond to spindle widths, providing a richer statistical description of the underlying noise.
Sources
- Meeting of squared Bessel flow lines and application to the skew Brownian motion
- The Aldous diffusion: a stationary evolution of the Brownian CRT
Related papers
- Rank-One HCIZ Integrals and Multiplicative Convolutions
- Repeated averaging on random d-regular graphs
- Precise cover times for branching random walks on Hamming graphs: (iterated) logarithmic corrections
- Markov Chain CLTs: Resolving Open Problems
- A Sharp Refinement of Samuels' Inequality via Cap Moves
- A zero-one law for swap-invariant events in uniform spanning forests