[Paper Review] Finite time extinction of super-Brownian motions with catalysts

Donald A. Dawson, Klaus Fleischmann+1 more|ArXiv.org|Sep 21, 1998

Stochastic processes and statistical mechanics17 references3 citations

TL;DR

This paper establishes that super-Brownian motion with a catalyst measure—specifically a stable random measure of index $\gamma \in (0,1)$—exhibits finite-time extinction almost surely when started with compactly supported, finite initial mass. The proof uses a decomposition into 'good' and 'bad' historical particle paths: good paths, which collide significantly with the catalyst, are analyzed via time change and comparison to Feller's branching diffusion, while bad paths are controlled via hitting probabilities and local time, ensuring the total mass vanishes in finite time.

ABSTRACT

Consider a catalytic super-Brownian motion $X=X^Γ$ with finite variance branching. Here `catalytic' means that branching of the reactant $X$ is only possible in the presence of some catalyst. Our intrinsic example of a catalyst is a stable random measure $Γ$ on $R$ of index $0< gamma <1$. Consequently, here the catalyst is located in a countable dense subset of $R$. Starting with a finite reactant mass $X_0$ supported by a compact set, $X$ is shown to die in finite time. Our probabilistic argument uses the idea of good and bad historical paths of reactant `particles' during time periods $[T_{n},T_{n+1})$. Good paths have a significant collision local time with the catalyst, and extinction can be shown by individual time change according to the collision local time and a comparison with Feller's branching diffusion. On the other hand, the remaining bad paths are shown to have a small expected mass at time $T_{n+1}$ which can be controlled by the hitting probability of point catalysts and the collision local time spent on them.

Motivation & Objective

• To investigate the long-time behavior of super-Brownian motion when branching is dependent on a catalyst measure.
• To determine whether such processes with finite initial mass die out in finite time rather than persisting indefinitely.
• To analyze the role of catalyst location and structure—specifically a stable random measure of index $\gamma \in (0,1)$—in driving extinction.
• To develop a probabilistic framework using historical paths and local time to control extinction time.

Proposed method

• Decompose the historical paths of reactant particles into 'good' and 'bad' sets based on collision local time with the catalyst.
• Apply time change to good paths using their collision local time, transforming them into a process comparable to Feller's branching diffusion.
• Use comparison techniques with Feller's diffusion to show that good paths lead to extinction in finite time.
• Bound the expected mass of bad paths using hitting probabilities of point catalysts and the local time spent on them.
• Control the cumulative contribution of bad paths across successive time intervals $[T_n, T_{n+1})$ via moment estimates.
• Combine estimates from good and bad paths to show that the total mass of the superprocess vanishes in finite time almost surely.

Experimental results

Research questions

• RQ1 Does super-Brownian motion with a catalyst measure of index $\gamma \in (0,1)$ exhibit finite-time extinction when started with finite, compactly supported initial mass?
• RQ2 How does the local time of particle paths with respect to the catalyst influence the extinction behavior?
• RQ3 Can the extinction time be bounded using a decomposition of paths into those with significant catalyst interaction versus those without?
• RQ4 What role does the spatial distribution of the catalyst—specifically a countable dense set—play in the extinction mechanism?
• RQ5 To what extent can the behavior of the superprocess be controlled via comparison with Feller's branching diffusion?

Key findings

• The super-Brownian motion $X^\Gamma$ with a stable catalyst measure $\Gamma$ of index $\gamma \in (0,1)$ almost surely dies out in finite time when started with finite, compactly supported initial mass.
• The extinction time is almost surely finite, even though the catalyst is supported on a countable dense subset of $\mathbb{R}$, indicating that sparsity of catalysts does not prevent extinction.
• The decomposition into good and bad paths allows for a sharp control of the expected mass at each time step, with bad paths contributing negligibly over time.
• The use of time change based on collision local time enables a direct comparison with Feller's branching diffusion, which is known to die out in finite time.
• The hitting probability of point catalysts and the local time spent on them are key to bounding the contribution of bad paths.
• The cumulative effect of the path decomposition ensures that the total mass of the process converges to zero in finite time almost surely.