Linking Stochastic Self-Propagating Star Formation and Spatio-Temporal Point Processes
Qihan Zou · 2025
Read Linking Stochastic Self-Propagating Star Formation and Spatio-Temporal Point Processes by Qihan Zou (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-12372.txt.
Introduction
Qihan Zou connects two ways of modeling how star birth spreads through a galaxy, in a 2025 paper whose head title names a spatio-temporal log-Gaussian Cox-Hawkes process. The classic self-propagating model says each new star can trigger or block nearby births. Point process theory offers a richer language for such event patterns.
Zou proves the classic update law can be rewritten as a conditional Poisson rule. The new model then stacks three drivers into one log-scale rate: fixed galactic structure, a wandering latent field, and the pull of past events. That pull can be signed, so past births may excite or inhibit new ones while the rate stays valid. The frame suits maps of star formation that cluster and spread.
Catalog record
| CREATOR | Qihan Zou |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-12372
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Linking Stochastic Self-Propagating Star Formation and Spatio-Temporal Point Processes Qihan Zou Abstract The stochastic self-propagating star-formation (SSPSF) model is an important theoretical framework for explaining how localised star-formation events trigger subsequent activity across galactic discs. While widely used to interpret spiral and irregular structures, its probabilistic rules have lacked a formal statistical foundation. In this work, we establish a connection between the SSPSF model and spatio-temporal point processes (STPP), which describe events in space and time through history-dependent intensities. We show that the SSPSF update law is equivalent to a separable spatio-temporal Hawkes process, and we derive a simple likelihood function that recovers SSPSF parameters from historical star-formation event data under simplifying assumptions. Beyond the statistical formulation, the framework provides a new approach to analysing the propagation of star formation in galaxies, enabling observational surveys of star-forming regions to be interpreted within the STPP framework. Furthermore, the approach naturally extends to continuous-time models, offering a more realistic representation of galactic dynamics. keywords galaxies: formation, galaxies: evolution, Spatio-temproal point processes 1 Introduction The Stochastic Self-Propagating Star Formation (SSPSF) model is one of the important framework developed to model the star formation processes of galaxies introduced by Mueller and Arnett 1976 . In contrast to the density wave theory for spiral galaxy (e.g. Lin and Lau 1979 , Lin and Shu 1987 , Shu 2016 ), SSPSF used stochastic self-propagating to explain the formation of various types of galaxies. This theory has been further developed by Gerola and Seiden 1978 ; Gerola et al. 1980 . Comins 1981 builds on the SSPSF framework, but allows the triggered star formation probability to vary with radius. This mimics the decrease in star formation efficiency in the outer disk. Statler et al. 1983 extend SSPSF into three dimensions by stacking multiple 2D layers with vertical neighbor connections. Subsequent work by Seiden and Schulman 1990 proposed percolation-based extensions to the SSPSF framework, emphasizing the role of local connectivity and threshold-induced activation. Jungwiert and Palous 1994 introduce anisotropy into the propagation rule, triggering probability is distributed elliptically around a star-forming cell, with ellipse orientation and eccentricity tied to the disk’s rotation curve. On the other hand, spatiotemporal point processes are a fundamental framework in statistics to model events in space and time (e.g. Ogata 1988 , Daley and Vere-Jones 2003 , Diggle 2006a , Diggle 2013 ) and have been applied to a wide range of phenomena characterized by clustered, self-propagating events. Ogata 1988 introduces likelihood‐based point process models for earthquake occurrences and a residual analysis method to detect features such as seismic quiescence for prediction. Diggle 2006b proposes a partial likelihood approach for inference in spatiotemporal point process models, avoiding the computational burden of full likelihood methods. Diggle 2006a reviews statistical models for spatio-temporal point process data, illustrating their application through real-world case studies in epidemiology and biology (e.g. foot and mouth disease, Gastroenteric disease, Bovine tuberculosis, etc.), and highlighting both empirical and mechanistic approaches to inference. Gabriel and Diggle 2009 extends the inhomogeneous K-function to the spatio-temporal domain, proposing a nonparametric estimator that adjusts for spatial and temporal inhomogeneity, and applying it to Campylobacter infection data. González et al. 2016 provide a comprehensive review for spatiotemporal point processes, covering fundamental concepts, and applications across epidemiology (e.g., Ebola) and natural harzard (e.g., Tornadoes). Reinhart 2018 reviews self-exciting spatio-temporal point processes, summarizing their theory, estimation methods, and applications in fields like earthquakes, crime, and disease spread. Recently, with the development of neural networks and deep learning, the integration of spatio-temproal point processes with deep learning has made greater progress and has been applied in a wider range. Wang et al. 2020 introduces a spatiotemporal point process framework using a convolutional RNN for tracking multiple objects. Zhou et al. 2022 propose a flexible deep spatio-temporal point process method with nonparametric intensity function that can infer irregular events. Yuan et al. 2023 introduces Spatio-Temporal Diffusion Point Processes framework that models spatio-temporal point processes without assuming conditional independence between time and space. One of the common processes in the spatio-temproal point processes framework is the spatio-temporal Hawkes process (e.g. Reinhart 2018 , Bernabeu et al. 2025 ). It is particularly well suited to systems in which events occur at points in space and time, each event may generate multiple secondary events after a finite delay, and the influence of an event decays with both temporal and spatial separation from its origin. In the astrophysical context, the stochastic self-propagating star formation (SSPSF) mechanism fits naturally into this formulation: the background rate \mu(\mathbf{s}) represents spontaneous star formation, while the triggering kernel g models the finite-time, finite-radius causal influence of a supernova shock front on subsequent star formation in surrounding regions. While there are many studies on the stochastic self-propagating star formation model and spatio-temporal point processes, these two frameworks have not appeared to be directly related. In this work, we build a bridge between the stochastic self-propagating star formation model and spatio-temporal point processes, establishing a link between these two important theories. The structure of the paper is organized as follows. In Section 2, we briefly review the classical stochastic self-propagating star formation model (SSPSF). In Section 3, we discuss the basic background of spatio-temporal point processes (STPP), with a focus on spatio-temporal Hawkes processes (STHP). In Section 4, we establish a connection between simplified spatio-temporal point processes and the simplified self-propagating star formation model under certain conditions. We then derive the log-likelihood function for the simplified STPP, which can be used to recover the spontaneous and triggered star-formation probabilities of the SSPSF from historical event data. In Section 5, we conduct simulation studies to examine the estimation of these probabilities using the proposed log-likelihood function, based on simulated SSPSF data under simplified settings, to further demonstrate the connection between the two frameworks. In Section 6, we provide further discussions and conclude the work. 2 The classical Stochastic self propagating star formation (SSPSF) model In this work, we focus mainly on the classical Stochastic Self-Propagating Star Formation (SSPSF) model introduced by Gerola and Seiden 1978 . This model represents a galactic disk as a two-dimensional array of concentric rings, each subdivided azimuthally into equal-area cells simulated regions in the galaxy where the star can be born. The disk rotates differentially according to a prescribed rotation curve, such that the set of nearest neighbours of a given cell changes over time as the rings shear past one another. Each cell of the disk can contain at most one star at a given time step, where a star refers to a star-forming region such as a young stellar cluster that is capable of producing massive stars that can trigger further star formation in surrounding regions. A supernova denotes a cell whose massive stars produce feedback strong enough to initiate star formation in its surrounding cells. This discrete, probabilistic stochastic star formation framework couples local star formation propagation to the global galaxy disk, generating persistent spiral structure without invoking any preimposed density wave. The model evolves according to the following rules: 1. Initialization of the model: randomly assign K cells with stars before simulation. Every star becomes a supernova in the very next time step after it is formed. 2. Triggered star: If a cell