Year: 2026
Type: article
Abstract: Causal interactions in dynamical systems are structured not only across variables but also across time. However, conventional causal representations typically collapse this temporal organization into variable-level links, obscuring when causal influences act. Here we introduce interventional spatiotemporal causality (ISTC), a framework that represents causal organization over the joint space of variables and temporal delays. ISTC reconstructs the high-dimensional state space of a dynamical system from observations, learns a digital twin of the embedded dynamics, and performs targeted interventions within the learned dynamics to quantify intervention responses. Synthetic benchmarks show accurate recovery of causal topology and characteristic delays, while real-world applications reveal interpretable spatiotemporal patterns of inferred causal influence consistent with domain knowledge. The framework integrates three key features: Joint modeling of causal relationships across spatial and temporal dimensions, intervention-based inference grounded in reconstructed dynamical mechanisms, and a flexible state-space reconstruction strategy that balances causal resolution with computational efficiency. By making the temporal organization of causal interactions directly accessible, ISTC provides a general framework for characterizing causal dependencies in complex spatiotemporal dynamical systems.
Source: Advanced Science
DOI: 10.1002/advs.78791
Authors: Tianying Chen, Jintong Zhao, Siyang Leng, Wei Lin
Language: English
Subfield: Statistical and Nonlinear Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $6,730
OpenAlex ID: W7236478701
