Climate

Tracing the space-time causal origins of Earth system extremes

Jhayron S. Pérez-Carrasquilla, J. Jake Nichol

Featured July 31, 2026

This analysis was generated by SciGrove. Upload your own PDFs or enter a DOI — and get the same AI breakdown on any paper.

Get started

AI-generated analysis — This is SciGrove's AI interpretation of the paper, not peer-reviewed content. Always refer to the original paper.

Simply

A new data-driven method traces the space-time origins of extreme weather events, revealing which factors causally contributed and by how much, like following a weather event's family tree backward.

In depth

The paper introduces Tracer of Causal Evolutions in Space and Time (TraCE-ST), a novel probabilistic Lagrangian framework designed to identify event-conditioned causal pathways in high-dimensional Earth system data. It recursively traces causal influence backward in space and time from a target extreme event, leveraging local causal discovery methods to group and probabilistically select causal parents, thereby quantifying their relative contributions to the event's evolution.

Key Takeaways

1
TraCE-ST provides a data-driven framework to trace multi-step, multivariate causal pathways leading to specific Earth system extreme events, moving beyond local causal links.
2
The probabilistic formulation of TraCE-ST allows for the estimation of relative causal contributions from multiple interacting variables and regions, offering a more complete explanation of complex events.
3
The framework is flexible, accommodating various underlying causal discovery engines (e.g., Elastic-Net, PCMCI), and has been validated across diverse synthetic and real-world extreme events, including tropical storms and heatwaves.

Conceptual Flow

HIGH LEVEL

Methodology: Tracing Causal Paths

The method starts from an event, finds its immediate causes, then finds the causes of those causes, tracing backward like following a river upstream.

Event Happens

Find Past Causes

Cause 1

Cause 2

Cause 3

Results: Quantifying Contributions

By repeating the tracing many times, the method shows which types of causes and locations were most important for the event.

Many Traced Paths

Count Important Causes

Cause A: High Impact

Cause B: Medium Impact

Cause C: Low Impact