Marten Scheffer

This study evaluates early warning signals for infectious disease outbreaks using critical slowing down indicators and time-to-event analyses across multiple diseases and regions, finding reliable lead times for intervention.

This paper compares methods for detecting resilience changes in systems, showing that burst sampling outperforms continuous monitoring with the same data points, suggesting it can improve early warning signals for critical transitions.

This paper introduces an improved potential analysis method to better handle ecological data limitations and enhance predictions of ecosystem dynamics and resilience.

The paper shows that stochastic regimes can obscure attractors in data, and that reconstruction algorithms can reveal them by analyzing the relationship between noise and system equilibria.