Designing Early Warning Systems for Approaching Tipping Points

• Applicants: Professor Dr. Florian Diekert; Professor Dr. Daniel Heyen
• Subject Area: Economic Policy, Applied Economics
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 576261184

Many complex systems – from ecosystems to climate systems – exhibit so-called tipping points: critical thresholds that, when transgressed, can trigger abrupt and often irreversible regime shifts. Examples range from collapsing fish stocks and melting ice sheets to abrupt climate transitions. A fundamental problem in managing such systems is that the exact location of these tipping points is usually unknown. Research has shown, however, that systems often emit characteristic "early warning signals" before reaching tipping points – statistically measurable changes such as increased variability or slowed recovery from perturbations. These signals have been empirically demonstrated in various natural and anthropogenic systems, from experimentally controlled populations to paleoclimatic data. Despite these scientific insights, a fundamental research gap exists: the literature on early warning signals focuses almost exclusively on their detection, but not on their systematic integration into decision-making processes. Our research project addresses this gap by connecting for the first time the natural science literature on early warning signals with economic decision theory. Our research objective is to develop theoretical frameworks and experimental evidence for the optimal design of early warning systems. We systematically analyze how central design dimensions – granularity (number and structure of signals) and range (detection distance) – influence optimal decisions under uncertainty. Using mathematical models, we theoretically characterize how these dimensions determine tipping risk and the economic value of early warning systems. In parallel, we conduct controlled behavioral economic experiments to empirically determine how decision-makers respond to different early warning system designs and which configurations prove optimal. We systematically examine various decision environments: from single-actor scenarios to strategic multi-actor situations, from static to dynamic decision processes. The integration of theoretical analysis and experimental methodology enables both fundamental insights into optimal system design and understanding of real decision-making behavior. The research findings should contribute in the long term to improving societal management capacities for systems with tipping risks – particularly in environmental and climate policy, where early detection of critical transitions could enable adaptive and preventive decisions.

• DFG Programme: Research Grants