Seminar Archive – Spring 2025

The quest for robust early warnings of noisy bifurcations

There is a lot of interest in finding “early warning signals” or “precursors” of tipping points due to bifurcations in cases where a system parameter is drifting. This is particularly relevant for tipping points where a fold bifurcation leaves no nearby solutions as stable. The idea of an early warning of a tipping point is compelling and potentially useful in a wide range of applications. In general, one examine the behaviour of a nonlinear system with weak noise, and finds properties of the noise response (critical slowing down) that indicates a tendency towards less stable solutions. The challenge is to understand when one can expect such a signal to be reliable and robust indicator of a future tipping event; this includes problems related to timescales, estimation and extrapolation, and begs the question of when a warning might be “too early” or “too late”. I will outline some ongoing work that aims to better understand when such warnings are reliable.

Tipping points and non-ergodicity in a rapid low-carbon transition.

A rapid low-carbon transition requires a rapid turnover of high-carbon capital to be replaced by low-carbon systems. This can lead to stranded assets, which happen when events occurring were not foreseen at investment time. Notably, a rapid transition to electric vehicles could destroy huge financial value in oil-related assets and cause difficulties to financial assets and pension funds. In extreme cases, stranded assets can trigger financial instability, as occurred in 2008. The way in which a rapid transition puts capital value at risk is not well understood, and has a deep connection with the non-ergodic and path-dependent nature of the process of economic transformation. Here we look at two systems, the transition to electric vehicles and its relationship with the oil sector, and the propagation of shocks in a stressed financial network. We introduce a simple algorithm-based model of low-carbon transition and related financial risk propagation across evolving networks of institutions with equity and debt relationships. Due to the complex network nature of financial relationships, and to domino effects of insolvencies that arise with shocks in highly-leveraged financial networks, the model generates heavy-tailed ‘sandpile’ or ‘earthquake’ behaviour very similar to those observed in standard financial market indices, which has a close relationship with financial tipping points. Measuring tail heaviness and mapping the phase diagram allows to construct a measure of systemic risk, and to estimate the amount of resilience the economy has towards financial instability. This can inform climate-related policy and financial regulation.

Positive tipping points for domestic heat decarbonisation: the need for social theory.

Implementation of climate mitigation policies is needed at an unprecedented rate to avoid crossing climate tipping points. While UK emissions have been cut by 50% since 1990, these emission reductions are largely due to the decarbonisation of the energy grid. While a crucial step, it has progressed without the need for significant public engagement. Tackling remaining emissions, however, will increasingly impact public and private life, necessitating social and behaviour change at an unprecedented pace and scale.

Domestic heat decarbonisation is an example of climate policy becoming personal, requiring almost every household in the UK to make a decision that will have a significant impact on their homes. The primary mechanism to support home heating decarbonisation in the UK is the Boiler Upgrade Scheme. Despite significant grant support, however, the scheme has seen lower than expected uptake. Meeting the government target of installing 600,000 heat pumps per year by 2028 will require an elevenfold increase in installations.

During this presentation I will make the argument that meeting UK climate targets is partially dependent on leveraging positive tipping points in this sector, but if a tipping point framework is to be applied to inform policy interventions, the social tipping points literature must engage more deeply with existing social and behaviour theories. I will outline how tipping points literature has engaged with social theory and behaviour change and make connections to why this is increasingly relevant to the decarbonisation of domestic heating in the UK.

---

Identifying effective policies for positive tipping points in the road freight transition.

The road freight sector is an important and growing contributor to global emissions.Compared with personal road transport, which is rapidly decarbonising in manycountries, progress to decarbonise road freight is much slower and characterised by costand logistical barriers. Public policy will be crucial to transition the sector away fromcombustion engine vehicles and toward battery electric and fuel cell vehicles.I present findings from FTT-Freight – a new model of technology decision-making in theroad freight sector based on evolutionary economics. The model aims to represent howconsumers make decisions on the purchase of new vans and trucks and how differentregulatory and market-based policies can impact uptake. I show how the effectivenessof different policies to trigger positive tipping points and increase the rate ofdecarbonisation in the sector varies substantially. Direct deployment policies are likelyto be particularly effective. Furthermore, significant coordination gains could be achieved if leading markets align policy objectives, making the transition cheaper for other countries around the world. 

---

An early warning indicator for tipping points in strongly forced systems.

Classical critical slowing down early warning signals (observing increasing trends in the autocorrelation and variance) have been developed to try and detect when a system is approaching a tipping point, typically represented mathematically by a bifurcation.  However, these signals often fail for strongly forced slow systems. Here we propose a new method that reconstructs the quasi-equilibrium state and therefore produces a robust indication of where the critical threshold may lie in a system. We show that the standard deviation of this reconstructed quasi-equilibrium state increases exponentially ahead of its critical threshold, at time t_crit, for both strongly forced fast and slow systems. Furthermore, we show that the standard deviation of the quasi-equilibrium state varies proportionally to 1/(t_crit – t)^alpha, where the exponent alpha can be derived analytically based on the type of bifurcation being crossed. Using both the reconstructed quasi-equilibrium state and fitting the above form to its standard deviation allows us to diagnose the location of the critical threshold without needing to have any further knowledge of the system’s parameter values.

---

Global tipping points: risks, opportunities, and early warning signals.

Any sufficiently complex system with alternative stable states – whether it is a biophysical Earth system or a human social system – can have a tipping point: A tipping point occurs when change in a system becomes self-perpetuating beyond a threshold, which often leads to abrupt and irreversible consequences. Tipping points in the natural world pose some of the gravest threats faced by humanity. Preventing them – justly and equitably – should become a collective, global governance goal. This will only be possible if we can rapidly transform our societies and economic systems using what we call positive social tipping points – for example, the exponential uptake of solar power, electric vehicles, and heat pumps.