Early whispers of change: Manuel Santos Gutiérrez on climate response, time, and tipping points

When people hear the phrase “climate sensitivity,” it often sounds like something fixed and well-defined: a number scientists have already pinned down. But in reality, much of climate change unfolds far from equilibrium, during long, uneven adjustment periods that are difficult to observe and even harder to model. Long before the climate reaches any new “steady” state, it passes through drawn-out periods of adjustment, instability, and surprise. These transitional phases, known as transient responses, may hold some of our best clues about what lies ahead.

Dr Manuel Santos Gutiérrez is a research associate based at the University of Leicester, working as part of the team developing the theoretical basis for understanding variability in models and paleoclimate data in Past to Future.

In this conversation, Manuel explains why those transient phases matter, how scientists search for early warning signs of tipping points, and what it means to study a climate system that is already in motion.

Climate sensitivity is often framed as a long-term number. What gets lost when we focus only on the final equilibrium and not on how the climate gets there?

The equilibrium climate sensitivity is a key metric for understanding how the Earth system responds to radiative forcings, such as carbon dioxide emissions. Transient response, on the other hand, focuses on the development of feedbacks before the system equilibrates. Both transient and equilibrium responses are considered in the IPCC report through various model comparisons. We believe that resolving transient features of climate’s response is crucial, since globally averaged temperatures can overshoot and surpass thresholds that take generations to relax to its equilibrium value.

Why is it so difficult or expensive to observe equilibrium climate states directly, either in data or in models?

While data provide our ground truth, the climate system is continuously and inhomogeneously forced in multiple ways—fluctuations in solar irradiance, greenhouse gas emissions, changes in land use—these factors take us away from equilibrium. Models, however, allow us to keep the parameters unchanged, but their high dimensionality and computational cost require robust curation of model outputs to ensure that we are indeed simulating equilibrium conditions.

Individual model runs (coloured lines) fluctuate strongly due to internal variability, obscuring the underlying climate response (black line). Recovering a stable signal requires many simulations, which is computationally expensive — especially for complex models. Approaches based on correlations aim to extract the same signal more efficiently from noisy data. Figure by M Santos Guitiérrez

You talk about replacing ensemble averages with correlations. For someone outside climate physics, what problem does this actually solve?

Taking averages from multiple model outputs is a fundamental procedure in weather and climate science; it provides accurate statistical predictions assuming some initial uncertainty. When the model at hand is extremely high-dimensional, a larger and larger number of ensemble members is required to better resolve the prediction. A few ensemble members produce noisy outputs, making it difficult to spot the pattern of the signal we are after. This is where correlations enter the picture: they yield enhanced signals for the same number of ensemble members, particularly for short or transient time horizons. In a nutshell, correlations help to better extract the signal from noisy data.

People often imagine tipping points as sudden, dramatic events. In reality, what does a system look like as it approaches a tipping point?

Tipping points in climate subsystems can be sudden or not, depending on the timescales one considers. With regard to the tipping phenomena that we, as people, are facing, they might not look as sudden. However, there are indicators that suggest what can happen. For instance, the subsystem’s variability tends to increase in proximity to tipping. This was realized for the Atlantic Meridional Overturning circulation, suggesting a possible collapse within a few generations. Still, finding the optimal way to measure proximity to tipping is an active area of research.

Advancing monsoon clouds and showers in Aralvaimozhy, near Nagercoil, India. Photo by w:user:PlaneMad, CC BY-SA 3.0, via Wikimedia Commons

The Atlantic Meridional Overturning Circulation (AMOC) features prominently in your work. Why is this system so important for understanding climate stability?

The AMOC is an ocean circulation system responsible for redistributing energy, heat, nutrients, and salt between tropical regions and the poles. In fact, it is also responsible for the warmer temperatures in Europe compared with other regions at similar latitudes. A weakening of the AMOC can affect weather patterns in Europe, as well as tropical wind convergence and the monsoons. Due to climate change, freshwater is being input into the Arctic regions, which puts the brake on such circulation. Will such a circulation collapse as warming progresses? How do we know that it will tip? In our work, we look for statistical indicators that signal a tipping point in this circulation.

You are using methods that borrow ideas from statistical physics and molecular systems. What do you find exciting about applying these tools to climate science?

The fields of statistical physics and climate science have largely evolved in parallel, although that has changed over the past few decades. There is a myriad of mathematical tools that could potentially be useful in the study of climate. While the maths is universal, it takes work to translate equations formulated for molecular fluids into the language of climate. Spotting when the formulas fail and adapting them to the nonlinear, chaotic equations governing our climate is an aspect of my work that I deeply enjoy.

Climate change is often framed in terms of endpoints, but Manuel’s work draws attention to how the climate behaves as it changes and how signs of instability may emerge before dramatic shifts occur. Learning to recognise these early whispers of change could improve how scientists assess risk and anticipate abrupt changes in the climate system.