Quantification of the cascading tipping probability from the AMOC to the Amazon rainforest

Valérian Jacques-Dumas, Henk A. Dijkstra · 2025

Read Quantification of the cascading tipping probability from the AMOC to the Amazon rainforest by Valérian Jacques-Dumas, Henk A. Dijkstra (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-13383.txt.

Introduction

Two of the climate system's feared tipping elements are linked: if the Atlantic overturning circulation collapses, the rainfall shifts could push parts of the Amazon toward degraded forest. Valérian Jacques-Dumas and Henk Dijkstra built a coupled conceptual model of both systems and asked, with numbers, how likely that cascade is and how long it would take.

Collapses are too rare to catch by ordinary simulation, so they used a rare-event algorithm that steers the ensemble toward transitions. For northwestern Brazil, they find a slide into degraded forest within 200 years very unlikely. Yet there it could only happen after an Atlantic collapse, whose drying effect favors extreme wildfires. The study quantifies a cascade often discussed only qualitatively: improbable on policy timescales, but conditional in a way that makes the first domino decisive.

Catalog record

CREATORValérian Jacques-Dumas, Henk A. Dijkstra
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE81 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-13383

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Quantification of the cascading tipping probability from the AMOC to the Amazon rainforest Valérian Jacques-Dumas, Henk A. Dijkstra Abstract The Amazon rainforest and the AMOC are two key components of the Earth system and may both collapse under climate change. Due to its influence on precipitation patterns, a collapsed AMOC influences the dynamics of the Amazon rainforest. We investigate this effect using a coupled conceptual AMOC-Amazon model. The Amazon model is based on empirical hydrological data controlled by the AMOC strength. The AMOC model and its influence on the Amazon are tuned using a simulated AMOC collapse in the Community Earth System Model (CESM). Since the collapse of both systems is very rare, we study it using a “rare-event” algorithm, which samples such events much more efficiently than direct numerical simulation. This algorithm also allows us to track many observables of interest of the coupled model. We find that in the centre of the Amazon basin an AMOC collapse is a necessary condition for the Amazon rainforest to collapse, due to its important drying effect. Moreover, we are able to quantify the importance of the AMOC in this tipping cascade by computing the conditional probability that the collapse of the Amazon rainforest follows that of the AMOC, given that the Amazon rainforest turns into a savannah within 200 200 years. 1 Introduction Tipping elements are important components of the Earth’s climate system because they may undergo abrupt transitions within a few decades, bringing them to a disrupted state [ 26 , 2 ] . We focus here on two subsystems: the Atlantic Meridional Overturning Circulation (AMOC) and the Amazon rainforest. The AMOC plays a major role in meridional heat transport, thus influencing the climate of the Northern Hemisphere and, more generally, of the entire planet. [ 41 ] was the first to suggest it may be prone to tipping by observing bistability in a conceptual model. Such behaviour has since then been confirmed in a hierarchy of models [ 43 ] , including fully coupled General Circulation Models (GCMs) [ 45 ] . This same paper has also highlighted a physics-based indicator that the AMOC is in a bistable regime, namely the freshwater it transports across its southern boundary. Available observations [ 5 , 17 ] of this indicator suggest that the AMOC is indeed in such a bistable regime. As for the Amazon rainforest, it is a crucial part of the global carbon cycle, an important reservoir of biodiversity, and cools the global climate through evapotranspiration [ 31 ] . It has also been shown that its tree cover may abruptly respond to changes in its hydrological conditions and could then exhibit multiple stable states [ 20 ] , including a (current) rainforest state, a savannah state and a treeless state. The Amazon rainforest may thus lie in a multistable regime, depending on its climatological and hydrological conditions. [ 15 ] has recently pinpointed several forcing factors that may bring the Amazon out of its rainforest state, including water stress, global warming and deforestation. Furthermore, separated regions of the forest are interconnected through atmospheric moisture flow [ 15 ] , which may increase their sensitivity to change [ 38 ] . Recent studies have shown that interconnected precipitation patterns within the forest act as positive feedback on the destabilization of the system and facilitate its abrupt transition [ 53 , 50 ] . However, such tipping events are very rare in simulations: they may occur only every few million years in a pre-industrial climate. As a consequence, to study them, one would have to simulate extremely long trajectories to sample at most a few of these events. This is not sufficient to analyse their statistical behaviour while being prohibitively expensive for large climate models. Rare-event algorithms were precisely developed to sample rare events much more efficiently by biasing ensembles of trajectories in a controlled way. There exist different algorithms in this family, which have been successfully applied to climate problems in recent years [ 36 , 42 , 35 , 23 , 10 ] . Here, we focus on an algorithm called Time Adaptive Multilevel Splitting (TAMS) [ 27 ] , which computes the probability that a system reaches a certain region of phase space before a time horizon. Although we have until now presented tipping elements separately, they do not evolve isolated. Therefore, tipping of one subsystem – called leading – might perturb the dynamics of another tipping element – called following, which may then tip in turn under this forcing. This phenomenon is called a tipping cascade [ 13 ] . This topic has gained considerable attention in recent years (see [ 50 ] for an extensive review) due to the global impact that such a cascade may have on the Earth system. [ 25 ] studied interactions between a network of five tipping elements, including the AMOC and the Amazon rainforest, and concluded that the effect of the AMOC on the Amazon rainforest is uncertain. An AMOC weakening would likely cause a southward shift in the InterTropical Convergence Zone (ITCZ) [ 21 , 3 , 4 ] , thus increasing precipitation over northern Brazil and the part of the Amazon rainforest located in the Southern Hemisphere. This also means that the Northern part of the Amazon rainforest may suffer from increased water stress. But as the southern part of the Amazon rainforest is the one mostly contributing to rainfall generation [ 39 ] , its strengthening may stabilize the rest of the system. Moreover, an AMOC weakening may shift the seasonal cycle, making the wet season dryer and the dry season wetter [ 34 , 4 ] , the effect of which change is still unclear. For these reasons, it is crucial to obtain quantitative understanding of the relationship between those systems by computing, for instance, the cascading tipping probability, or the probability that tipping of the leading system triggers a tipping of the following system. Numerous studies have tackled this issue already for the AMOC-Amazon coupled system [ 8 , 48 , 49 , 47 ] but they often have to rely on several assumptions. Indeed, the cost of coupled tipping studies often limits them to non-process based models [ 8 ] , where tipping elements are generally assumed to obey a simple double-fold bifurcation and to be linearly coupled [ 48 , 49 , 47 ] . Here, we show that, using TAMS, we are able to quantify the cascading tipping probability in a process-based coupled model. We can retrieve this probability by studying the tipping of the following system in the coupled model. Although TAMS has mostly been used to compute rare-event probabilities, it can effectively estimate a broader class of observables [ 7 , 12 ] . We use this flexibility to compute the conditional probability that any decrease in AMOC strength precedes a decrease in the Amazon mean tree cover, given that the latter occurs before a certain time. This directly quantifies the cascading tipping probability from the AMOC to the Amazon rainforest, and, more generally, the influence of the AMOC on the Amazon rainforest. In Section 2 , we describe the coupled conceptual model studied here. Then, we analyse in

Read the complete text