Large-scale modes of variability are present in the climate system. These modes are known to have influences on each other but are usually viewed as linear influences. The nonlinear connections among a set of key climate indices are explored here using tools from information theory, which allow us to characterize the causality between indices. It was found that quadratic nonlinear dependencies between climate indices are present at low frequencies, reflecting the complex nature of their dynamics.

An obstacle in studying climate extremes is the lack of robust statistics. We use a rare event algorithm to gather robust statistics on extreme Arctic sea ice lows with probabilities below 0.1 % and to study drivers of events with amplitudes larger than observed in 2012. The work highlights that the most extreme sea ice reductions result from the combined effects of preconditioning and weather variability, emphasizing the need for thoughtful ensemble design when turning to real applications.

Stratospheric aerosol injection (SAI) could be used alongside mitigation to reduce global warming. Previous studies suggest that more atmospheric CO₂ is taken up when SAI is deployed. Here, we look at the entire SAI deployment from start to after termination. We show how the initial CO₂ uptake benefit, and hence lower mitigation burden, is reduced in later stages of SAI, where the reduction in natural CO₂ uptake turns into an additional mitigation burden. 

Land cover and land management changes (LCLMCs) can alter climate even in intact areas, causing carbon changes in remote areas. This study is the first to assess these effects, finding they substantially alter global carbon dynamics, changing terrestrial stocks by up to dozens of gigatonnes. These results are vital for scientific and policy assessments, given the expected role of LCLMCs in achieving the Paris Agreement's goal to limit global warming below 1.5 °C.

Projecting future precipitation is essential for preparing for climate change, but current climate models still have large uncertainties, especially over land. This study presents a new method to improve precipitation projections by identifying which models best capture key climate patterns. By giving more weight to models that better represent these patterns, our approach leads to more reliable future precipitation projections over land.

We investigate the probabilities of triggering climate tipping points under various emission scenarios and how they are altered by additional carbon emissions from the tipping of the Amazon and permafrost. We find that there is a high risk for triggering climate tipping points under a scenario comparable to current policies. However, the additional warming and hence the additional risk of triggering other climate tipping points from the tipping of the Amazon and permafrost remain small.

We present a high-resolution paleogeographic reconstruction of the Baltic Sea for the Holocene period by combining eustatic sea-level change, glacio-isostatic movement, and sediment dynamics. In the northeastern part, morphological change is dominated by regression caused by post-glacial rebound that outpaces the eustatic sea level rise, whereas a transgression, together with active sediment erosion/deposition, constantly shapes the coastal morphology in the southeastern part.

Social tipping dynamics have received recent attention as a potential mechanism for effective climate actions – yet how such tipping dynamics could unfold remains largely unquantified. We explore how social tipping processes can develop by enabling necessary conditions (exemplified by climate change concern) and increased perceptions of localized impacts (sea level rise). The likelihood of social tipping varies regionally, mostly along areas with the highest exposure to persistent risks.

This work provides a relevant analysis of changes in extreme precipitation over the Caribbean and their link with warming in different ocean basins. It also improves our understanding of the impact of warming on extreme precipitation events, which can cause devastating damage to economic sectors such as agriculture, biodiversity, health, and energy.

This study investigated the computational benefits of using multiple models of varying cost and accuracy to quantify uncertainty in the mass change of Humboldt Glacier, Greenland, between 2007 and 2100 using a single climate change scenario. Despite some models being incapable of capturing the local features of the ice-flow fields, using multiple models reduced the error in the estimated statistics by over an order of magnitude when compared to an approach that only used a single accurate model.

AI is impacting science, providing key data insights, but most algorithms are statistical requiring cautious "out-of-sample" extrapolation. Yet climate research concerns predicting future climatic states. We consider a new method of AI-led equation discovery. Equations offer process interpretation and more robust predictions. We recommend this method for climate analysis, suggesting illustrative application to atmospheric convection, land–atmosphere CO₂ flux, and global ocean circulation models.