This article provides a concrete example of how chaotic fluctuations in Greenland ice streams limit numerical predictability of ice sheet collapse. The paper identifies a limit to the horizon of predictability and attributes it to the physical structure of the ice sheet. For a fixed warming magnitude and an ensemble of warming rates and initial conditions, the timing of the collapse can differ by tens or hundreds of thousands of years.

This review is of interest to the whole Baltic Sea community and beyond as it can motivate future studies using the proposed framework to systematically investigate the impact of teleconnections across multiple time scales as well as analyzing both physical and biogeochemical processes hand in hand.

This study evaluates how Antarctic ice sheet surface mass balance has evolved over the past 125 years and its deep connections to ocean cooling, wind patterns, and dynamic ice loss. By integrating Earth System modeling with paleoclimate data, the study reconciles observed and modeled trends in snowfall, sea surface temperatures, and large-scale climate variability. The findings not only explain long-standing climate model discrepancies but also confirm an anthropogenic increase in Antarctic snowfall, with implications for global sea level rise and regional climate change.

This paper is a very timely contribution that can be expected to be of broad interest to the Earth System Modelling and Integrated Assessment Modelling community. It deals with the topic of how to accurately assess the Earth system responses to Carbon Dioxide Removal (CDR) methods. It proposes approaches to create a no-CDR baseline against which the Earth system effects of activity-driven CDR can be evaluated, as well as a framework to quantify the direct effect of CDR on atmospheric carbon dioxide as well as the resulting feedbacks. Such a framework is essential for assessing the efficiency of activity-driven CDR in emissions-driven simulations, such as those proposed for CMIP7.

The ocean component of the Earth System response to anthropogenic land cover change has previously received little attention. This paper demonstrates its potential importance, both for scientific understanding and for policies considering reforestation to mitigate climate change.

Groundwater depletion is among the most pressing global environmental challenges today, with a recent UN report warning of an "era of global water bankruptcy.” This study sheds light on how groundwater resources are likely to evolve through to 2100 under different climate and socioeconomic scenarios at hyper‑resolution (~1 km). All model simulations are openly accessible to support research and decision‑making in regions where observational data are limited.

The manuscript's findings of improved sea-ice cover accuracy and the quickness of future declines is of interest to a general audience. In addition, the complexity of the relationship between deep water formation and the over-all AMOC strength is an important aspect of the science that should be understood by a general audience when AMOC declines and its associated climate impact is discussed.

This paper finds that hysteresis and partial irreversibility in permafrost thawing result in a potentially large increase in cumulative carbon emissions from permafrost. This highlights the risks that come with overshoot scenarios, and the processes that may counteract the expected decrease in temperatures in the later part of such scenarios.

This paper provides important insights on possible abrupt shifts and state transitions under future climate change. The comprehensive catalogue presented in the paper spans changes in multiple components of the Earth system. These can have major impacts on both the climate system and other dependent systems, from ecosystems to society and the economy. The paper also makes the case for the possibility of Strong Nonlinear Surprises happening earlier than previously anticipated.

This paper presents data-driven equation discovery as a powerful scientific method to study climate variability, and thereby exemplifies how machine-learning methods do not only work as black boxes, but can also help us better understand complex Earth system dynamics.