Operational avalanche forecasting has so far been done almost exclusively by human forecasters. For the first time, an automated machine learning approach allows to reach forecasting skills close to human forecasters.

This multi-model inter-comparison presents a state-of-the-art assessment of Latin America for the first time which is a region often ignored in air quality studies.

Coastal erosion is a process with widespread human impact. This well-written paper explains the NEWTS coastal erosion model in terms which will be accessible for geoscientists and beyond. It is recommended reading for anyone with an interest in this application area.

This paper is significant for both the geoscience community and the general public. For geoscientists, the novel HydroFATE model provides an innovative tool to estimate and track the presence of household and pharmaceutical contaminants in the world's river systems, aiding in global pollution studies and environmental planning. For the public and media, it highlights the widespread issue of water contamination from commonly used substances, illustrating their potential impacts on environmental and public health. HydroFATE can inform decision-making across sectors - from water testing prioritization by local governments to ecological considerations by pharmaceutical companies, making it a compelling narrative for the media. The antibiotic sulfamethoxazole's use as a test case further links this work to global health discussions on antibiotic resistance.

Robust outlier detection is a challenge for all areas of science that deal with real data. Here, the author describes a new approach to this in the field of geodesy, but does so in a readable and accessible way. It will therefore be valuable reading for those beyond that field.

This manuscript provides a novel and comprehensive assessment of uncertainty associated with sea level rise. The model description is thorough and it is applied to a number of possible scenarios. The conclusions are important for framing future discussions on sea level rise.

The integration of neural networks into PDE solvers to simulate systems for which the PDE models are incomplete is a key advance at the cutting edge of geoscientific modelling. The approach presented here is applicable far beyond the realm of ice modelling, and will be of interest to model developers and users across geoscience and beyond.

Grave wave (GW) parameterisations currently used in state-of-the-art weather and climate models are based on a purely columnar approach, which does not allow for any horizontal propagation of GWs and has been identified as potential source of systematic biases in the simulation of middle atmospheric dynamics. The study by Eichinger and colleagues presents now a computationally efficient method to emulate the effects of lateral propagation of orographic GWs in climate models by horizontal momentum flux redistribution using redistribution maps derived from a GW ray-tracing model. The presented approach is an important step towards a better representation of orographic GWs in climate models, which might improve long-standing problems in atmospheric modelling.

This paper investigates an impactful topic, is easily digestible to non-scientists, is well written, has nice visuals, uses novel objective identification methods and has well documented and accessible code.

Sea level rise represents one of the most compelling aspects of anthropogenic climate change. The potential social and economic impacts are enormous, with little that can be done to mitigate them. It is therefore of critical importance that we are able to correctly anticipate these impacts in advance. This study presents a new, open-source platform that integrates numerical modelling with socioeconomic and physical datasets, whilst also allowing for the uncertainty in climate change projections. This tool therefore allows for new and improved estimates of the global costs of future sea level rise and is likely to be of widespread interest.

Achieving both performance and portability in a whole dynamical core implemented in a high-productivity language such as Python is an eye-opening result which rebuts some widely held assumptions in the geoscientific modelling community. This is a paper which everyone who writes geoscientific models should read.

Many papers are currently being published applying deep learning to geoscientific applications. However, most of them only offer proof of concept results on highly idealised scenarios. This paper combines deep learning approaches with the structural causal models popularized by the work of Judea Pearl, and it applies this methodology to a real problem, analyzing soil moisture-precipitation coupling in climate reanalysis data. In contrast to many papers in this field, this promises actual insight in the scientific application of the work.

Having major Earth system model components make full use of new architectures is a critical step on the pathway to exascale climate simulation. This paper documents just this for the widely used WAVEWATCH III model.

Modelling the interactions of the atmosphere and cryosphere is essential to understanding our changing climate. This paper presents the coupling of the widely used WRF atmosphere to the SNOWPACK snow model. This work creates an important new tool for the modelling community.

Surface albedo is critical to understanding the Earth's energy balance. It emerges that snow grain shape and internal mixing of impurities are both important to getting this right, but until now these have not been incorporated into Earth System Models. This paper appears to be the first time that these effects have been included in an Earth System Model.

This paper provides key insight into the methodology of Working Group III of the IPCC Sixth Assessment Report on climate mitigation. The paper will be essential reading for anyone wishing to understand the assessment of climate outcomes of mitigation pathways in the context of the Paris Agreement.

Fire is a pervasive feature of the Earth system, and a cause of significant carbon emissions. This manuscript presents a higher resolution fire emissions data set than previously available, thereby providing a valuable resource to the scientific community.

Finding the horizon is commonplace for humans, and evocative when imagining journeys on the coast, in the mountains, or in endless plains. This paper shows a way to optimize a machine's ability to complete the same task, with the goal of bettering our ability to understand nature and climate.

This work removes the massive specialist-knowledge barrier to running hydrological models, making them usable by a much broader swath of scientific community -- and potentially beyond.

This paper introduces Firedrake, a new automatic system to generate code and solve partial differential equations using finite element methods. This capability is a core need of many models, and consequently a source of significant redundant software development effort. Because it does not prescribe a particular set of equations, the Firedrake software is applicable to a wide range of geoscientific models. Firedrake demonstrates remarkable computational efficiency, scaling beyond 12,000 computing cores. It is also free-libre open source software, contributing to improvements in scientific computational replicability and reproducibility.