Recent advances in Earth Observation have focused on large-scale foundation models. However, these models are computationally expensive, limiting their accessibility and reuse for downstream tasks. In this work, we investigate compact architectures as a practical pathway toward smaller general-purpose EO models. We propose a Metadata-aware Mixture-of-Experts Masked Autoencoder (MoE-MAE) with only 2.5M parameters. The model combines sparse expert routing with geo-temporal conditioning, incorporating imagery alongside latitude/longitude and seasonal/daily cyclic encodings. We pretrain the MoE-MAE on the BigEarthNet-Landsat dataset and evaluate embeddings from its frozen encoder using linear probes. Despite its small size, the model competes with much larger architectures, demonstrating that metadata-aware pretraining improves transfer and label efficiency. To further assess generalization, we evaluate on the EuroSAT-Landsat dataset, which lacks explicit metadata, and still observe competitive performance compared to models with hundreds of millions of parameters. These results suggest that compact, metadata-aware MoE-MAEs are an efficient and scalable step toward future EO foundation models.

Through a series of experiments, we provide evidence that the GraphDOP model - trained solely on meteorological observations, using no prior knowledge - develops internal representations of the Earth System state, structure and dynamics as well as the characteristics of different observing systems. Firstly, we demonstrate that the network constructs a unified latent representation of the Earth System state which is common across different observation types. For example, cloud structures maintain physical consistency whether viewed in predictions for satellite radiances from different sensors, or for direct in-situ measurements of the cloud fraction. Secondly, we show examples that suggest that the network learns to emulate viewing effects - learned observation operators that map from the unified state representation to observed properties. Microwave sounder limb effects and geometric viewing effects, such as sunglint in visible imagery, are both well captured. Finally, we demonstrate that the model develops rich internal representations of the structure of meteorological systems and their dynamics. For instance, when the network is only provided with observations from a single infrared instrument, it is able to infer unobserved, non-local structures such as jet streams, surface pressure patterns and warm and cold air masses associated with synoptic systems. This work provides insights into how neural networks trained solely on observations of the Earth System spontaneously develop coherent internal representations of the physical world in order to meet the training objective - enhancing our understanding and guiding future development of these models.

Progress within physical oceanography has been concurrent with the increasing sophistication of tools available for its study. The incorporation of machine learning (ML) techniques offers exciting possibilities for advancing the capacity and speed of established methods and also for making substantial and serendipitous discoveries. Beyond vast amounts of complex data ubiquitous in many modern scientific fields, the study of the ocean poses a combination of unique challenges that ML can help address. The observational data available is largely spatially sparse, limited to the surface, and with few time series spanning more than a handful of decades. Important timescales span seconds to millennia, with strong scale interactions and numerical modelling efforts complicated by details such as coastlines. This review covers the current scientific insight offered by applying ML and points to where there is imminent potential. We cover the main three branches of the field: observations, theory, and numerical modelling. Highlighting both challenges and opportunities, we discuss both the historical context and salient ML tools. We focus on the use of ML in situ sampling and satellite observations, and the extent to which ML applications can advance theoretical oceanographic exploration, as well as aid numerical simulations. Applications that are also covered include model error and bias correction and current and potential use within data assimilation. While not without risk, there is great interest in the potential benefits of oceanographic ML applications; this review caters to this interest within the research community.

An exponential growth in computing power, which has brought more sophisticated and higher resolution simulations of the climate system, and an exponential increase in observations since the first weather satellite was put in orbit, are revolutionizing climate science. Big data and associated algorithms, coalesced under the field of Machine Learning (ML), offer the opportunity to study the physics of the climate system in ways, and with an amount of detail, infeasible few years ago. The inference provided by ML has allowed to ask causal questions and improve prediction skills beyond classical barriers. Furthermore, when paired with modeling experiments or robust research in model parameterizations, ML is accelerating computations, increasing accuracy and allowing for generating very large ensembles at a fraction of the cost. In light of the urgency imposed by climate change and the rapidly growing role of ML, we review its broader accomplishments in climate physics. Decades long standing problems in observational data reconstruction, representation of sub-grid scale phenomena and climate (and weather) prediction are being tackled with new and justified optimism. Ultimately, this review aims at providing a perspective on the benefits and major challenges of exploiting ML in studying complex systems.

Firedrake is a new tool for automating the numerical solution of partial differential equations. Firedrake adopts the domain-specific language for the finite element method of the FEniCS project, but with a pure Python runtime-only implementation centred on the composition of several existing and new abstractions for particular aspects of scientific computing. The result is a more complete separation of concerns which eases the incorporation of separate contributions from computer scientists, numerical analysts and application specialists. These contributions may add functionality, or improve performance. Firedrake benefits from automatically applying new optimisations. This includes factorising mixed function spaces, transforming and vectorising inner loops, and intrinsically supporting block matrix operations. Importantly, Firedrake presents a simple public API for escaping the UFL abstraction. This allows users to implement common operations that fall outside pure variational formulations, such as flux-limiters.

Despite continuous improvements, precipitation forecasts are still not as accurate and reliable as those of other meteorological variables. A major contributing factor to this is that several key processes affecting precipitation distribution and intensity occur below the resolved scale of global weather models. Generative adversarial networks (GANs) have been demonstrated by the computer vision community to be successful at super-resolution problems, i.e., learning to add fine-scale structure to coarse images. Leinonen et al. (2020) previously applied a GAN to produce ensembles of reconstructed high-resolution atmospheric fields, given coarsened input data. In this paper, we demonstrate this approach can be extended to the more challenging problem of increasing the accuracy and resolution of comparatively low-resolution input from a weather forecasting model, using high-resolution radar measurements as a "ground truth". The neural network must learn to add resolution and structure whilst accounting for non-negligible forecast error. We show that GANs and VAE-GANs can match the statistical properties of state-of-the-art pointwise post-processing methods whilst creating high-resolution, spatially coherent precipitation maps. Our model compares favourably to the best existing downscaling methods in both pixel-wise and pooled CRPS scores, power spectrum information and rank histograms (used to assess calibration). We test our models and show that they perform in a range of scenarios, including heavy rainfall.

Recent advances in Earth Observation have focused on large-scale foundation models. However, these models are computationally expensive, limiting their accessibility and reuse for downstream tasks. In this work, we investigate compact architectures as a practical pathway toward smaller general-purpose EO models. We propose a Metadata-aware Mixture-of-Experts Masked Autoencoder (MoE-MAE) with only 2.5M parameters. The model combines sparse expert routing with geo-temporal conditioning, incorporating imagery alongside latitude/longitude and seasonal/daily cyclic encodings. We pretrain the MoE-MAE on the BigEarthNet-Landsat dataset and evaluate embeddings from its frozen encoder using linear probes. Despite its small size, the model competes with much larger architectures, demonstrating that metadata-aware pretraining improves transfer and label efficiency. To further assess generalization, we evaluate on the EuroSAT-Landsat dataset, which lacks explicit metadata, and still observe competitive performance compared to models with hundreds of millions of parameters. These results suggest that compact, metadata-aware MoE-MAEs are an efficient and scalable step toward future EO foundation models.

Estimating background-error covariances remains a core challenge in variational data assimilation (DA). Operational systems typically approximate these covariances by transformations that separate geostrophically balanced components from unbalanced inertio-gravity modes - an approach well-suited for the midlatitudes but less applicable in the tropics, where different physical balances prevail. This study estimates background-error covariances in a reduced-dimension latent space learned by a neural-network autoencoder (AE). The AE was trained using 40 years of ERA5 reanalysis data, enabling it to capture flow-dependent atmospheric balances from a diverse set of weather states. We demonstrate that performing DA in the latent space yields analysis increments that preserve multivariate horizontal and vertical physical balances in both tropical and midlatitude atmosphere. Assimilating a single 500 hPa geopotential height observation in the midlatitudes produces increments consistent with geostrophic and thermal wind balance, while assimilating a total column water vapor observation with a positive departure in the nearly-saturated tropical atmosphere generates an increment resembling the tropical response to (latent) heat-induced perturbations. The resulting increments are localized and flow-dependent, and shaped by orography and land-sea contrasts. Forecasts initialized from these analyses exhibit realistic weather evolution, including the excitation of an eastward-propagating Kelvin wave in the tropics. Finally, we explore the transition from using synthetic ensembles and a climatology-based background error covariance matrix to an operational ensemble of data assimilations. Despite significant compression-induced variance loss in some variables, latent-space assimilation produces balanced, flow-dependent increments - highlighting its potential for ensemble-based latent-space 4D-Var.

Can we improve machine-learning (ML) emulators with synthetic data? If data are scarce or expensive to source and a physical model is available, statistically generated data may be useful for augmenting training sets cheaply. Here we explore the use of copula-based models for generating synthetically augmented datasets in weather and climate by testing the method on a toy physical model of downwelling longwave radiation and corresponding neural network emulator. Results show that for copula-augmented datasets, predictions are improved by up to 62 % for the mean absolute error (from 1.17 to 0.44 W m$^{-2}$).

High-performance object stores are an emerging technology which offers an alternative solution in the field of HPC storage, with potential to address long-standing scalability issues in traditional distributed POSIX file systems due to excessive consistency assurance and metadata prescriptiveness. In this paper we assess the performance of storing object-like data within a standard file system, where the configuration and access mechanisms have not been optimised for object access behaviour, and compare with and investigate the benefits of using an object storage system. Whilst this approach is not exploiting the file system in a standard way, this work allows us to investigate whether the underlying storage technology performance is more or less important than the software interface and infrastructure a file system or object store provides.

It is tested whether machine learning methods can be used for preconditioning to increase the performance of the linear solver -- the backbone of the semi-implicit, grid-point model approach for weather and climate models. Embedding the machine-learning method within the framework of a linear solver circumvents potential robustness issues that machine learning approaches are often criticized for, as the linear solver ensures that a sufficient, pre-set level of accuracy is reached. The approach does not require prior availability of a conventional preconditioner and is highly flexible regarding complexity and machine learning design choices. Several machine learning methods are used to learn the optimal preconditioner for a shallow-water model with semi-implicit timestepping that is conceptually similar to more complex atmosphere models. The machine-learning preconditioner is competitive with a conventional preconditioner and provides good results even if it is used outside of the dynamical range of the training dataset.

The NEMO general circulation ocean model is extended to incorporate three physical processes related to ocean surface waves, namely the surface stress (modified by growth and dissipation of the oceanic wave field), the turbulent kinetic energy flux from breaking waves, and the Stokes-Coriolis force. Experiments are done with NEMO in ocean-only (forced) mode and coupled to the ECMWF atmospheric and wave models. Ocean-only integrations are forced with fields from the ERA-Interim reanalysis. All three effects are noticeable in the extra-tropics, but the sea-state dependent turbulent kinetic energy flux yields by far the largest difference. This is partly because the control run has too vigorous deep mixing due to an empirical mixing term in NEMO. We investigate the relation between this ad hoc mixing and Langmuir turbulence and find that it is much more effective than the Langmuir parameterization used in NEMO. The biases in sea surface temperature as well as subsurface temperature are reduced, and the total ocean heat content exhibits a trend closer to that observed in a recent ocean reanalysis (ORAS4) when wave effects are included. Seasonal integrations of the coupled atmosphere-wave-ocean model consisting of NEMO, the wave model ECWAM and the atmospheric model of ECMWF similarly show that the sea surface temperature biases are greatly reduced when the mixing is controlled by the sea state and properly weighted by the thickness of the uppermost level of the ocean model. These wave-related physical processes were recently implemented in the operational coupled ensemble forecast system of ECMWF.

The Australian bushfires around the turn of the year 2020 generated an unprecedented perturbation of stratospheric composition, dynamical circulation and radiative balance. Here we show from satellite observations that the resulting planetary-scale blocking of solar radiation by the smoke is larger than any previously documented wildfires and of the same order as the radiative forcing produced by moderate volcanic eruptions. A striking effect of the solar heating of an intense smoke patch was the generation of a self-maintained anticyclonic vortex measuring 1000 km in diameter and featuring its own ozone hole. The highly stable vortex persisted in the stratosphere for over 13 weeks, travelled 66,000 km and lifted a confined bubble of smoke and moisture to 35 km altitude. Its evolution was tracked by several satellite-based sensors and was successfully resolved by the European Centre for Medium-Range Weather Forecasts operational system, primarily based on satellite data. Because wildfires are expected to increase in frequency and strength in a changing climate, we suggest that extraordinary events of this type may contribute significantly to the global stratospheric composition in the coming decades.