497 papers
Tracing the origin of tropical North Atlantic Sargassum blooms to West Africa
We simulate the dynamics of pelagic \emph{Sargassum} rafts as systems of finite-size floating particles, governed by a Maxey--Riley law with nonlinear elastic interactions. Using surface ocean currents and wind data from reanalysis systems for clump transport, we computed trajectories within a domain covering the tropical and subtropical north Atlantic. The subsequent motion is reduced using Ulam's discretization method into a time-inhomogeneous Markov chain that simulates a background \emph{Sargassum} concentration. Bayesian inversion, combined with nonautonomous transition path theory, was used to infer the origin of the first significant recorded bloom in the tropical North Atlantic, which unfolded in April 2011. Both methodologies independently identified the bloom's origin as near the West African coast, up to two years before it was detectable via satellite imagery on the basin's western side. This finding supports anecdotal evidence of \emph{Sargassum} strandings on the Ghanaian coast in 2009. Moreover, it correlates with unusual environmental conditions -- such as increased nutrient loads from significant upwelling linked to a pronounced Dakar Niña and Saharan dust deposition -- that promote bloom proliferation. Additionally, it aligns with the observation that the species of \emph{Sargassum} in the 2011 bloom differ from those in the Sargasso Sea, which might otherwise be considered a natural origin.
2604.04866Apr 2026
View
On the instability of some upward propagating, exact, nonlinear mountain waves
Using the short-wavelength instability method, we investigate the linear instability of an exact solution describing upward-propagating mountain waves, derived in A. Constantin, \emph{J. Phys. A: Math. Theor.} (2023), under the assumption of a dry adiabatic flow. Within this approach, the stability problem reduces to analysing a system of ordinary differential equations along fluid trajectories. Our results show that the flow becomes unstable when the wave steepness exceeds the critical threshold of $\frac{1}{3}$. Given the representation of the solution in Lagrangian coordinates, the instability analysis will show the existence of an unstable layer of few hundred meters beneath the tropopause where instability may occur, finally leading to a chaotic 3-dimensional fluid motion.
2604.03620Apr 2026
View
Physics-Constrained Adaptive Flow Matching for Climate Downscaling
Regional climate information at kilometer scales is essential for assessing the impacts of climate change, but generating it with global climate models is too expensive due to their high computational costs. Machine learning models offer a fast alternative, yet they often violate basic physical laws and degrade when applied to climates outside of their training distribution. We present Physics-Constrained Adaptive Flow Matching (PC-AFM), a generative downscaling model that addresses both problems. Building on the Adaptive Flow Matching (AFM) model of Fotiadis et al. (2025) as our baseline, we add soft conservation constraints that keep the downscaled output consistent with the large-scale input for precipitation and humidity, and use gradient surgery via the ConFIG algorithm to prevent these constraints from interfering with the generative objective. We train the model on Central Europe climate data, evaluate it on a 10-time downscaling task (63km to 6.3km) over six variables (near-surface temperature, precipitation, specific humidity, surface pressure, and horizontal wind components) across a comprehensive set of metrics including bias, ensemble skill scores, power spectra, and conservation error, and test the generalization on two held-out climate regions. Within the training distribution, PC-AFM reduces conservation errors and improves ensemble calibration while matching the baseline on standard skill metrics. Outside the training distribution, where unconstrained models develop large systematic errors by extrapolating learned statistics, PC-AFM halves precipitation wet bias, reduces conservation error and improves extreme-quantile accuracy, all without any information about the target climate at inference time. These results indicate that physical consistency is a practical requirement for deploying generative downscaling models in real-world applications.
2604.03459Apr 2026
View
High-resolution probabilistic estimation of three-dimensional regional ocean dynamics from sparse surface observations
The ocean interior regulates Earth's climate but remains sparsely observed due to limited in situ measurements, while satellite observations are restricted to the surface. We present a depth-aware generative framework for reconstructing high-resolution three-dimensional ocean states from extremely sparse surface data. Our approach employs a conditional denoising diffusion probabilistic model (DDPM) trained on sea surface height and temperature observations with up to 99.9 percent sparsity, without reliance on a background dynamical model. By incorporating continuous depth embeddings, the model learns a unified vertical representation of the ocean states and generalizes to previously unseen depths. Applied to the Gulf of Mexico, the framework accurately reconstructs subsurface temperature, salinity, and velocity fields across multiple depths. Evaluations using statistical metrics, spectral analysis, and heat transport diagnostics demonstrate recovery of both large-scale circulation and multiscale variability. These results establish generative diffusion models as a scalable approach for probabilistic ocean reconstruction in data-limited regimes, with implications for climate monitoring and forecasting.
2604.02850Apr 2026
View
MAG-Net: Physics-Aware Multi-Modal Fusion of Geostationary Satellite and Radar for Severe Convective Precipitation Nowcasting
Radar-based convective precipitation nowcasting suffers from rapid performance degradation beyond 30 minutes due to missing thermodynamic variables. Existing deep learning models also face blurring effects, training instability, and limited interpretability. To address this, we propose MAG-Net, a Physics-Aware Multi-modal Attention-guided Generator Network. It integrates radar dynamics with selected geostationary satellite channels (IR 10.8, WV 7.1, BTD) to incorporate thermodynamic and microphysical precursors. MAG-Net features a Dual-Stream Encoder for heterogeneous modalities and a Symmetric Dual-Head Decoder optimizing reflectivity regression and event probability via an uncertainty-weighted multi-task strategy. Furthermore, an inference-time Gradient-Preserving Fusion (GPF) strategy combines probabilistic constraints with regression details for better high-frequency texture retention. Experiments on a large-scale dataset (2018-2023) over southeastern China show MAG-Net outperforms deterministic (e.g., CPrecNet) and generative (e.g., DGMR) baselines. Specifically, it improves CSI40 by 0.083 (0.172 to 0.255) over CPrecNet, enhancing intense convective echo detection. Finally, Integrated Gradients (IG) analysis reveals the model's reliance on satellite inputs increases with forecast lead time and convective intensity, confirming that satellite data captures critical precursors for severe weather prediction.
2604.02818Apr 2026
View
On the White-Noise Limit of the Colored Linear Inverse Model
A recent paper by Lien et al. (2025) introduces the "colored linear inverse model" (colored LIM), in which stochastic forcing is modeled using Ornstein-Uhlenbeck colored noise rather than idealized white noise. In that work, it is shown that the derivative-based identification formulas used to estimate model parameters do not admit a regular white-noise limit due to the loss of differentiability of the lag-correlation function at zero lag. Here we revisit the white-noise limit from the perspective of the underlying stochastic differential equations. Treating the colored LIM as an augmented Ornstein-Uhlenbeck system, we show that as the correlation time tau -> 0 the colored-noise-driven system reduces to the classical LIM, and the corresponding stationary covariance satisfies the standard fluctuation-dissipation relation. Re-examining the same linear system used by Lien et al. (2025), we illustrate this convergence numerically. These results highlight a distinction between the singular behavior of derivative-based identification formulas and the regular limiting behavior of the underlying stochastic model. Taken together with recent results showing convergence of estimated parameters in the white-noise limit, they provide a consistent interpretation in which the colored LIM recovers the classical LIM at the level of stochastic dynamics even though certain estimation procedures become ill-defined in that limit.
2604.02519Apr 2026
View
2604.02283A proposal for the safety and controllability requirements that SRM systems should meet
Solar Radiation Modification (SRM) may be the only way to limit global warming in the coming decades, leading to increased interest in the subject and to the expansion of related research & development (R&D) activity. Defining the safety and controllability requirements that any SRM system should meet is crucial for directing R&D activities and enabling governments to make informed decisions on the development and possible implementation of such systems. We present an initial proposal for this set of requirements, which also guides Stardust's R&D, as a basis for further discussion and consideration. While we focus on SRM systems based on Stratospheric Aerosol Injection (SAI), the proposed principles may be applicable more broadly.
2604.02283Apr 2026
View
Assessing the ability of a stretched-grid deep-learning weather prediction model to capture physical balances
Weather forecasting has traditionally relied on Numerical Weather Prediction (NWP) models, which simulate weather by solving the governing fluid equations. Recently, the emergence of Deep Learning Weather Prediction (DLWP) models has opened a new era in weather forecasting, offering a data-driven alternative to classical NWP approaches. Regional DLWP models such as the stretched-grid model Bris developed by Met Norway, have demonstrated performance on par with, or even slightly better than regional NWP models across a range of standard forecast metrics. By overcoming the coarse horizontal resolution that constrained earlier global data-driven models, the operational use of regional DLWP systems now appears increasingly promising. Nevertheless, the performance of such models during extreme events is generally inferior to that of regional NWP models, and comprehensive evaluations of their ability to generate physically realistic forecasts are still lacking. Here, we present a study comparing the physical consistency of the deterministic version of Bris with the control run of the operational MetCoOp Ensemble Prediction System (MEPS) in forecasting the severe extratropical cyclone Poly, which hit the Netherlands on 5 July 2023. We examine whether Bris accurately represents deviations from key atmospheric balances and whether it reproduces expected dynamics of the storm. We show that, despite its relatively good performance in terms of RMSE, Bris struggles to capture important mesoscale features of the event and that it significantly disrupts several atmospheric balances. This unrealistic disruption is mainly linked to the fine-scale noise evidenced in its output fields, which leads to incorrect and unrealistic spatial gradients. These results raise critical questions for improving AI-based models to better represent extreme events and how to ensure physical consistency in their predictions.
2604.01454Apr 2026
View
Gray Swan Factory: Making Extreme Events from Ordinary Cyclones
Gray swans, plausible but unobserved extreme events, broaden our understanding of the range of hazards beyond those observed during the short observational record. They are useful for dynamical studies, synthetic training data, emergency planning, infrastructure design, and insurance hazard assessment. We propose a method to produce gray swans from the observational record using gradient descent on a loss function with a differentiable weather prediction model. Minimizing the loss corresponds to perturbed initial conditions that produce a measurable outcome at a future time, subject to constraints, such as the size of the initial perturbations. We illustrate the method by altering hurricane Fiona (2022), which tracked northward over the Atlantic Ocean, to produce a gray-swan outcome similar to hurricane Sandy (2012), which made landfall on the East Coast of the United States after a unique westward turn. The Fiona gray-swan solution, involving small perturbations to reanalysis initial conditions, produces an extratropical cyclone with a Sandy-like track, a warm core, and a minimum sea-level pressure more than 20 hPa lower than Sandy. Perturbations to the extratropical state are more important than to the hurricane, leading to interactive strengthening, and merger, of an upper-level trough and the hurricane. Similar gray swans are found for four other Atlantic hurricanes. A major weakness of this work is that the hurricane core is not resolved by the model used for optimization, and the impact of this is unknown. Furthermore, although these solutions present plausible outcomes, they do not inform on their probability of occurrence.
2604.00348Apr 2026
View
ACE2-NEMO: Coupling an ML atmospheric emulator to a full-depth dynamical ocean model
Understanding how fast atmospheric variability shapes slow climate variability and sensitivity remains a central challenge in Earth-system science. Recent advances in machine-learned (ML) atmospheric models have demonstrated remarkable skill on weather timescales, but their emergent behaviour in a fully coupled climate system remains largely unexplored. We present early results from a new hybrid modelling framework, in which the ACE2 ML atmospheric emulator is interactively coupled to the NEMO ocean model. We report on a set of 70-year coupled simulations (1950-2020 historical forcing and fixed-1950s control). These experiments represent, to our knowledge, the first multi-decadal integrations of a machine-learned atmosphere interacting with a full-depth dynamical ocean. Several historical and fixed-1950s control simulations from the fully dynamic global coupled climate model EC-Earth, which has the same ocean component used in ACE2-NEMO, are also considered for comparison. We assess the behaviour of the coupled system, with particular focus on low-frequency tropical variability and the climate response to greenhouse-gas forcing. Analysis of potentially emergent El Niño-like variability reveals realistic fast timescale air-sea coupling in the tropical Pacific, but the temporal variability is unrealistic, with very low amplitude oscillations; this appears to be due to weak atmospheric feedback in the tropical Pacific. The response to CO2 forcing shows initial agreement with EC-Earth3P, but deviates due to reduced downward short-wave radiation in ACE2. These results provide a unique test of physical realism for atmospheric emulators, and evaluate the possible role of entirely machine-learned components in next-generation Earth system models.
2603.28704Mar 2026
View
Remote Influences of Land Surface Temperature and their Implications for Sea Surface Temperature Patterns
The spatial pattern of sea surface temperature (SST) plays a central role in shaping the climate system, yet the influence of land surface temperature (LST) remains poorly understood. Using a state-of-the-art coupled ocean--land--atmosphere model, we examine the model's response to regional LST perturbations imposed through LST nudging and idealized time-dependent ramp warming simulations. We find that LST warming over South America strengthens the tropical Pacific zonal SST gradient, yielding a more La Niña--like mean state. Enhanced LST increases the zonal contrast in diabatic heating and excites stationary Rossby wave responses, which reinforce alongshore winds and coastal upwelling in the eastern Pacific. This provides a dynamical pathway linking regional land warming to changes in the equatorial Pacific mean state. Similar responses are found for warming over North America, accompanied by North Pacific cooling, and for warming over Central Africa, coupled with tropical Atlantic cooling. In contrast, warming over the Maritime Continent or the Tibetan Plateau does not induce significant SST pattern changes. Historical simulations nudged toward observed LST exhibit cooling in the tropical southeast Pacific, with the tentative implication that uncertainty in LST may contribute to model-simulated SST biases during the historical period.
2603.28481Mar 2026
View
Statistical Models of Ember Wash and Their Impact on Wildfire Area Growth
Wildfire spread is strongly influenced by the transport and ignition of embers. While long-range spotting driven by plume lofting has received significant attention, embers transported near the surface by turbulent winds can also influence fire propagation. We develop a stochastic model for near-surface ember transport, referred to as ember wash. The model represents ember motion as a sequence of short displacements analogous to saltation-like transport and incorporates a probabilistic ignition process that depends on ember survival during transport. This formulation leads to an exponential distribution of ember flight times. The model is implemented within a simplified fire spread model to examine burn patterns and growth dynamics. Simulations demonstrate that ember wash produces spread behavior that differs fundamentally from classical plume-driven spotting. These results suggest that ember wash provides a plausible mechanism for wildfire spread regimes that differ from those predicted by geometric or plume-driven spread models.
2603.28222Mar 2026
View
Downscaling weather forecasts from Low- to High-Resolution with Diffusion Models
We introduce a probabilistic diffusion-based method for global atmospheric downscaling implemented within the Anemoi framework. The approach transforms low-resolution ensemble forecasts into high-resolution ensembles by learning the conditional distribution of finer-scale residuals, defined as the difference between the high-resolution fields and the interpolated low-resolution inputs. The system is trained on reforecast pairs from ECMWF IFS, using coarse fields at 100 km to reconstruct fine-scale variability at 30 km resolution. The bulk of the training focuses on recovering small-scale structures, while fine-tuning in high-noise regimes enables the generation of extremes. Evaluation against the medium-range IFS ensemble target shows that the model increases probabilistic skill (FCRPS) for surface variables, reproduces target power spectra at small scales, captures physically consistent multivariate relationships such as wind-pressure coupling, and generates extreme values consistent with those of the target ensemble in tropical cyclones.
2604.03303Mar 2026
View
Calibrated Conformal Prediction Intervals for Microphysical Process Rates
Conformal prediction can yield statistically valid prediction intervals for any regression model, with no model modifications and small computational costs. To assess its practical value, we apply conformal methods to quantify uncertainty in machine learning emulators of six microphysical process rates. Microphysical process rates describe small-scale processes in atmospheric clouds such as precipitation formation and aerosol-cloud interactions, and help understand weather and climate. The emulators are trained on simulation output from the ICOsahedral Nonhydrostatic (ICON) model in a limited-area numerical weather prediction configuration. We compare split conformal prediction for deterministic emulators with conformalized quantile regression for quantile regression emulators. Both conformal prediction methods yield well-calibrated and sharp prediction intervals on average, but conformalized quantile regression provides more consistent intervals across several orders of magnitude, making it preferable for the uncertainty quantification of climate variables.
2603.27699Mar 2026
View
AIFS-COMPO: A Global Data-Driven Atmospheric Composition Forecasting System
We introduce AIFS-COMPO, a skilful medium-range data-driven global forecasting system for aerosols and reactive gases. Building on the ECMWF Artificial Intelligence Forecast System (AIFS), AIFS-COMPO employs a transformer-based encoder-processor-decoder architecture to jointly model meteorological and atmospheric composition variables. The model is trained on Copernicus Atmosphere Monitoring Service (CAMS) reanalysis, analysis, and forecast data to learn the coupled dynamics of weather, emissions, transport, and atmospheric chemistry. We evaluate AIFS-COMPO against a range of atmospheric composition observations and compare its performance with the operational CAMS global forecasting system IFS-COMPO. The results show that AIFS-COMPO achieves comparable or improved forecast skill for several key species while requiring only a fraction of the computational resources. Furthermore, the efficiency of the approach enables forecasts beyond the current operational horizon, demonstrating the potential of AI-based systems for fast and accurate global atmospheric composition prediction.
2604.03300Mar 2026
View
StretchCast: Global-Regional AI Weather Forecasting on Stretched Cubed-Sphere Mesh
Global AI weather forecasting still relies mainly on uniform-resolution models, making it hard to combine regional refinement, two-way regional-global coupling, and affordable training cost. We introduce StretchCast, a global-regional AI forecasting framework built on a variable-resolution stretched cubed-sphere (SCS) mesh that preserves a closed global domain while concentrating resolution over a target region. Within this framework, we develop a one-step predictor, SCS_Base Model, and a rollout-oriented multistep predictor, SCS_FCST4 Model, to test the feasibility of SCS-based forecasting and the benefit of joint multistep training. Experiments use ERA5 with 69 variables over 1998-2022. Because training compute remains limited, this study uses a coarse-resolution proof-of-concept configuration rather than a final high-resolution system. Even with only about 7,776 effective global grid cells and roughly 0.875 degree resolution over the center-refined face, the 23M-parameter SCS_Base Model yields stable multivariate forecasts. With 83M parameters and training cost on the order of hours, SCS_FCST4 Model delivers competitive medium-range anomaly-correlation evolution over the target region after unified reprojection, especially for geopotential height, specific humidity, and part of the lower-tropospheric winds, while maintaining smooth cross-face continuity and realistic multiscale structure in typhoon and spectral analyses. These results support StretchCast as a practical lightweight foundation for global-regional AI weather forecasting.
2603.27288Mar 2026
View
Stochastic coupling of climate variables and ice volume over the Late Pleistocene glacial cycles
Understanding the interactions between ice sheets and global climate forcings over geological timescales is essential for projecting their future. Previous studies have highlighted the role of ice dynamics and climate interactions in establishing the 100,000-year glacial cycles, particularly regarding the growth of the North American ice sheet. Researchers have reconstructed consistent time series for ice volume, temperature, and carbon dioxide by applying inverse forward modeling to benthic oxygen isotope records. Here we model the stochastic behavior of paleoclimate time series to evaluate the coupling between climate variables during the Pleistocene glacial cycles. We quantify the behavior of these time series using multifractal time-weighted detrended fluctuation analysis, which differentiates between near-red-noise and white-noise behavior below and above the 100,000-year glacial cycle, respectively, in all records. This study builds upon the work of Keyes et al. [Chaos vol. 33, 093132 (2023)] by incorporating ice volume into a five-variable model that includes carbon dioxide, methane, nitrous oxide, and temperature, along with intervariable coupling terms to capture potential relationships among these variables. Our analysis shows that ice volume, carbon dioxide, and temperature have a stabilizing effect upon each other. To test our model, we compute response functions for each pair of variables and compare these with empirical data, confirming our predictions regarding intervariable stability and coupling. This study provides a comprehensive overview of glacial-interglacial dynamics and highlights the role of cryosphere-climate feedbacks in shaping Earth's climate evolution.
2603.26937Mar 2026
View
Impact of geophysical fields on Deep Learning-based Lagrangian drift simulations
We assess the influence of different Eulerian geophysical input fields on Lagrangian drift simulations using DriftNet, a learning-based method designed to simulate Lagrangian drift on the sea surface. Two experiments are conducted: a fully numerical experiment (Benchmark B1) and a real-world drifters-based experiment (Benchmark B2). Both experiments are performed in two regions with different ocean dynamics: North East Pacific and Gulf Stream regions. The performance of DrifNet is evaluated with three different metrics: separation distance between simulated and ground-truth trajectories, the normalized cumulative Lagrangian separation and the autocorrelation of Lagrangian velocities. In both regions, results from B1 show that combining assimilated sea surface currents (SSC) with fully observed sea surface height (SSH) leads to greatest improvement in trajectory simulation. This configuration reduces separation distance by over 50\% and significantly decreases normalized cumulative Lagrangian separation and metrics related to velocities autocorrelation functions compared to the baseline using SSC alone. On the other hand, the inclusion of sea surface temperature (SST) either alone or in combination with SSC generally degrades performance. In B2, using satellite-derived SSH, Ekman and winds velocities improves surface drifters trajectories simulation, particularly in the North East Pacific. While the satellite-derived SST in combination with reanalysis-based SSC configuration leads to better trajectories simulation in the Gulf Stream. Overall, we highlight the added value of combining multiple geophysical fields to improve Lagrangian drift simulation on both numerical and real-world experiments.
2604.03292Mar 2026
View
Do Climate Models Need Microphysical and Convective Parameterizations to Generate Accurate Precipitation Fields?
Accurately representing surface precipitation is crucial for the operational use of weather and climate models. Presently, global numerical weather prediction (NWP) models struggle to accurately generate precipitation due to their parametrization of unresolved deep convective clouds and, in regions of grid-resolved ascent, inadequate parameterizations of cloud microphysics. Here we bypass these parameterizations with a machine learning model that diagnoses precipitation from 13 ERA5 fields that are easily observed and assimilated, as opposed for example, to fields like rain or cloud liquid water. We train a pair of models; ML_ERA5 using ERA5 precipitation as the target, and ML_IMERG using a satellite based precipitation product. ML_ERA5 closely reproduces the ERA5 precipitation at all intensities. When evaluated against the satellite dataset, ML_IMERG closely matches observations, notably reproducing the diurnal cycle of the satellite product. ML_IMERG generally captures extremes better than ERA5 while also reducing ERA5's overproduction of light precipitation. When evaluated against a third ground-and-radar-based dataset, ML_IMERG inherits the strengths of the satellite dataset which is superior to ERA5 in the summer months.
2603.25936Mar 2026
View
Toward Artificial Intelligence Enabled Earth System Coupling
Coupling constitutes a foundational mechanism in the Earth system, regulating the interconnected physical, chemical, and biological processes that link its spheres. This review examines how emerging artificial intelligence (AI) methods create new opportunities to enhance Earth system coupling and address long-standing limitations in multi-component models. Rather than surveying next-generation modelling efforts broadly, we focus specifically on how state-of-the-art AI techniques can strengthen cross-domain interactions, support more coherent multi-component representations, and enable progress toward unified Earth system frameworks. The scope extends beyond climate models to include any modelling system in which Earth spheres interact. We outline emerging opportunities, persistent limitations, and conceptual pathways through which AI may enhance physical consistency, interpretability, and integration across domains. In doing so, this review provides a structured foundation for understanding the role of AI in advancing coupled Earth system modelling.
2604.03289Mar 2026
ViewPage 1 of 25
