Peer Revenue Publish or profit

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Physics Physics Atmospheric and Oceanic Physics

physics.ao-ph Watch Atmospheric and Oceanic Physics

Thursday 27 August

Atmospheric and Oceanic Physics

When is logistic error growth a spectral reduction?

Malaquias Peña

Forecast-error growth has been studied through complementary approaches: scalar models describe the evolution of bulk error, whereas spectral theories follow decorrelation and error transfer across scales. Their connection matters when a fitted growth curve is used to infer a physical mechanism or a limit of predictability. We establish a conditional link between these approaches. Within a bilinear spectral exchange model, assume a stationary reference spectrum and a remainder that vanishes on the proposed common-shape evolution. For a common initial decorrelation fraction strictly between zero and one, the spectral shape is preserved and its amplitude grows logistically exactly when every participating scale shares one positive growth coefficient. The converse inference fails: a logistic aggregate need not imply proportional decorrelation or a common coefficient. An exact two-shell construction demonstrates this failure with identical aggregate curves and different spectral dynamics. For heterogeneous amplitudes, an exact averaging identity separates the contributions of growth-rate mismatch, shape variance, rate–shape covariance, evolving spectral weights, and residual tendencies. A finite-time bound relates their combined effect to departure from a reference logistic curve. Forced surface-quasigeostrophic simulations illustrate how similar retrospective agreement with a logistic reference can coexist with heterogeneous effective shell rates and substantial instantaneous tendency discrepancies. Thus a scalar growth curve alone cannot identify how uncertainty is distributed across scales or justify interpreting a fitted rate as a common spectral growth coefficient. The framework specifies the additional spectral and budget evidence needed to give scalar predictability models that interpretation.
Read the full abstract

Wednesday 26 August

Atmospheric and Oceanic Physics · Cross-listed from physics.plasm-ph

Apokamp-type Gas Discharge Phenomenon: Experimental and Theoretical Backgrounds

Vasily Yu. Kozhevnikov, Andrey V. Kozyrev, Aleksandr O. Kokovin et al. (+5 more)

The apokamp discharge is an atmospheric-pressure plasma jet generated at the bending point of the discharge channel between high-voltage and floating-potential electrodes. The jet propagates perpendicularly to the interelectrode discharge channel without convection. In ambient air, the apokamp length reaches several centimeters, and the temperature of its tip is 100-250 {\deg}C. The typical apokamp propagation speed ranges from 100 to 220 km/s. The proposed time-dependent theoretical model explains the phenomenon. The discharge modeling confirms the tendencies observed in various experiments on the problem.
Read the full abstract

All authors Vasily Yu. Kozhevnikov, Andrey V. Kozyrev, Aleksandr O. Kokovin, Aleksey G. Sitnikov, Eduard A. Sosnin, Victor A. Panarin, Victor S. Skakun, Victor F. Tarasenko

Atmospheric and Oceanic Physics · Cross-listed from gr-qc

Atmospheric Newtonian noise in torsion-balance measurements of the gravitational constant $G$

Jyotirmaya Mohanta and Yutaka Shikano

Measurements of Newton's gravitational constant remain limited by environmental and apparatus-dependent systematics whose treatment is often less explicit than that of instrumental noise. Among these, atmospheric density fluctuations generate unshieldable gravity gradients that couple directly to torsion-balance observables as atmospheric Newtonian noise. Here, we develop a GUM-consistent framework for propagating this contribution through the torque estimator into the uncertainty budget of torsion-balance measurements. We derive closed-form spatial transfer functions for two benchmark geometries: a two-mass dumbbell, which provides an upper-coupling reference, and a perfectly symmetric cross, which serves as an idealized rejection limit and exposes the trade-off between suppressing low-order environmental coupling and preserving signal response. We also separate stationary correlated inputs from non-stationary baseline drift, using the Ornstein–Uhlenbeck process only as a benchmark for the former. Atmospheric pressure benchmarks indicate that the resulting background contribution is below present reference uncertainty levels, but can become relevant as systematic floors approach the part-per-million regime. This framework provides a practical route for incorporating site-specific environmental gravity gradients into future torsion-balance uncertainty budgets.
Read the full abstract

Atmospheric and Oceanic Physics

A place for stabilization alongside tipping cascades: the AMOC-cryosphere system

Sacha Sinet

The Atlantic Meridional Overturning Circulation (AMOC) and cryosphere are central to tipping-point research, where their interactions are typically framed as pathways for cascading transitions. Here, I review evidence that these interactions can also oppose change, affecting its magnitude and timing or, in some cases, whether a transition occurs at all. I argue that a more complete assessment of tipping risk requires progress on three fronts: systematic attention to stabilizing alongside destabilizing dynamics, conceptual frameworks that accommodate competing and context-dependent effects, and methods to detect, anticipate, and attribute stabilization.
Read the full abstract

Atmospheric and Oceanic Physics · Cross-listed from cs.CV

Precipitation Downscaling Using Foundation Model-Conditioned Diffusion

Victor Nascimento Ribeiro, Jorge Guevara, Jorge Sebastian Moraga et al. (+7 more)

High-resolution precipitation fields are essential for hydrological impact assessment, yet global climate model outputs are too coarse and biased for direct use. AI-based statistical downscaling with diffusion models offers a promising approach, but the mechanism by which large-scale atmospheric predictors condition generation remains largely unexplored. We investigate three conditioning strategies for a denoising diffusion probabilistic model applied to daily precipitation downscaling: channel concatenation of upsampled coarse predictors, cross-attention conditioning with a learned convolutional encoder, and cross-attention conditioning with the frozen encoder of the pretrained Prithvi WxC weather foundation model. All strategies are evaluated against an unconditioned baseline under identical conditions using probabilistic, distributional, spectral, and extreme-event metrics for the Colorado River Basin. Concatenation conditioning achieves the lowest point-wise CRPS and MSE, but tends to produce over-smoothed fields that suppress high-intensity events. In contrast, cross-attention conditioning provides substantially better distributional realism and modest improvements in spectral fidelity. Improvements are greatest for extremes: the Prithvi-WxC conditioned model retains over half of >100mm/day events, although estimates are uncertain due to limited samples. When trained on the full dataset, the learned convolutional model performs similarly to the foundation model-conditioned approach while requiring lower computational resources. However, the Prithvi-WxC-conditioned model achieves comparable performance with only five years of training data. These results indicate that cross-attention conditioning offers advantages over simple concatenation for probabilistic precipitation downscaling, and that pre-trained foundation model representations may offer benefits in data-limited settings.
Read the full abstract

All authors Victor Nascimento Ribeiro, Jorge Guevara, Jorge Sebastian Moraga, Chris Lucas, Natalie Lord, Andrew Taylor, Edward Lockhart, Will Trojak, Johannes Schmude, Anne Jones

Atmospheric and Oceanic Physics

Missing the Butterfly and Predicting the Past: Features or Bugs of Accurate AI Weather Models?

Pedram Hassanzadeh, Weidong Li, Y. Qiang Sun et al. (+4 more)

AI weather prediction (AIWP) models rival physics-based models, yet the sources of their unexpected forecast accuracy and the degree of their physical fidelity remain unclear. Here, across a hierarchy spanning observation-based reanalysis, a general circulation model, and the multi-scale Lorenz system, we show that AI models can be trained to skillfully predict the past (backcast), though backcasts are systematically less accurate than forecasts. However, skillful backcasting appears to violate the second law of thermodynamics, and all these forecasting and backcasting models miss the butterfly effect. We trace the surprising forecast accuracy, missing butterfly, and skillful backcasting to a single cause: inevitable coarse-graining of training data, which removes fast, small scales and/or some variables. From the Lorenz system to official Pangu-Weather models, reducing coarse-graining makes AI predictions more physics-like (arrow of time and butterfly-like effects emerge), but forecast accuracy declines. Results offer an explanation for AIWP models' forecast skill: unlike physics-based models, they implicitly learn how fast, small scales affect large scales without inheriting their rapid error growth. Broader implications are that AI models' proliferation calls for revisiting predictability theories and long-term climate emulation strategies, and backcasting offers a useful, new lens for such analyses.
Read the full abstract

All authors Pedram Hassanzadeh, Weidong Li, Y. Qiang Sun, Jiangdi Wang, Alexander Wikner, Justin Finkel, Jonathan Q. Weare

Atmospheric and Oceanic Physics · Cross-listed from cs.LG

Tropospheric temperature and humidity profile retrieval from Meteosat Flexible Combined Imager based on deep learning

Alejandro Salgueiro, Johannes Rausch, Julie Thérèse Villinger et al. (+1 more)

The Meteosat Third Generation (MTG) Flexible Combined Imager (FCI) offers new opportunities for tropospheric temperature and humidity profiling, at higher spatio-temporal resolutions and expanded spectral coverage relative to its predecessor. Vertically resolved retrievals from broadband imagers are inherently challenging, and operational retrieval algorithms typically rely on numerical weather prediction (NWP) background fields to compensate for limited infrared spectral resolution, reducing the retrievals' independence. We develop a spatially aware deep learning framework to retrieve all-sky tropospheric temperature and humidity profiles from FCI, without forecast profiles as input. A Residual U-Net that exploits spatial context across all 16 FCI channels was trained on 14 months of collocated FCI observations and CERRA reanalysis targets over Europe. Validated against independent radiosondes, retrieved temperatures show biases below 0.4 K and standard deviations of 1.5-1.9 K. Retrieved relative humidity standard deviations range from 12-20 %, compared to 9-19 % for CERRA. Performance degrades modestly under clouds, with standard deviation increases below 0.4 K and 3 % RH beneath cloud tops despite limited direct radiative information. Ablation experiments show that spatial context improves retrievals, with the largest gains below cloud tops. Feature sensitivity analysis indicates broad consistency with FCI bands' established radiative transfer characteristics. Visible and near-infrared channels contribute despite not being commonly used in physics-based profile inversions. These results demonstrate that spatially aware deep learning models can extract statistically reliable tropospheric profiles from geostationary imager observations, independent of NWP forecast fields, enabling more rapid autonomous monitoring of the atmosphere.
Read the full abstract

All authors Alejandro Salgueiro, Johannes Rausch, Julie Thérèse Villinger, Angela Meyer

Tuesday 25 August

Atmospheric and Oceanic Physics

A Station-Based Evaluation of Machine Learning-based Weather Forecasting Models in Northern Norway

Siyan Chen, Lars Uebbing, Eirik Mikal Samuelsen et al. (+5 more)

Recent machine learning weather prediction (MLWP) models have demonstrated remarkable forecasting skill on global reanalysis-based benchmarks. However, their performance remains unclear in challenging environments such as Northern Norway, where narrow fjords and rapidly changing weather result in highly variable local wind conditions. In this case study, we evaluate FourCastNet3 (FCN3), GraphCast, and ECMWF High Resolution Forecast (HRES) for wind speed forecasting using multi-year station observations from Northern Norway, focusing on their relative performance, generalization beyond the training period, and performance under high-wind conditions. Our results show that HRES slightly outperforms FCN3 and GraphCast, with an overall RMSE of 2.89 $\mathrm{m\,s^{-1}}$, compared to 2.96 $\mathrm{m\,s^{-1}}$ for FCN3 and 2.94 $\mathrm{m\,s^{-1}}$ for GraphCast. Notably, the MLWP models maintain comparable performance beyond their respective training periods, with no clear evidence of noticeable degradation. FCN3 performs best under high-wind conditions, although all models substantially underestimate strong winds. Our findings suggest that MLWP has become competitive with NWP for local wind, but further refinements are still needed to capture complex terrain better.
Read the full abstract

All authors Siyan Chen, Lars Uebbing, Eirik Mikal Samuelsen, Georgios Leontidis, Arnt-Børre Salberg, Sébastien Lefèvre, Robert Jenssen, Kristoffer Wickstrøm

Atmospheric and Oceanic Physics · Cross-listed from cs.LG

AFDBench: A Reasoning-First AI Scientist for NationalWeather Service Forecast Discussions

Manmeet Singh, Somnath Luitel, Prabhjot Singh et al. (+3 more)

Large language models (LLMs) hallucinate numerical values when generating high-stakes meteorological text, posing risks for weather communication. We present AFDBench, an AI meteorologist that generates professional Area Forecast Discussions (AFDs) by reasoning through structured AI weather forecast data from Google's WeatherNext 2. We introduce AFDBench, the first benchmark for evaluating generative meteorological reasoning, comprising 7,732 expert written discussions from 13 National Weather Service (NWS) offices paired with real AI weather forecast inputs, and three complementary metrics: Met-Align (numerical accuracy), Style-Align (professional dialect adherence), and Input-Grounding (fidelity to source weather data). Zero-shot evaluations reveal that open-source LLMs achieve low Style-Align (~0.33) and moderate Input-Grounding (~0.88), failing to write in the professional NWS register or faithfully use their input data. We apply Group Relative Policy Optimization (GRPO) with domain-specific rewards targeting temperature accuracy, synoptic correctness, and format compliance. On 1,033 held-out samples from two unseen NWS offices, GRPO nearly doubles Style-Align from 0.318 to 0.619 and improves Input-Grounding from 0.881 to 0.940, demonstrating that reinforcement learning teaches a 7B-parameter model to write like a professional meteorologist and faithfully interpret AI weather data.
Read the full abstract

All authors Manmeet Singh, Somnath Luitel, Prabhjot Singh, Manraaj Banga, Naveen Sudharsan, Josh Durkee

Atmospheric and Oceanic Physics

AICON: An operational global machine learning weather forecasting model

Tobias Goecke, Marek Jacob, Florian Prill et al. (+9 more)

We introduce AICON, a global machine learning weather prediction (MLWP) model which generates forecasts at 13 km spatial resolution with a 3-hour time step, trained on the high-resolution, non-hydrostatic ICON-DREAM dataset. AICON is in full operational use at Deutscher Wetterdienst since 2nd of March 2026. The model employs a graph neural network (GNN) architecture with an encoder-processor-decoder structure, where node updates are performed using a graph attention mechanism. A key feature of AICON is its use of an icosahedral multi-mesh derived from the native grid of the ICON model, ensuring consistency with the training data. ICON's terrain-following vertical SLEVE coordinate is one of the major distinctions from existing emulators. AICON's training strategy prioritizes small-scale fidelity by avoiding autoregressive multi-step rollout and longer forecast horizons during training, a design choice motivated by the hypothesis that this approach preserves fine-scale features often damped in models optimized for longer-range forecasts. We describe the prognostic and diagnostic variables used for training, the transfer learning protocol employed to accelerate convergence, and the model's performance across a range of evaluation metrics. An extensive evaluation, including routine verification against observation, a tropical cyclone case and spectral analysis reveal the strengths and limitations in the representation of atmospheric variability across scales. Routine verification against observations demonstrates competitive skill relative to the operational ICON model, particularly for near-surface variables in the short to medium forecast range.
Read the full abstract

All authors Tobias Goecke, Marek Jacob, Florian Prill, Michael Denhard, Felix Fundel, Jan Keller, Roland Potthast, Hendrik Reich, Britta Seegebrecht, Sven Ulbrich, Arianna Valmassoi, Sabrina Wahl

Atmospheric and Oceanic Physics

Predictability of El Niño from Delayed Observations

Francisco J. Beron-Vera

Using monthly Niño-3.4 anomalies through July 2026, we investigate how much predictive information is contained in delayed observations of the index. Ridge regression identifies informative delays, while multilayer perceptron and sparse identification of nonlinear dynamics (SINDy) models test whether nonlinear complexity provides additional direct forecast skill; gated recurrent unit (GRU) and long short-term memory (LSTM) networks provide a complementary test in which the temporal representation is learned internally. Delayed observations substantially improve forecasts over persistence and climatology at leads of up to six months, but increasing model complexity provides no systematic improvement. Historical recursive experiments favor a simple explicit SINDy recurrence and select shallow recurrent architectures, with no appreciable gain from learning the temporal representation internally. These results support a compact predictive representation of Niño-3.4 evolution in which the representation of past information is more consequential than model complexity. As a prospective application, the selected models are used to forecast the developing 2026 event beyond the last available observation and to compare its predicted evolution with completed historical El Niño events.
Read the full abstract

Monday 24 August

Atmospheric and Oceanic Physics

Persistent memory and tail-risk amplification in Atlantic Meridional Overturning Circulation variability: a Volterra integral framework calibrated with CMIP6

Mauricio Herrera-Marín

The Atlantic Meridional Overturning Circulation (AMOC) carries multi-decadal memory that ensemble-mean risk projections do not capture. We quantify this memory and its consequences for tail risk with a Volterra integral framework applied to ensemble-mean-subtracted variability in 14 CMIP6 models (1850-2100, three SSP scenarios), so that results reflect intrinsic thermohaline memory rather than the common anthropogenic forcing trend. Four results follow from leave-one-out cross-validation and an annealed-quenched tail decomposition. First, the ensemble-mean DFA1 Hurst exponent is H = 0.781 +/- 0.219, with 12 of 14 models showing long-range dependence (H > 0.5), consistent with thermohaline adjustment timescales near 33 yr. Second, a first-order Volterra model reduces out-of-sample RMSE by 16.8% over the best autoregressive baseline and by 12.3% over an unconstrained 20-lag distributed baseline, both robust to 200 circular-shift placebos (p < 0.05): the physically motivated kernel shape carries genuine predictive advantage. Third, the rolling 30-year lower-tail frequency rises 1.9-3.1x above the historical baseline depending on scenario, and the memory amplification index exceeds 1 in 8 of 14 models under SSP5-8.5 (median 1.15): persistence-driven clustering of weak-AMOC states amplifies tail frequency beyond forcing-only projections. Fourth, model-specific optimal memory horizons (15-52 yr) correlate with thermohaline regime (r = 0.74, p < 0.01), and the ensemble-mean Hurst exponent already exceeds a detection threshold H* = 0.70 in 9 of 14 models, with theoretical lead times to near-tipping conditions of 10-35 yr depending on scenario. These results support trajectory-specific, memory-aware AMOC risk assessment under moderate-to-high forcing and a physically grounded, scenario-conditional early-warning framework.
Read the full abstract

Atmospheric and Oceanic Physics

Statistics of Temperature Extremes and Implications for Electrical Power Infrastructure in Continental France

Peter Werner, Bastien Cozian, Yoann Robin et al. (+2 more)

In this article, we present extreme value statistics of temperature extremes over continental France with a focus on their implications for electrical power infrastructure. These are obtained by fitting a non-stationary generalised extreme value distribution using a Bayesian setup, which also provides errors or uncertainty values for all estimates. Within this method, a combination of simulated data from a collection of 28 CMIP6 models and measured records from the E-OBS dataset is used. The investigated climate scenarios are SSP2-4.5, SSP3-7.0 and SSP5-8.5, considering both historical and future climates spanning the years from 1850 to 2099. The method provides full spatial resolution on a 0.25 degree grid, allowing to asses extreme temperatures at arbitrary locations. Leveraging this particular aspect, various maps revealing the spatial structure of annual maximum high temperature extremes over continental France for the median and the statistical upper bound are shown together with summary information for the administrative regions in France towards the end of the century. It is statistically possible that the south west region of Occitanie could reach temperatures of up to 57{\deg}C under a high emission scenario by 2080, which is the highest compared to all other regions in continental France. We also perform a comparison to the reference climate adaptation trajectory for France (TRACC - Trajectoire de réchauffement de Référence pour l'Adaptation au Changement Climatique), showing that it potentially underestimates the stated maximum temperatures which could be surpassed by up to +8{\deg}C. Furthermore, five reference electrical power infrastructure locations are investigated on how they are potentially affected by temperature extremes with the quantified intensities.
Read the full abstract

All authors Peter Werner, Bastien Cozian, Yoann Robin, Laurent Dubus, Freddy Bouchet

Sunday 23 August

Atmospheric and Oceanic Physics

Elveslocation: A new approach for studying thunderstorm atmosphere

S. A. Sharakin and R.E. Saraev

Thunderstorm electrical discharges, which vary in their type of initiation and development (including compact intracloud discharges and initial breakdown pulses), also manifest differently as recorded electromagnetic signals in both the radio and optical ranges. A unique ionospheric optical imprint of a discharge is an ELVES - a sub-millisecond glow at altitudes around 90 km with a characteristic spatiotemporal pattern in the form of an expanding ring, reaching a diameter of several hundred or even a thousand kilometers. The position of the elve, its expansion velocity, the azimuthal and radial distribution of the glow intensity (and their temporal dynamics) contain information about the location and orientation of the discharge, as well as its current function. This information is most fully represented in the data recorded by orbital detectors of dynamic images with a wide field of view and microsecond temporal resolution, such as the TUS and Mini-EUSO. Recovering the parameters of a discharge from orbital detector data is associated with solving a complex inverse problem, since the phenomenon of interest is mediated by several processes, including the interaction of the discharge's electromagnetic pulse with the ionosphere and the non-trivial instrument response function of the detector. These additional sources of uncertainty are most consistently accounted for within the Bayesian paradigm. The application of simulation-based inference (SBI) allows for the reconstruction of discharge parameters within a dynamic model, thereby taking a step towards creating a new method for studying the thunderstorm atmosphere – elveslocation. This work presents a dynamic model of an elve, which enables both the identification of characteristic spatiotemporal patterns of an event (generator mode) and the reconstruction of discharge parameters (SBI simulator mode).
Read the full abstract

Atmospheric and Oceanic Physics · Cross-listed from cond-mat.soft

A Reduced-Order Dynamical Model for the Ignition of Diver-Induced Cohesive Silt-Out on Sloping Beds

Sandy Hardian Susanto Herho, Iwan Pramesti Anwar, Umar Abdurrahman et al. (+4 more)

A downward fin thrust near a cohesive seabed lofts sediment. Whether this cloud settles or organizes into a self-sustaining down-slope current determines if a diver loses visibility briefly or catastrophically. We reduce the layer-averaged balances of fluid mass, sediment mass, and momentum to a temporal slab of fixed thickness in the weakly entraining limit. This yields an autonomous planar vector field in dimensionless near-bed speed and suspended load, governed by three dimensionless groups: drag against settling, erosion strength, and near-bed concentration. The field possesses a quiescent equilibrium at the origin and, above a critical bed slope obtained in closed form as the ratio of near-bed concentration times drag-settling number to erosion-strength number, an interior saddle equilibrium. The saddle's stable manifold partitions the state space into basins of decay and unbounded growth, establishing the exact threshold between self-limiting and igniting disturbances. On a flat bed, the origin attracts globally for all admissible erosion laws. For representative silt, the ignition threshold is near a 17-degree bed angle. The closed-form structure is confirmed numerically to machine precision. A closure study demonstrates the critical slope is robust in existence but closure-dependent in value, identifying the linear erosion law as the conservative choice. The model thus isolates bed slope as the controlling parameter for ignition.
Read the full abstract

All authors Sandy Hardian Susanto Herho, Iwan Pramesti Anwar, Umar Abdurrahman, Faruq Khadami, Alfita Puspa Handayani, Karina Aprilia Sujatmiko, Dasapta Erwin Irawan

Atmospheric and Oceanic Physics

A score-based particle flow filter for non-Gaussian data assimilation in high-dimensional chaotic systems

Zheqi Shen, Youmin Tang and Yuewei Fang

Current particle flow filters rely on Gaussian prior assumptions that fail to capture the non-Gaussian attractor structure of chaotic systems. This study proposes a Score-based Particle Flow Filter (Score-PFF) that replaces the parametric prior gradient with a neural network-learned score function via denoising score matching. This enables flexible characterization of multimodal, skewed, and complex prior distributions in chaotic dynamics. Pure prior adjustment experiments demonstrate correct gradient directions toward the attractor (26.9%-29.0% error reduction over Gaussian priors). Under linear observations, Score-PFF significantly outperforms both Gaussian PFF and EAKF (Cohen's d = 1.03 and 1.40), preserving non-Gaussian structure that EAKF progressively Gaussianizes. Under nonlinear observation operators, Score-PFF maintains robust performance with up to 60% RMSE reduction in strongly non-Gaussian regimes. On the 1000-dimensional Lorenz-96 system, Score-PFF achieves 49.5% RMSE reduction over PFF while reducing per-assimilation cost by replacing SVD-based covariance inversion with neural network inference. Score-PFF establishes a computationally tractable, non-Gaussian data assimilation framework suitable for high-dimensional geophysical systems.
Read the full abstract

Atmospheric and Oceanic Physics

Quantifying AI data center nitrogen oxide (NO$_x$) emissions from space

Kevin D. Gauld, Daniel J. Varon, Nicholas Balasus et al. (+1 more)

AI data center power demand is spurring rapid deployment of on- and near-site natural gas turbines. Nitrogen oxide (NO$_x$) pollution from this equipment is a growing concern but has not previously been quantified with atmospheric observations. Here we demonstrate space-based detection and quantification of NO$_x$ emissions from the SpaceXAI Colossus 2 power plant in Southaven, Mississippi. Using observations from the geostationary TEMPO satellite instrument, we detect a strong increase in local mean NO$_2$ column concentrations after the plant began operations in late 2025. We then use TEMPO to estimate two-week-average NO$_x$ source rates from August 2025 to mid-August 2026, calibrating against continuous emission monitoring system (CEMS) data from US power plants. TEMPO first detected NO$_x$ emissions in December 2025 at 460$\pm$180 kg h$^{-1}$. We find that emissions increased through August 2026, averaging 730$\pm$185 kg h$^{-1}$ after February 2026, roughly 16 times higher than expected from the facility's March 2026 permit for 41 turbines operating under best available control technology (BACT) requirements ($\sim$47 kg h$^{-1}$). Emissions at the expected level would be undetectable by our TEMPO analysis.
Read the full abstract

All authors Kevin D. Gauld, Daniel J. Varon, Nicholas Balasus, Daniel H. Cusworth

Saturday 22 August

Atmospheric and Oceanic Physics

Structured Neural Modeling of Daily Arctic Sea-Ice Concentration Evolution: Physical-Trajectory-Driven Learning and Forecast-Domain Adaptation

Maqun Zhang, Feng Gao, Wankun Chen et al. (+3 more)

Accurate modeling of the daily evolution of sea ice concentration (SIC) is central to improving the credibility and operational forecasting capability of deep learning-based sea ice prediction. However, existing deep learning methods often couple the underlying sea ice evolution relationships and data errors within high-dimensional nonlinear mappings, making it difficult to construct a stable and verifiable evolution core and apply it reliably to practical forecasting. To address this issue, this study proposes a reanalysis-forecast dual-domain decoupled framework for learning sea ice evolution operators. The framework builds upon a lightweight physical baseline to generate daily evolution trajectories, employs a temporally constrained joint multi-lead compensation network to com?pensate for unresolved processes, and introduces an ice-mass?aware transport mechanism to suppress numerical dissipation. In the forecasting stage, the parameters of the base evolution core are fixed, while a lightweight variable-semantic adaptation mechanism calibrates inter-domain distributions and evolution responses, thereby separating forecast-domain errors from base evolution errors. Experiments show that the constructed base evolution core can accurately and stably simulate daily sea ice evolution at both short-term and annual scales under reanal?ysis forcing, and can be effectively transferred to the forecast domain through lightweight adaptation, achieving stable prac?tical forecasting capability while preserving the base evolution structure. The source code will be made publicly available at https://github.com/zhangmaqun65535/SNM upon acceptance of this manuscript.
Read the full abstract

All authors Maqun Zhang, Feng Gao, Wankun Chen, Hui Yu, Yanhai Gan, Junyu Dong

Atmospheric and Oceanic Physics · Cross-listed from cs.LG

DySCo: Dynamically consistent data-driven downscaling of extremes in climate projections

S. Stamatelopoulos, M. Wang, I. Lopez-Gomez et al. (+5 more)

Regional climate risk assessment is critical for applications such as infrastructure design, disaster forecasting, and insurance resource allocation. However, estimating regional (i.e., high-spatial-resolution) risk with global climate models (GCMs) remains computationally prohibitive, which has driven the development of downscaling methods for coarse GCM outputs. Downscaling is vital for rare events, since quantifying their extreme properties requires high spatial resolution and very long GCM simulations. These methods non-intrusively increase GCM resolution while correcting statistical biases from unresolved fine-scale processes, thereby improving the accuracy of extreme event statistics with long return periods. A key challenge is preserving dynamical consistency, as freely evolving GCM trajectories are not expected to track the observational dataset used for training the correction operator. This is critical for causal extreme event analyses, where storyline-based risk assessment, i.e., extreme event catalogs, is necessary for effective planning. We address this challenge by introducing Dynamically and Statistically Consistent downscaling (DySCo), a non-intrusive framework yielding high-resolution climate projections consistent with coarse GCM dynamics. DySCo relies on a data-driven reformulation of nudging to create dynamically paired training trajectories without intrusive GCM modifications. Using these paired trajectories, we train a dynamically and statistically consistent, two-stage operator. We evaluate the method by downscaling the Community Earth System Model v2 Large Ensemble (LENS2) in time and space towards historical reanalysis. Results show DySCo achieves superior dynamical consistency with the coarse GCM trajectories, essentially applying a minimal, causal correction to the GCM, preserving top statistical performance comparable to state-of-the-art unsupervised models.
Read the full abstract

All authors S. Stamatelopoulos, M. Wang, I. Lopez-Gomez, L. Zepeda-Nunez, Z. Y. Wan, R. Carver, F. Sha, T. P. Sapsis

Atmospheric and Oceanic Physics

Video Diffusion for Satellite-based High-Dynamical-Fidelity Precipitation (HiDFiP) Field Generation

Runze Li, Yan Xia, Yongquan Qu et al. (+6 more)

High spatiotemporal fidelity precipitation products that accurately capture storm spatial organization, propagation, and lifecycle evolution, are essential for advancing hydrometeorological research and operations at regional and global scales. Satellite products offer the only near-global precipitation observations, but they still fall short of reproducing the spatiotemporal structure of ground-based references, due largely to dynamic distortions from the inhomogeneity, intermittency, and indirectness of satellite retrievals. Here we propose a video-diffusion framework for satellite-based High-Dynamical-Fidelity Precipitation (HiDFiP) field generation beyond the space-time coverage of ground-radar, using radar-rich CONUS as a testbed. The framework performs explicit spatiotemporal modeling with IMERG as the primary source and leverages four-dimensional storm-environment information from about 40 ERA5/ERA5-Land atmospheric/land fields to compensate for the temporal information deficit inherent to satellite retrievals. We introduce an extensive metric suite to assess HiDFiP dynamical fidelity in temporal-reconstruction and spatial-transfer settings against MRMS ground-radar precipitation over CONUS. Relative to IMERG and image-wise diffusion baselines, HiDFiP accurately reproduces the storm space-time spectral characteristics; storm timing, location, and directional propagation; precipitation-event episodicity and temporal structure; precipitation-system morphology and spatial organization; and storm-track kinematics and lifecycle evolution. Transferability experiments indicate that HiDFiP generalizes reasonably well to an unseen region. This work advances a video-diffusion paradigm for satellite-based high-spatiotemporal-fidelity precipitation generation and provides an algorithmic and diagnostic foundation for long-term global radar-grade precipitation records.
Read the full abstract

All authors Runze Li, Yan Xia, Yongquan Qu, Dongwei Fu, Clement Guilloteau, Judy Hoffman, Stephan Mandt, Pierre Gentine, Efi Foufoula-Georgiou

Friday 21 August

Atmospheric and Oceanic Physics

Probabilistic inference of surface parameters for Monin-Obukhov similarity theory

Ethan YoungIn Shin and Michael F. Howland

In simulations of atmospheric flow, the grid spacing typically exceeds the size of the roughness elements at the surface by an order of magnitude. The unresolved effects of surface morphology and roughness on the flow are represented by effective surface parameters and specified as a surface flux boundary condition, most often through a formulation based on Monin-Obukhov similarity theory (MOST). These surface parameters are known to depend on both surface and flow properties, yet they are generally estimated as deterministic quantities with no characterization of associated uncertainty. In this study, we use a Bayesian approach to infer surface parameters for MOST and quantify their uncertainties. For the aerodynamic roughness length $z_0$ inferred in isolation, a normal-normal conjugate update yields the posterior and posterior predictive distributions in closed form. We first demonstrate our method on idealized conventionally neutral boundary layers generated by large-eddy simulation, where $z_0$ is prescribed, and quantify how prior- and observation-related choices shape the inferred posterior. We then apply the method to field observations from the Atmospheric Radiation Measurement Southern Great Plains observatory, from which we infer $z_0$ distributions conditioned on month, on wind direction, and on both jointly. By leveraging a prior informed by sample statistics of all near-neutral observations in the training years, we demonstrate the advantage of the Bayesian inference method for $z_0$ relative to the state-of-practice least-squares profile-fitting method in conditions of data sparsity. Predictions for unseen observations-evaluated at the posterior mean-reduce root mean squared error and mean absolute error in data-sparse wind directions, while posterior predictive distributions consistently reduce the continuous ranked probability score by approximately $20$-$30\%$.
Read the full abstract

Atmospheric and Oceanic Physics

Remote sensing of the temporal and spatial variability of atmospheric refractivity using ADS-B interferometry (ADSBi)

Ollie Lewis, Chris Brunt, Malcolm Kitchen et al. (+2 more)

The extreme variability of humidity in the lower atmosphere over short spatial and temporal scales presents an enormous challenge for existing observing systems. Variations in humidity can be detected due to changes in the refractive properties of the atmosphere, which influence the propagation of electromagnetic radiation. We demonstrate the ability to retrieve high-resolution refractivity gradient measurements using an interferometer to measure the refraction of the Automatic Dependent Surveillance-Broadcast (ADS-B) radio transmission routinely broadcast by commercial aircraft. The ADS-B interferometry (ADSBi) technique is sensitive to small-scale spatial and temporal variability in the refractivity gradient due to changes in humidity. We demonstrate how the retrieval of refractivity profiles is possible exploiting the differences in the observed refraction of radio transmissions broadcast by aircraft at different distances and altitudes.
Read the full abstract

All authors Ollie Lewis, Chris Brunt, Malcolm Kitchen, Neill E. Bowler, Edmund K. Stone

Atmospheric and Oceanic Physics · Cross-listed from physics.flu-dyn

Short spatial mooring-tilt variations from deep Mediterranean observations

Hans van Haren

Interaction between energy-abundant mesoscale eddies and internal waves can lead to convection-turbulence generation and may prove important for deep-sea life and circulation. However, the size of scales of interacting flows is not well known. In this paper, a diagnostic tool of tilt is tested near the single top-buoyancy of 40 mooring lines 9.5 m apart horizontally and compared with 50-m scale relative vorticity and waterflow above a 2500-m deep flat Northwestern-Mediterranean seafloor. Whilst tilt relates to first order with flow-speed squared induced by mooring-line drag, considerable deviations from this relationship and larger tilt occur when the amplitude of relative vorticity attains values O(f), f the inertial frequency of planetary vorticity. The sign of relative vorticity is of no importance. Such larger-tilt events occur during most intense convection turbulence via warm-water slanting from above. During these events, variations in tilt-angle magnitude are as large as the average tilt O(0.1)degree, thereby reducing variational scales from 50 to 9.5 m. Thus, in a deep-sea environment where flow speeds are <0.07 m s^-1, O(0.01) m s^-1 flow-speed variations provide important turbulent mixing, without deep dense-water formation.
Read the full abstract

Atmospheric and Oceanic Physics

A Graph Neural Network Framework for Characterizing Rainfall Variability Regimes across India

Pradyumnan Raghuveeran, Gaurav Chopra, Ajay Bankar et al. (+1 more)

The Indian Summer Monsoon shows significant spatial variation. While prior work primarily focused on forecasting rainfall amounts, little attention has been given to how consistently a location's seasonal rainfall trajectory repeats from year to year. We introduce a graph-based machine learning framework to classify locations across India by this inter-annual consistency. Using 2001 to 2022 GSMaP ISRO data (excluding 2012), we constructed graphs for 29,026 grid points where nodes represent individual years and edges denote cosine similarity. A Graph Convolutional Network classified locations as either consistent or erratic with 96.8% accuracy. Applied to the Indian landmass, the model successfully identified the Western Ghats, Northeast India, and parts of central India as consistent regions. This classification was rigorously validated through statistical testing and temporal stability analysis, showing 93.6% agreement across two independent timeframes. Crucially, the results reveal a previously unreported coupling: regions with higher rainfall volumes are also the most temporally repeatable year-to-year, demonstrating an emergent, spatially coherent structure.
Read the full abstract

All authors Pradyumnan Raghuveeran, Gaurav Chopra, Ajay Bankar, R. I. Sujith

Thursday 20 August

Atmospheric and Oceanic Physics · Cross-listed from physics.flu-dyn

Wind farms as sensor arrays of turbulent boundary layer spatio-temporal flow structure

Manuel Ayala, Dennice Gayme and Charles Meneveau

Temporal fluctuations of wind farm-generated power arise from the interaction between atmospheric turbulence, turbine properties, and wind-farm layout. However, accurately characterizing these fluctuations remains an open challenge. We here present and extend an analytical framework to predict the temporal spectrum of wind farm power-fluctuations, and compare its predictions with detailed large-eddy-simulations (LES) of a wind farm operating within a conventionally neutral boundary layer. The modeling framework assembles several established concepts from turbulent boundary layer physics: a spatio-temporal turbulence spectral model accounting for mean advection and assuming random sweeping by large eddies, a top-down wind farm model of a fully developed wind turbine array boundary layer flow providing the required mean-flow and turbulence scales, and a spatial sampling kernel representing turbine positions and finite rotor size. The latter is extended to three dimensions to represent filtering of spatial fluctuations of turbulence along the vertical direction. Using only atmospheric, turbine, and layout parameters, the model predictions are evaluated against an extensive LES database of a large wind farm on flat terrain. The model accurately predicts the aggregate power frequency spectrum, including peaks associated with advection between turbine rows, the decay of inertial-range turbulence fluctuations due to rotor averaging, and spectra of aggregate power signals from various arrangements of groups of turbines within the array (e.g. staggered or random subsets). The ability to predict wind power fluctuation spectra from fundamental fluid dynamics and existing boundary layer turbulence models could help improve wind farm grid integration.
Read the full abstract

Atmospheric and Oceanic Physics · Cross-listed from physics.flu-dyn

Actuator Disk Models reproduce Actuator Line Model power and thrust fluctuation statistics in wind farm simulations

Manuel Ayala, Dennice Gayme and Charles Meneveau

Actuator-disk models (ADM) are widely used in wind-farm simulations, but their accuracy in reproducing power and thrust temporal fluctuation statistics has not yet been compared to the more accurate and detailed predictions of actuator-line models (ALM). This work provides a detailed comparison of such model predictions under three distinct atmospheric conditions. Data is obtained from a database (JHTDB-Wind) containing actuator-line turbine-response and time-resolved flow field data from a large-eddy simulation of a small windfarm operating over a full diurnal cycle. Power and thrust time series corresponding to ADM are constructed from disk-averaged velocity signals and compared with the corresponding more detailed actuator-line signals over three temporal windows within the diurnal cycle. The analysis compares time series, power spectral densities, and both the fluctuation and increment probability density distributions. The results show that the ADM time-series capture the dominant temporal, spectral, and statistical features of the ALM values at resolved and disk-averaged flow-field time scales, i.e. slower than the rotor frequency. Effects of temporal filtering that are often applied to ADM inputs are examined. Overall, the results provide strong support for the use of ADM for predicting power and thrust fluctuation statistics in wind-farm simulations.
Read the full abstract

Atmospheric and Oceanic Physics

GOES-East full-disk AI nowcasting of cloud evolution in observation space

Dhamma Kimpara, Omid Bagheri, Ivette Hernandez Banos et al. (+2 more)

Clouds affect aviation, solar energy, remote sensing, and storm prediction, yet they remain among the hardest atmospheric features to forecast, particularly at convective scales. Because clouds are shaped by processes spanning a wide range of space and time scales, numerical weather prediction, extrapolation methods, and existing machine learning (ML) approaches are each limited by some combination of accuracy, domain size, and temporal resolution. We present DOP+, an ML approach for clouds that forecasts GOES-East full-disk infrared brightness temperatures by extending direct observation prediction (DOP) with conditioning on meteorological fields. The domain covers tropical, midlatitude, and marine regimes across $\sim 10^8 ~ km^2$, roughly a fifth of Earth's surface. DOP+ forecasts cloud evolution at 10-minute resolution and outperforms persistence, synoptic-scale NWP, and a pure DOP baseline across 0-6 h lead times in fractions skill score and mean absolute error skill score. Convective structure is retained out to 2-3 h. DOP+ thus achieves a state-of-the-art combination of accuracy, temporal resolution, and spatial coverage. Our work lays the foundation for fully global cloud nowcasting at convective timescales.
Read the full abstract

All authors Dhamma Kimpara, Omid Bagheri, Ivette Hernandez Banos, Byoung-Joo Jung, Chris Snyder

Atmospheric and Oceanic Physics

The impact of feature engineering and an optimisation framework for ocean colour machine learning

Edson Silva, Julien Brajard, Simon Cappe et al. (+2 more)

Machine learning (ML) is widely used for the development of ocean colour algorithms, but most studies focus on model parameter training and hyperparameter tuning. The optimisation of the data that feeds the models - i.e., Feature Engineering (FE) - is not fully explored. We assess the impact of FE in ocean colour machine learning models and we propose an optimisation framework that includes seven sequenced levels of data transformation: i. band choice, ii. log scaling, iii. spectral shape normalisation, iv. index extraction, v. principal component analysis, vi. feature scaling, and vii. zero-to-one scaling. We demonstrate the application for Multi-layer perceptron, Support Vector Machines, and eXtreme Gradient Boosting Trees on Sentinel-3 OLCI observations in the Norwegian coastal waters. The models are trained to estimate Chlorophyll-a concentration [Chl-a] and Secchi disk depth (Zsd). Results show that accuracy is highly variable among FE found in six studies using Sentinel-3 OLCI and the ones that we optimise. The R range from 0.01 to 0.55 for [Chl-a] and from 0.15 to 0.68 for Zsd, where the optimised FE shows the top results. The ML models with optimised FE could also improve by two times the R and reduce up to 63% of the mean absolute error when compared to CHL_OC4ME and CHL_NN standard algorithms. Nevertheless, no common optimised FE is found for all target variables and ML models, suggesting that FE optimisation is necessary for each application. Therefore, our proposed framework can be key for improving the accuracy of water quality monitoring in coastal waters.
Read the full abstract

All authors Edson Silva, Julien Brajard, Simon Cappe, Lasse H. Pettersson, François Counillon

Wednesday 19 August

Atmospheric and Oceanic Physics · Cross-listed from q-bio.PE

From objective discovery to prediction of global ocean eco-provinces: A pathway for trustworthy learning

Makayla McDevitt, Maike Sonnewald and Stephanie Dutkiewicz

Marine ecosystems are increasingly impacted by climate change, necessitating tools to identify and predict spatial habitat information. To build such tools, ecological marine provinces, "eco-provinces", ecologically meaningful regions in the global ocean can be used. We use unsupervised machine learning (ML) to identify eco-provinces with corresponding uncertainty measures based on output of a global simulation of phytoplankton functional types. Our work aims to create a proof of concept to predict eco-provinces based on satellite ocean color data. To do so, we develop a hierarchy of explainable dense ensemble networks to infer how well the eco-provinces can be detected from modeled ocean color fields. Key results include that the delineated eco-provinces are both ecologically meaningful and can be inferred with high skill. However, no straightforward relationship was found where adding more input data to the network consistently improves inference skill, and there is an intricate tradeoff between inputs and prediction fidelity. Our work is a case for optimism and a cautionary tale of needing uncertainty quantification and careful validation of prediction fidelity.
Read the full abstract

Atmospheric and Oceanic Physics

Coupled multiscale paleoclimate reconstruction with four-dimensional variational data assimilation

Zilu Meng, Gregory J. Hakim, Julien Emile-Geay et al. (+2 more)

Paleoclimate archives extend climate knowledge beyond the instrumental era, registering different seasons, variables, time averages, and memory lengths. A longstanding problem is to integrate these heterogeneous sources of information within a unified methodology. Here we present a new data-assimilation framework, Last Millennium Reanalysis 4D-Var (LMR4D-Var), which reconstructs climate trajectories from these heterogeneous datasets while balancing errors in the model, observations, and initial conditions. We compare results using LMR4D-Var to assimilate proxies from PAGES2k, Temp12k, and borehole temperature profiles without treating them as instantaneous equivalents. Instrumental verification shows that LMR4D-Var achieves the highest skill compared with previous reconstructions. Borehole assimilation preserves skill against withheld annually resolved records, increases agreement between reconstructed 300–2000-m ocean heat content and independent estimates, and yields a cooler reconstructed Little Ice Age ocean. Results for Temp12k demonstrate assimilation of decadal-to-millennial records and the potential for Holocene and deeper-time applications with suitable emulators.
Read the full abstract

All authors Zilu Meng, Gregory J. Hakim, Julien Emile-Geay, Tanaya Gondhalekar, Eric J. Steig