Papers

The deployment of a geoengineering program to cool Earth's climate may be imminent. It is crucial that tools be developed to ensure that such a program would achieve its objectives while minimizing disruption. Here we exploit recently developed differentiable atmospheric models to demonstrate a novel geoengineering control strategy. In the differentiable primitive-equation atmospheric model JAX-GCM we impose a uniform $+4$\,K ocean warming and ask what pattern of sea-surface temperature cooling – in five ocean-masked zonal bands of prescribed SST forcings whose amplitudes are free – returns land near-surface air temperature closest to the model's own unwarmed climatology. This idealized set-up represents a cooling pattern that could be delivered physically either by marine cloud brightening or stratospheric aerosol injection. Gradients through chaotic dynamics decorrelate from the true sensitivity beyond the Lyapunov horizon, so we optimize greedily over segments of 8 to 14 days, following receding-horizon control. The learned strategy removes $92.3 \pm 0.4\%$ of the realized land warming across a ten-member ensemble of two-year rollouts, and a three-year run sustains it. If we use the spatial pattern of land temperature as the optimization objective, the distributions of precipitation, evaporation, and specific humidity over land are restored as well, even though they are not included in the objective function. The learned strategy from JAX-GCM replayed in the AI emulators LUCIE and NeuralGCM without re-optimization is successful, suggesting robustness. These promising results demonstrate a strategy for designing optimal climate interventions that can be applied broadly for geoengineering scenarios under consideration.

Estimating the impact of observations newly added to an existing numerical weather prediction (NWP) system requires reintegrating the ensemble forecast from the analysis that assimilates the additional observations, which is laborious and computationally expensive. We therefore propose preemptive EFSO (PEFSO; preemptive ensemble forecast sensitivity to observations), which estimates observation impact without model reintegration by invoking a stronger tangent-linear approximation than that used in EFSO. The estimated impact, however, is merely a scalar quantity at a specified verification time, so it is also valuable to obtain the ensemble forecast in which the remaining additional observations are assimilated after denying the detrimental ones. We refer to this procedure as ADD-SEL, but carrying it out requires reintegration. We therefore propose two methods for updating the ensemble forecast without reintegration as approximations to ADD-SEL, collectively termed ADD-SEL-PRE: one recomputes the ensemble transform matrix (i.e., the Kalman gain), and the other modifies only the innovations at lower cost. Experiments with the Lorenz-96 model examine whether both can be approximated without reintegration. PEFSO estimates observation impact comparable to that obtained by EFSO within the range where the tangent-linear approximation remains valid, even under more practical conditions: an ensemble size of 10, the analysis as the reference state, and only the additional observations available. When additional observations are assimilated after denying detrimental ones, both ADD-SEL-PRE methods update the ensemble forecast, giving forecast errors comparable to those from ADD-SEL, particularly for lead times up to two days, where the tangent-linear approximation is expected to hold.

Observations at high frequency have become dramatically more abundant with recent technological advances. Ultra-rapid data assimilation (URDA) has been proposed to exploit them, frequently updating ensemble forecasts at lower cost without re-integrating the forecast model. In this study, we examine the applicability of URDA to a realistic numerical weather prediction (NWP). Specifically, we conducted idealized experiments for the heavy rainfall event of August 2021, using the regional atmospheric model Scalable Computing for Advanced Library and Environment (SCALE-RM). In the experiments, pseudo-observations emulating the Automated Meteorological Data Acquisition System (AMeDAS) were assumed to become available every 10 min. The results show that the threat score for hourly precipitation is generally improved in the forecasts updated by URDA relative to the pre-existing baseline forecasts, through the ensemble-based error covariance. Furthermore, for sea level pressure, temperature, relative humidity, and winds, the root mean square error of the URDA-updated forecasts is reduced relative to that of the baseline forecasts as the forecasts are successively updated by assimilating additional observations. These results indicate the potential of URDA to operate effectively with a realistic NWP model.

Accurate wind forecasts are essential for operational decision-making and public safety, yet forecasts tend to miss near-surface high wind speeds in complex terrain. In response, advances in machine learning (ML) weather prediction methods have demonstrated the ability to improve forecast skill beyond traditional numerical weather prediction (NWP) models. However, the absence of a benchmark dataset to evaluate NWP and ML models with sufficient, quality-controlled wind speed observations in complex terrain poses challenges to the development and intercomparison of high-quality surface wind forecasts across the Conterminous United States (CONUS). We develop the Wind IN-situ Data Benchmark (WIND-Bench), a benchmark dataset from in-situ observations in the Meteorological Assimilation Data Ingest System (MADIS) observational network. WIND-Bench integrates multiple sensor networks with quality control that distinguishes sensor failures from high-wind conditions, using a framework that validates observations against forecasts from the National Oceanic and Atmospheric Administration (NOAA) High-Resolution Rapid Refresh (HRRR) model. WIND-Bench provides a standardized benchmark for evaluating ML and NWP models and for quantifying forecast skill, accelerating the development, evaluation, and operational deployment of skilled near-surface wind forecasts.

We consider narrow-band deep-water surface gravity waves propagating above a background flow whose spatial scale is comparable to the wavelength. We focus on the regime where the background-flow speed is comparable to the Stokes drift of the waves to derive a two-way coupled model between the fast waves and the slow background flow. Nonlinear wave interactions arise at the same asymptotic order as the two-way coupling and must therefore be included in the derivation. The resulting model consists of a nonlinear version of the reduced wave equation obtained in part I, coupled to a Craik-Leibovich equation governing the evolution of the background flow. The model describes the evolution of the wave field and background flow over the slow turnover frequency of the background flow, thus saving the computational burden of temporally resolving the fast wave frequency. The inviscid model exactly conserves wave action, mechanical energy and horizontal momentum, while the viscous model exactly conserves horizontal momentum. The coupling terms are cast in compact form for ease of physical interpretation and numerical implementation.

We consider deep-water surface gravity waves propagating above a background flow whose spatial scale is comparable to the wavelength, focusing on the regime where the flow is slow compared to the group velocity of the waves. We introduce an "equivalent solvability condition" method to construct reduced equations, demanding that, upon multiple-timescale expansion, the reduced equations share the same leading-order solution and first solvability condition as the original system. This approach turns the full 3D problem into a 2D reduced equation for the wave field. We derive such reduced equations for broad-band waves above a depth-invariant background flow, and for narrow-band waves above a fully 3D background flow. In the latter case the reduced equation takes the form of a Schrodinger equation involving the near-surface vorticity of the background flow only, with the impact of the near-surface horizontal flow divergence shown to be subdominant. Beyond the reduction in spatial dimensionality, the latter reduced equation describes the evolution of the wave field over the slow advective timescale of the background flow, thereby eliminating the computational burden of time-resolving the fast wave period. We illustrate the capabilities of the reduced equations through an analytical solution for the weak scattering of a wave packet by a patch of organized flow, followed by numerical solutions for stronger scattering of a wave packet by a patch of disorganized flow.

Accurate kilometer-scale near-surface winds are important for understanding atmospheric processes over complex terrain, yet remain difficult to reconstruct from sparse and unevenly distributed observations. Here we introduce KiloDA, a diffusion framework for hourly kilometer-scale wind reconstruction from surface stations. KiloDA learns the statistical distribution and spatial structure of wind fields from historical 3-km Weather Research and Forecasting (WRF) model forecasts. At each reconstruction time, no contemporaneous WRF field is used. Instead, station observations provide the only constraints on the current atmospheric state and guide posterior sampling from the learned prior. In idealized WRF experiments, KiloDA recovers localized wind structures when only 0.24% of grid cells are observed and shows an overall advantage over conventional interpolation across terrain conditions and wind speed regimes. This capability largely transfers to real observations. In a fully withheld region, KiloDA reduces the median wind speed root mean square error (RMSE) by 19% relative to ERA5 reanalysis, using only observations outside the region, with the largest improvements over high-elevation and high-relief terrain. A random station holdout further confirms that this advantage extends across different complex-terrain locations and holdout configurations. These results show that historical model archives can provide useful structural knowledge for reconstructing kilometer-scale wind fields from sparse observations without requiring an accurate model estimate of the current atmospheric state.

Significant wave height (SWH) retrieval from high-frequency (HF) radar typically relies on a weak second-order Doppler continuum that is sensitive to noise, interference, and spectral leakage. This letter presents a Windowed Envelope Statistics Estimator (WESE) that operates directly on beam-formed time-domain voltages. A second-order term obtained from a Neumann expansion of the rough-surface field equation motivates quadratic compensation of localized radar features. WESE extracts the mean, standard deviation, or variance from overlapping windows of the in-phase, quadrature, or envelope-magnitude sequence, followed by quadratic compensation, rank ordering, least-squares regression, and causal smoothing. Evaluation used 335 synchronized hourly observations from a 13.385 MHz, 12-element WERA system at Argentia, Newfoundland and Labrador. The optimal configuration used quadrature variance, a 16-sample window, 896 retained chronological samples, and 30-h smoothing, achieving an RMSE of 0.152 m and a Pearson correlation of 0.978. This represents RMSE reductions of 32.1% and 18.7% relative to previously reported linear and second-order compensated ordered-statistics models, respectively. The results demonstrate robust time-domain SWH estimation without explicit Doppler-spectrum construction.

Medium-range weather forecasting underpins operational and planning decisions across the energy industry. Developing competitive weather models was once the domain of national meteorological centers, but recent advances in machine-learned weather prediction (MLWP) have opened the field to industry. We present Aries, a SwinTransformer-based MLWP model developed at InCommodities. Aries is trained on ERA5 reanalysis data at 0.25\textdegree{} resolution, predicting 74 prognostic and 11 diagnostic atmospheric variables. We evaluate the model on 2025 ECMWF Analysis initializations, ensuring a recent and strictly out-of-sample test period for all models compared. On 10-metre wind speed, Aries outperforms both ECMWF HRES and AIFS in terms of RMSE for lead times up to four days, while on 2-metre temperature it achieves RMSE on par with AIFS operational. These results demonstrate that proprietary development of competitive weather models is technically viable, supporting a broader set of forecasts available for operational and planning applications in the energy industry.

The potential of crossing climate tipping points (TP) has reached the attention of many researchers and the general public. On the one hand, the basis for this concern is strengthening, with simulations showing that abrupt transitions might occur even for moderate emission scenarios. On the other hand, our understanding of what constitutes such transitions mathematically is becoming more nuanced. This leads to challenges for the fidelity of early-warning signals (EWS), which accompany bifurcations in systems that closely track a slowly changing steady state. Different kinds of out-of-equilibrium dynamics in a rapidly changing climate, as well as chaos, blur the critical tipping threshold and imply limits in predictability. Using EWS to detect loss of local stability from data requires careful evaluation in the high-dimensional system, and may have limited predictive power in a highly multistable climate, where there is more than a single, well-known alternative state. This calls for new methods using non-equilibrium statistical mechanics and dynamical systems to efficiently probe the global stability properties of climate models and observations.

Because precipitation is regulated by intricate interactions between topography, cloud microphysics,cumulus convection, radiation, and boundary-layer processes, it is still difficult to accurately simulate orographically enhanced monsoon rainfall over the Western Ghats.

This study performs a scale analysis for the dynamics of equatorial flow systems like the Madden$-$Julian oscillation (MJO). This analysis makes three assumptions about the geometry of the flow and focuses on the leading order terms of the mass continuity and momentum equation system. First, the flow is meridionally narrow. Second, the zonal structure is planetary wavenumber 1. Last, the flow is vertically shallow. When constrained by these geometric assumptions, the system shows a time scale much longer than a day depending on the meridional narrowness. Specifically, for equatorial Kelvin waves, the oscillation period ranges from 36 to 64 days given a ratio of the planetary radius to the meridional width scale from 6 to 8 (i.e., an e-folding latitude from 9.59$^\circ$ to 7.18$^\circ$ either side of the equator). Given the long, narrow, and shallow geometry, the dynamics can sustain Kelvin waves only at low frequencies. This result suggests that the intraseasonal time scale of the MJO arises from the Kelvin wave frequency limit due to the planetary wave geometry. This planetary intraseasonal scale relation is independent from thermodynamics, and an open question remains on what physical processes coordinate the thermodynamics in the same length or time scale.

Local heatwave hazard depends on fine-scale air temperature, but ground stations are sparse and reanalysis products such as ERA5 cannot resolve the terrain and land-surface contrasts that shape real heat exposure. We present a conditional diffusion emulator for high-resolution 2-m temperature downscaling, conditioned on static geography, a training climatology, exact-time ERA5 temperature, and solar and temporal features, guided at inference by score-based data assimilation (SDA): a differentiable Gaussian observation likelihood steers the diffusion score toward sparse revealed temperature observations without any retraining. On a controlled 32-case synthetic-grid protocol over AORC, guidance improves hidden-cell reconstruction over both ERA5 and a strong observation-proximal nearest-neighbor baseline once observation density reaches 1% (RMSE 0.318 vs.\ 0.431~K, winning all 32 cases), while sparser regimes still favor direct interpolation. We further map the full guidance-strength landscape across three observation densities, showing that the optimal strength shifts systematically with density and that over-guiding causes sharp, predictable degradation – giving a concrete operating recipe rather than a single untuned setting. The resulting fields are intended as a temperature layer for downstream heatwave-hazard products such as threshold exceedance and cumulative heat-burden. The present evidence is a controlled synthetic-grid validation; station-network and held-out-year evaluations are the next steps toward deployment.

Subseasonal-to-seasonal (S2S) forecasting is societally critical, supporting decision-making in sectors ranging from water and agricultural management to disaster risk reduction, energy planning, and insurance. Achieving reliable predictions at these timescales requires representing the ocean and its dynamics, but traditional physics-based Ocean General Circulation Models (OGCMs), are computationally expensive and difficult to develop and improve because of the code complexity. In this work, we propose Neptune, an end-to-end data-driven framework for global ocean and sea-ice components emulation tailored for S2S timescales, up to 60 days. Neptune combines Convolutional Neural Networks (CNNs) and Spherical Fourier Neural Operators (SFNOs) to effectively capture local features and global cross-scale interactions, thereby obtaining a coherent representation of the ocean state. Forced by prescribed daily atmospheric fields, Neptune emulates ocean state variables, from temperature and salinity, to zonal and meridional currents, from sea surface height to sea ice thickness and concentration, with daily outputs at the ocean surface and through the water column. Specifically, we propose two variants of Neptune, Neptune-1 and Neptune-025, capable of emulating the ocean state at 1{\deg} and 0.25{\deg} resolution, respectively. Evaluated against a suite of metrics, including statistics (RMSE, CRPS and ACC), physical coherency (Ocean Heat Content, Eddy Kinetic Energy and Ice Brier Score) and climate indices (ENSO and Z20 metric, IOD), Neptune successfully reproduces the spatio-temporal evolution of the oceanic fields up to 60 days, and is stable over long timescales. Neptune provides compelling evidence that end-to-end data-driven ocean emulators can become a powerful component of next-generation S2S forecasting systems, emulating ocean state at high spatio-temporal resolution.

This study examines how a risk-based attribution analysis of European heat wave events under climate change varies when explanatory variables are included in the analysis. Instead of relying solely on temperature statistics, the analysis is conducted while conditioning on large-scale circulation or pre-existing climate anomalies. Using the hot European summers of 2006 and 2007 as examples, we observe several systematic effects. Controlling for the presence of atmospheric blocking explains much of the natural temperature variability and greatly improves the separation of present and pre-industrial distributions in both years. In 2006, the predominant strong blocking raised the expected temperatures, so that, by comparison, the observed anomaly no longer appeared unusually hot, even in a pre-industrial climate. More complex effects occur when the conditions are part of the causal chain of climate change, for example, acting as moderators or mediators. By conditioning on such variables, we may remove part of the climate change signal and thus fundamentally alter the outcome of the attribution study. Using a simplified model we demonstrate why attribution studies with different kinds of conditions can reach different, seemingly contradictory conclusions.

Machine-learning weather models (MLWMs) now match or outperform operational numerical weather prediction (NWP) at global medium-range forecasting, at far lower inference cost. Many deployed MLWMs are deterministic, producing a single forecast with no estimate of its own uncertainty, whereas a growing family of trained-probabilistic models generate calibrated ensembles directly, at the price of a dedicated training run. We ask instead how much uncertainty can be extracted from a deterministic checkpoint that already exists, without retraining it. Where physical ensembles represent model uncertainty by stochastically perturbing parametrisation tendencies, we perturb the network's raw weight tensors at inference time, a scheme we call stochastically perturbed weights (SPW). We also ask whether it works, where and on which scales to inject the noise, and where it fails. A three-phase ablation across four deterministic backbones, Aurora, GraphCast, SFNO, and AIFS, selects one production baseline per model, benchmarked against the trained-probabilistic AIFS-ENS, FourCastNet 3 and Atlas as well as the operational ECMWF ensemble (IFS-ENS) over 112 initialisation times. At a 240 h (10-day) lead time the SPW ensembles reach continuous ranked probability skill scores (CRPSS) between 0.04 and 0.13 below the best trained-probabilistic baseline, at zero marginal training cost. No injection site works across models: the productive tensor group is architecture-specific, so SPW is at present a tuning procedure rather than a plug-and-play recipe. Its main failure mode is a coherent whole-field offset that overdisperses the domain mean, and restricting the noise to coarse scales or perturbing the initial conditions each repair part of it.

Projecting the impacts of climate change requires large ensembles of climate variables that match historical observations, align with the warming ranges assessed by the IPCC, and can efficiently run new future emissions scenarios, including the newest generation of climate model scenarios (CMIP7) and pathways consistent with countries' Paris Agreement pledges. Generating such ensembles at the scale needed for impact studies is normally computationally prohibitive. We close this gap with GCMagicc, a hybrid model that pairs a simple physical climate model with machine learning to generate ensembles of 10 climate variables at the resolution of full-scale Earth system models, without relying on GPU resources or retraining for new scenarios. Trained on 32 CMIP6 Earth system models and observational/reanalysis data, GCMagicc complements rather than replaces Earth system models. We apply it to a range of future pathways: the canonical SSP scenarios of the latest IPCC report (1.2-6.1{\deg}C warming, min-max across scenarios of 5-95 percentile ranges), current policies (2.3-4.0{\deg}C), national pledges under the Paris Agreement (1.5-3.3{\deg}C) and the CMIP7 range from the 'VL' to 'H' scenarios (1.2-4.2{\deg}C), releasing a large public dataset. As an illustration, we perform an attribution analysis of the severe 2025 Iranian drought using GCMagicc ensembles, with three CMIP6 large ensembles for comparison, with and without anthropogenic forcings. The results suggest a strong anthropogenic signal: a median probability of drought at least as severe as observed of 29% with anthropogenic forcing, and zero under natural-forcing-only simulations. In the future, drought conditions are projected to materially worsen, amplifying the potential for agricultural and food security impacts and geopolitical conflicts that use water scarcity as a weapon. GCMagicc data is available at https://gcmagicc.org.

Despite growing interest in AI-based prediction of climate extremes, event- or label-based AI-ready extreme climate datasets remain limited, constraining efforts to systematically characterize and forecast such phenomena. To address this gap, we present EastAsiaClimateExtremes, an open dataset that provides ERA5/OISST reanalysis-based weekly extreme labels and event-based metrics for anomalously high temperature (AHT), heavy rainfall (HR), and marine heatwaves (MHW) over East Asia, together with analysis workflows hosted on GitHub to facilitate reproducibility and adaptation. The dataset is fully documented and co-registered with ECMWF S2S hindcast outputs on a common spatial grid and temporal framework, thereby enabling direct comparison between reanalysis-derived labels and dynamical model forecasts. This unified dataset serves as a reference framework for East Asian climate extreme research and AI-based subseasonal-to-seasonal prediction. It supports both quantitative characterization of the spatiotemporal occurrence of regional extremes and systematic diagnosis of S2S model skill in reproducing extreme signals. Beyond these immediate applications, the dataset enables a broader range of studies such as extreme event attribution and compound risk analysis. The accompanying analysis workflows characterize the historical statistics of reanalysis-based weekly extremes-including occurrence frequency and mean and maximum intensity-together with their climatological means and trend characteristics, and further assess the skill of the ECMWF hindcast.

Earth System Models rely on tunable subgrid-scale parameterizations, but optimizing these parameters is computationally expensive, particularly when nonlinear interactions require many simulations. We present a hybrid Physics-ML multi-fidelity framework that combines Green's Function Optimization (GFO) with Gaussian Process or Neural Network surrogate optimization. Using a quasi-geostrophic turbulence model, GFO first ranks parameter sensitivities in normalized coordinates and selects a reduced active subset. Nonlinear surrogates then explore this subset using inexpensive 30-day simulations before refining promising candidates with 180-day simulations. Across seven strategies and a 35-member ensemble, GFO-MultiGP and GFO-MultiNN achieved mean improvements of 64.6 percent and 65.2 percent, respectively, while reaching practical saturation after 3,060 and 2,520 simulation-days. The corresponding standalone GP and NN achieved 61.1 percent and 42.0 percent improvements and required 7,740 and 6,660 simulation-days. These results demonstrate an end-to-end sample-efficiency advantage for the tested hybrid pipelines. Because screening, dimensionality reduction, initialization, and fidelity scheduling change simultaneously, their individual contributions are not isolated.

We introduce the Terrain-Weighted Forest Fire Model (TFFM), a lattice model in which fire spreads on a spatially correlated Gaussian height field with the asymmetric bond probability $p_{i\to j}=\mathrm{clip}[e^{-\beta+\gamma(h_j-h_i)},0,1]$, plus an additive wind bias. Simulations on lattices up to $L=8192$ reveal a sharp active-to-inactive transition whose critical suppression threshold $\beta_c$ is even in $\gamma$, decreases with $|\gamma|$, and decreases as the terrain correlation length $\sigma_h$ is reduced: slope asymmetry acts as a suppressant because downhill bonds are penalized and fire stalls at local elevation maxima. For rough terrain and low tree density the fire fails to percolate even at zero suppression. Finite-size scaling on a fine $\beta$ grid at $L=2048$–$8192$ gives a correlation-length exponent $\nu=1.8\pm0.17$ from both the susceptibility peak and the width of the transition, and a front-velocity exponent $\delta=0.34\pm0.03$, identical for smooth and rough terrain; neither matches directed percolation ($\nu_\perp=0.73$, $\nu_\parallel-\nu_\perp\approx0.56$) or isotropic percolation ($\nu=4/3$, $\approx0.18$). The single-seed survival probability at $\beta_c$ is independent of $L$ and decays extremely slowly, with a running exponent falling from $\approx0.09$ to $\approx0.04$, excluding directed percolation and pointing to a survival probability that remains finite at criticality, consistent with the $L$-independent value $P^*\approx0.5$ at which $P_{\rm surv}$ drops to zero. Wind raises $\beta_c$ by a factor of $2$–$4$, produces a sharp onset of downwind fire-scar drift at weak coupling, and, at high terrain coupling, decreases the burned fraction at boundary crossing—a terrain-wind competition effect absent from isotropic bond-disorder models. The model yields fire-risk thresholds and fire-scar signatures comparable to satellite burn-scar data.

Medium-range precipitation forecasts are impaired by persistent systematic biases, lead-time-dependent error accumulation, and coarse spatial resolution, restricting their reliability for flood-drought risk assessment. Existing AI correction techniques lack dedicated modeling for multi-day dynamic bias evolution and proper meteorological constraints, often generating over-smoothed rainfall structures, and cannot meet operational deployment demands. This work introduces PCSDiff, a cascaded task-decoupled diffusion framework targeting 10-day precipitation bias correction and downscaling. To jointly counteract temporal error drifts and reconstruct physically plausible local precipitation details, PCSDiff integrates the Precipitation Intensity-aware Multi-branch Decoder (PIMD) module for dynamic multi-day error mitigation using synoptic-temporal features, followed by a two-phase conditional diffusion super-resolution module to restore fine-scale precipitation patterns. Evaluated against CMA-CRA observations over China after global-data training, PCSDiff cuts RMSE by 16.1% and lifts ACC by 13.9% relative to raw ECMWF forecasts at 3-10-day lead times, and consistently outperforms mainstream deep-learning baselines on both general and extreme-precipitation metrics. Benefiting from a streaming inference pipeline, our method achieves low-latency rolling forecasting for practical meteorological operations.

A theory for the general circulation of the atmosphere must be based on a theory of its macroturbulence. A central component is the near-surface eddy heat flux, which is closely tied to the mass transport of the surface branch of the circulation. Existing scaling theories are largely formulated within the two-layer quasi-geostrophic framework, which assumes supercritical states and an inverse cascade of kinetic energy—conditions that are frequently not satisfied in planetary atmospheres, including Earth's. Nevertheless, such theories have shown some empirical success, including that the Rhines scale is associated with the mixing length even in the absence of any inverse energy cascade. Here, we analyze hundreds of idealized, dry general circulation model simulations spanning wide ranges of rotation rates, meridional temperature gradients, vertical stratification, and seasonality. Most simulations reside in the marginally critical regime where strongly nonlinear scaling theories fail. We propose a new theory that captures both domain-averaged and local eddy heat flux behavior across all simulations. The theory uses scaling arguments for the zonal momentum balance and accounts for kinematic effects on eddy mixing. The scaling depends on a non-dimensional thermal Rossby number and implicitly accounts for the role of nonlinear eddy momentum fluxes across jets. We demonstrate the skill of the proposed scaling in explaining global near-surface temperature distributions across a wide range of climate states in a simple energy balance model. Implications for general circulation theory and extending the scalings to moist atmospheres are discussed.

A coupled analytical framework is developed to predict the internal boundary layer (IBL) height, wall shear stress and the mean velocity profile downstream of a surface roughness transition in a neutral atmospheric boundary-layer. The new model combines a three-layer analytical velocity formulation with a modified diffusion analogy for the IBL growth rate. The diffusion analogy links the growth rate of the IBL to turbulent diffusion and to mean vertical advection. Our formulation rectifies physical inconsistencies in a previous model that assumed that the turbulent diffusion is controlled only by the upstream surface conditions and that the characteristic streamwise velocity difference caused by streamline displacement is independent of downstream distance from the roughness transition. Key model parameters, namely the turbulent diffusion coefficient and the eddy viscosity augmentation coefficient, are modelled as functions of the upstream-to-downstream aerodynamic roughness length ratio. The coupled model, along with three other analytical models, is tested against twelve datasets covering a wide range of roughness ratios, including wind-tunnel experiments and large-eddy simulations. The new model accurately predicts wall shear stress, mean velocity, and IBL height for all combinations of upstream and downstream roughness values evaluated, for both smooth-to-rough and rough-to-smooth transitions.

Long-term climate data exhibit variations composed of internal and forced components. Internal variability arises from natural processes that could occur within a stable climate. Forced variability, on the other hand, reflects climate changes induced, for example, by anthropogenic greenhouse gas and aerosol emissions. Accurately distinguishing between these types of variability is crucial for attributing climate fluctuations and understanding internal variability processes and impacts. In this study, we apply a U-Net convolutional network, a model commonly used in computer vision, to separate forced and internal climate variability from 1950 to 2022 using a multi-model dataset. The dataset includes multiple fields from five single-model initial-condition large ensembles. Cross-validation is conducted by training the U-Net using the data from the ensembles of four models, leaving out the data from one model to assess performance. Validation results yield errors ranging from 0.1{\deg}C to 0.4{\deg}C for the forced variability of local-scale monthly surface air temperature, which is no more than half the magnitude of external forcing. The U-Net achieves better performance than a simple approach based on a fourth-order polynomial trend for estimating forced variability. The error is mainly due to insufficient sampling and poor agreement among the models, as the U-Net underestimates the warming for the model with the highest transient climate sensitivity during validation. Performance is lower for variables like sea-level pressure and precipitation because of their low ratio of forced to internal variability. This framework might be enhanced by incorporating a larger multi-model dataset in the training and validation.

As global warming intensifies, the frequency and intensity of extremely high temperatures are expected to increase. This will impact the production of photovoltaics (PVs), which are increasingly adopted as an effective alternative to replace fossil fuel-based energy sources and reduce CO2 emissions. Furthermore, extremely high temperature days account for a considerable proportion of days with high PV power potential (PVpot). Therefore, this study investigates changes in PVpot on future extremely high temperature days over East Asia, a region with high greenhouse gas emissions and vulnerability to extreme climatic events. The East Asia-averaged PVpot for extremely high temperature days was estimated to decrease across all scenarios and future periods. The East Asia-averaged PVpot for extremely high temperature days was predicted to decrease more substantially toward the late 21st century, with a larger magnitude of decrease expected under the high-carbon emissions scenario compared to the low-carbon emissions scenario. By the mid-and late 21st century, PVpot for extremely high temperature days was projected to decrease in PV hotspot areas, particularly in the regions of northern China and southern Mongolia, by up to -7.2 %. The signs of PVpot projections vary across sub-regions under summer mean conditions, while on extremely high temperature days, PVpot is consistently expected to decrease in all regions. This suggests that extremely high temperatures further intensify the decrease in PVpot. Moreover, under extremely high-temperature conditions, near-surface air temperature has been identified as the primary driver of projected decreases in PVpot among the climate variables considered; its influence is expected to intensify over time, thereby accelerating PVpot decreases under the high-carbon emissions scenario.

Subgrid parametrizations of vertical mixing in ocean models depend on free coefficients that cannot be measured directly and must be calibrated against high-fidelity references such as large-eddy simulations (LES). Existing approaches return point estimates and leave the associated uncertainty unquantified, a limitation when the inverse problem is ill-posed or when distinct parameter configurations fit the data comparably well. Simulation-based inference (SBI) addresses exactly this: given a prior and access to the simulator, it approximates the full posterior over parameters without requiring a tractable likelihood, at a cost set by the number of simulator evaluations. We apply it to tunax, a JAX-based single-column ocean model, to calibrate the coefficients of its $k$–$\varepsilon$ closure. A blockwise PCA summary statistic compresses the simulator output along the depth axis while preserving its forcing–horizon–variable structure, making inference tractable at modest budgets. We compare neural posterior estimation and its sequential variants against a recent training-free approach built on a tabular foundation model. The latter recovers informative posteriors from a few hundred simulator calls, outperforming the trained estimators at every budget considered.

The 26 August 2026 Gyirong cascade connected a high-mountain failure to developed downstream valleys. Event-bracketing observations delineate a preferred changed/source surface of 1.01 km^2, within alternative interpreted envelopes of 0.49-1.84 km^2, and a representative 21.8 km source-to-port route descending 3.40 km. Across two precipitation products, five antecedent windows and six fixed spatial supports, all 60 matched-year ranks remain below the 90th-percentile wet threshold. The source-nearest 7-day temperature mean of 9.43 degrees C exceeds all 25 matched years from 2001-2025. Positive temperature anomalies extend to every tested support, but ranks vary from the 88th percentile to above all historical values. Within the subsequently affected 37.354 km^2 UNOSAT footprint, modelled built-up surface increases from 0.036 to 0.441 km^2 between 1975 and 2020; its fraction of the fixed area rises from 0.096% to 1.181%. Delivered rapid maps contain 695 building points and 15.671 km of roads graded destroyed. These measurements connect the physical footprint with environmental context and historical exposure development. They complement existing process reconstruction while leaving the initiation mechanism and the mechanical role of warming unresolved.

Hail can cause large financial losses and poses risks to public safety, making reliable nowcasts essential for timely warnings. Deep-learning approaches have emerged as a strong alternative to conventional methods, but how input data choices affect performance has not been deeply explored. We investigate how the skill of a deep-learning hail nowcasting model can be improved without changing the model architecture. Sensitivity experiments assess the impact of training-data volume, random data augmentation (mirroring and rotation) and the number of input timesteps. Increasing the years of training data substantially improves forecast skill, by up to 25 minutes at later lead times. Augmentation improved performance for larger datasets but interestingly degraded performance for smaller ones. Sensitivity to input timesteps was weaker than sensitivity to training years. These findings show that improvements in data selection and preprocessing can yield substantial gains even with a fixed architecture, offering guidance for future deep-learning nowcasting development.

High-resolution cross-correlation spectroscopy (HRCCS) is a powerful technique for detecting molecules whose individual spectral lines are too weak to be identified directly, but its sensitivity is limited by the availability and quality of high-resolution opacity data. Many molecules of atmospheric and astrobiological interest lack complete line lists, restricting traditional template-based searches. In this work, we use Titan as a controlled testbed to develop and validate a new cross-section-based methodology for HRCCS template construction. We analyse K-band CRIRES+ observations of Titan (1.99 - 2.48 {\mu}m), and compute cross-correlation functions using both line-by-line and cross-section-based templates. Our analysis recovers known hydrocarbons such as methane (CH4) and acetylene (C2H2), and yields the first HRCCS detection of ethane (C2H6) with a significance of SNRpeak = 5.17 {\pm} 0.07. The ethane detection was made possible exclusively through cross-section-based templates, as no high-resolution line list currently exists for this molecule. These results demonstrate that cross-section-based template construction is a practical and powerful strategy for extending HRCCS to molecules that currently lack reliable line lists, and establish Titan as a benchmark for calibrating molecular detection techniques that can be applied to both solar system and exoplanet atmospheres. Future applications of this approach to other ground-based high-resolution spectrographs, as well as to JWST's highest-resolution modes and next-generation facilities such as the ELT, could significantly expand the inventory of molecules detectable in planetary atmospheres.

Accurate cloud-cover forecasts are important for temperature prediction, radiation forecasting, and solar-power operations. Short-range forecasting methods can preserve observed cloud placement during the first forecast hours, but their skill decreases when cloud fields evolve through formation, dissipation and deformation. Longer lead times require accounting for atmospheric evolution, but operational numerical weather prediction (NWP) forecasts may not accurately represent the satellite-observed cloud state at initialization. We develop CloudCast v2, a machine-learning model for 12-hour cloud-cover forecasting from observation-based initial conditions. The model is first trained on the Copernicus European Regional Reanalysis (Ridal2024) to learn cloud-evolution dynamics, and is then adapted to satellite-derived cloud fields using conditional flow matching (Lipman2023), a generative method that transforms noise into cloud-cover forecasts conditioned on the observed initial cloud fields and NWP inputs. CloudCast v2 reduces mean absolute error by 10% relative to its predecessor, CloudCast v1 (Partio2025), over the 1-12 h range. It also overtakes CloudCast v1 in fractions skill score, a neighborhood-based measure of spatial agreement, after approximately 3-6 h, depending on the cloudiness category. These results show that observation-initialized machine-learning forecasts can extend beyond the usual 1-3-hour nowcasting range while retaining spatial detail from satellite cloud fields.