Wednesday, 4 November 2026: 5 new papers, 1 cross-list and 0 replacements in Atmospheric and Oceanic Physics (physics.ao-ph), in arXiv's order, with full abstracts. The same papers as arXiv's own listing, set as a newspaper; every title links to arXiv.
This work presents a lightweight dataset of tropical cyclone (TC)-following meteorological variables to facilitate the development and application of machine learning (ML) methods in TC research. Using outputs from our previous climate downscaling and idealized simulations, the dataset is constructed as multichannel video sequences that track TC centers throughout their lifetimes. To demonstrate the dataset utility, we present an application in which a regional ML model is trained to predict TC intensity with explicit incorporation of lateral boundary conditions. Our results highlight the importance of lateral boundary information for predicting TC intensity with ML models inside a limited-area domain, underscoring the need to properly account for lateral boundaries when developing regional ML models for TC intensity forecast.
Inferring upper-ocean velocity from altimetric sea surface height (SSH) remains challenging below 50~km, scales now resolved by the Surface Water and Ocean Topography (SWOT) mission. Conventional methods rely on geostrophic or cyclogeostrophic balance. Here, we demonstrate that deep learning can predict the full velocity vector from SSH and environmental data and recover derivative-based diagnostics, including spectral energy flux. We trained a U-Net encoder--decoder on hourly fields from the pre-SWOT Cape Basin extraction of the LLC4320 ocean simulation, with 1.5-km grid spacing ($11.8$--$15.8^\circ$E, $41$--$45^\circ$S). Inputs were SSH, sea surface temperature (SST), and wind stress; targets were horizontal velocities at the surface and 50~m, where wind influence is weaker. At 50~m, the best model achieves a speed RMSE of $0.049\,\mathrm{m\,s^{-1}}$, 29% lower than the geostrophic approximation. Predicted relative vorticity has a correlation of 0.93, compared with 0.75 for the geostrophic approximation. The model also recovers divergence with a correlation of 0.66, whereas geostrophic velocity is nondivergent by definition. Tests of training strategies yield four findings: SSH gradients outperform raw SSH as inputs; wind stress is the most useful covariate for total velocity prediction; SST benefits vary seasonally; and including vorticity and divergence in the loss improves accuracy. These results offer a path for transferring high-fidelity simulation physics to SWOT applications, where kilometer-scale SSH snapshots contain superposed balanced and unbalanced signals.
The land carbon sink explains most of the variability of the atmospheric CO2 growth rate, but the Global Carbon Budget estimates it with dynamic global vegetation models (DGVMs) run once a year. Emulators could provide it sooner if they reproduce each model outside its training period. Here we show that convolutional neural network emulators of 19 of these models, driven by ERA5 climate and their own past predictions, reproduce the global land sink of these models and its spread over two years not used in training. Global monthly net biome production (NBP) anomalies were reproduced with a median R^2 of 0.81. The error of the multi-model mean annual NBP was at most 0.23 GtC yr^-1. At the grid-cell level, the median R^2 of monthly anomalies was 0.30, partly because of errors that accumulate in the emulated NBP history.
Ground-based weather radar mosaics contain systematic blind zones, especially at low altitudes, where missing reflectivity lacks corresponding complete observations for direct verification. We present RadarFill, an observation-constrained conditional diffusion framework for gap filling in three-dimensional radar reflectivity mosaics. RadarFill conditions on partially observed reflectivity and an observation mask, enforces available observations throughout reverse sampling, and preserves valid measurements in the final field. For training and controlled evaluation, structured synthetic missingness is designed from altitude-dependent missing-data statistics and the vertical nesting of blind zones in an operational 3-D radar mosaic product. Across 4,384 held-out test times and representative convective and typhoon cases, RadarFill recovers principal precipitation structures and moderate-to-strong echoes under extensive low-level missingness and generally outperforms conventional interpolation and a learning-based reference model. For real blind-zone evaluation, RadarFill model is applied directly to Fujian Province and compared with 90 matched volume scans from an independent phased-array radar. Within the principal 1-2 km low-level blind zones, RadarFill yields lower reflectivity errors than conventional interpolation, and the gap-filled precipitation structures show broad agreement with independent observations. In two controlled nowcasting experiments, improvements in RadarFill-initialized fields persist into subsequent forecasts, with the largest benefits at low altitudes and short lead times. These results provide complementary evidence from controlled synthetic missingness, independent radar observations, and downstream nowcasting, supporting the potential of observation-constrained conditional generation for gap filling in real 3-D radar blind zones.
Artificial intelligence (AI) based ocean state forecasts now approach or surpass those produced by operational ocean forecasting systems as measured by conventional pointwise metrics. However, these metrics may favor spatially smoothed forecasts and cannot determine whether forecast variables remain physically coherent. Forecast evaluation should therefore move beyond overall pointwise accuracy toward assessing forecast trustworthiness. We propose a multidimensional framework comprising overall pointwise accuracy, physical consistency, and distributional consistency. Overall pointwise accuracy summarizes forecast and reference agreement over all valid samples; physical consistency evaluates dynamical balances using model outputs alone; and distributional consistency evaluates distribution tails and multiscale spatial structures. Three AI models were deployed as representative testbeds in a unified operational environment. This comprehensive assessment shows that lower overall pointwise errors cannot preclude substantial dynamical imbalances and suppressed fine-scale variance. Conversely, higher pointwise errors may arise from spatial displacement of coherent structures rather than a poor representation of their amplitude. The results demonstrate that the evaluation of AI ocean forecasting models should be based on the multi-dimensional assessment of forecast trustworthiness, rather than solely relying on a single metric. The study provides a preliminary framework for AI model intercomparison, which is also informative for targeted model refinement and operational implementation.
Twilight clouds are small, optically faint warm clouds whose droplets fluctuate between haze and activated growth. We develop a nonequilibrium statistical field theory linking this regime to cloud-field organization through K\"ohler microphysics, supersaturation transport, and turbulent mixing. At parcel level, supersaturation rapidly tracks slow droplet growth with a finite lag. Mixing-induced supersaturation removal drives a saddle-node bifurcation, shifting the activation threshold and supporting intermittent hovering under turbulent fluctuations. Starting from turbulent transport of supersaturation coupled to droplet growth, we derive a droplet-supersaturation field system and formulate it within the Martin-Siggia-Rose framework. A Gurari-type variational reduction inspired by polaron theory eliminates the fast bath. Turbulence dresses the slow droplet field, shrinks the forbidden zone below the equilibrium curve, and raises the mixing threshold required to suppress activation. Each droplet population induces a screened supersaturation halo that couples neighboring cloud elements; its delayed adjustment generates dynamic inertia and temporally correlated noise. The resulting nonlinear, non-Markovian pseudoparabolic stochastic equation has a bounded frozen-background response at short wavelengths. Its long-wave coupling amplifies contrasts where the dressed slow manifold turns downward. Comparisons with the parent two-field system show that a spatially correlated Markovian closure preserves droplet-radius distributions and cloudlet organization. Near the activation fold, restoring residual-bath fluctuation memory improves recurrent-hovering frequency and patch extent without recalibrating the drift or fitting a correlation time. More broadly, the framework identifies how slow fields acquire interactions and inertia through the relaxing, diffusive environments they modify.