An ML model trained only on harmonized gridded observations achieves competitive medium-range weather forecast skill with the IFS for several upper-air and surface headline scores when verified against observations.
Baseline reference
Title resolution pending
Baseline reference. 55% of citing Pith papers use this work as a benchmark or comparison.
45
Pith papers citing it
22.2k
external citations · Crossref
Baseline
55% of classified citations
citation-role summary
dataset 7
background 4
citation-polarity summary
representative citing papers
Observation-only machine learning can generate multi-decade global atmospheric reanalyses with large-scale skill near ERA5 and surface errors between ERA-Interim and ERA5, in a single day of compute.
HO-FNO extends standard FNO with n-linear spectral mixing and shows improved accuracy on nonlinear PDE benchmarks, sometimes with a single layer beating deeper FNO models.
A novel observationally constrained probabilistic trigger for mesoscale convective systems improves spatiotemporal scales of tropical precipitation and ensemble spread in NWP models compared to prior MCSP schemes.
GPROF-IR is a CNN-based retrieval that uses temporal context in geostationary IR observations to produce precipitation estimates with lower error than prior IR methods and climatological consistency with PMW retrievals for integration into IMERG V08.
Under one protocol, most AI climate models reproduce historical climatology and ENSO response as well as a CMIP6 model, but some underestimate warming trends and all diverge on +2/+4K SST experiments.
A multilinear operator learned on PCA coefficients maps time-since-ignition inputs to smoke outputs, matching Monte Carlo accuracy with half the model calls and outperforming prior classifiers on holdout data.
Cast3 translates NWP principles into a data-driven model using cubed-sphere grids, super-ensembles, and generative nudging to achieve state-of-the-art ensemble predictions that outperform baselines.
Time series foundation models match the performance of specialized models for day-ahead load forecasting while providing explanations that match domain knowledge on weather and calendar effects.
SCARFACE provides a harmonized annual panel dataset of over 2,700 socio-economic, environmental, and agricultural indicators for 256 sub-regions in the Po Valley, Italy, spanning 2011 to 2024.
Relaxed unitary convolutions for GNNs on meshes balance smoothness preservation with natural smoothing in dynamics, outperforming unitary convolutions and other models on PDEs and weather tasks.
FPPF learns a conditional flow-matching proposal that approximates the optimal particle-filter proposal, retains exact importance weights, and with localization outperforms classical and generative DA baselines on chaotic systems.
MTGNN with hybrid adjacency matrix reconstructs GRACE-like TWS anomalies to 1940, reaching basin-mean correlation 0.94 and competitive performance with fewer predictors than baselines.
Online conformal prediction post-processing guarantees calibrated uncertainty coverage for GenCast, NeuralGCM, and AIFS-ENS forecasts of temperature and precipitation including extremes.
Training climate emulators on random-CO2 runs that break SST–CO2 correlation, plus an energy constraint, yields a data-efficient model that works on AMIP+4K and abrupt 4xCO2 cases prior ACE models mishandled.
A zero-shot pipeline with physics-informed synthetic histories lets time-series foundation models outperform baselines by 1.7-2x in cold-start PV forecasting on 440 sites across four datasets.
NTK-UQ produces 31-37% sharper 90% prediction intervals than split conformal prediction for extreme weather forecasts, with adaptive scaling via architecture-dependent eigenvalue truncation and ICA decomposition of last-layer features.
Distributional autoencoders trained on climate model simulations model full conditional distributions of European temperature fields to enable probabilistic storyline attribution, illustrated by higher intensities and probability ratios for a 2003-like heatwave in 2028 and 2053.
ML climate emulators degrade under seasonal distribution shifts that proxy long-term climate change, but physically motivated compositional decompositions improve out-of-distribution performance with modest in-distribution trade-offs.
AxiomOcean deploys a 3D encoder-backbone-decoder architecture that jointly predicts upper-ocean variables and outperforms prior AI models by 20-35% in day-1 RMSE while preserving eddy kinetic energy and vertical consistency.
Commutativity regularization mitigates transient error amplification in autoregressive neural simulators by penalizing non-normality and non-commutativity of Jacobians, yielding stable long-horizon rollouts.
A new optimization algorithm with double machine learning for wildfire spread estimation enables better crew assignments that reduce total area burned.
M-CaStLe generalizes local stencil-based causal discovery to the multivariate case and decomposes resulting graphs into reaction and spatial components for interpretation in space-time gridded data.
ESFM is a single open foundation model that unifies heterogeneous Earth data sources and forecasts missing regions while preserving inter-variable physical relationships.
citing papers explorer
-
AIFS-DOP: End-to-End Medium-Range Weather Prediction from Observations Alone with Machine Learning
An ML model trained only on harmonized gridded observations achieves competitive medium-range weather forecast skill with the IFS for several upper-air and surface headline scores when verified against observations.
-
Global reanalysis from observations alone with machine learning
Observation-only machine learning can generate multi-decade global atmospheric reanalyses with large-scale skill near ERA5 and surface errors between ERA-Interim and ERA5, in a single day of compute.
-
Higher-Order Fourier Neural Operator: Explicit Mode Mixer for Nonlinear PDEs
HO-FNO extends standard FNO with n-linear spectral mixing and shows improved accuracy on nonlinear PDE benchmarks, sometimes with a single layer beating deeper FNO models.
-
An observationally constrained probabilistic trigger for organized deep convection in an NWP ensemble
A novel observationally constrained probabilistic trigger for mesoscale convective systems improves spatiotemporal scales of tropical precipitation and ensemble spread in NWP models compared to prior MCSP schemes.
-
GPROF-IR: An Improved Single-Channel Infrared Precipitation Retrieval for Merged Satellite Precipitation Products
GPROF-IR is a CNN-based retrieval that uses temporal context in geostationary IR observations to produce precipitation estimates with lower error than prior IR methods and climatological consistency with PMW retrievals for integration into IMERG V08.
-
AIMIP Phase 1: systematic evaluations of AI weather and climate models
Under one protocol, most AI climate models reproduce historical climatology and ENSO response as well as a CMIP6 model, but some underestimate warming trends and all diverge on +2/+4K SST experiments.
-
Enabling Real-Time Training of a Wildfire-to-Smoke Map with Multilinear Operators
A multilinear operator learned on PCA coefficients maps time-since-ignition inputs to smoke outputs, matching Monte Carlo accuracy with half the model calls and outperforming prior classifiers on holdout data.
-
Cast3: Translating numerical weather prediction principles into data-driven forecasting
Cast3 translates NWP principles into a data-driven model using cubed-sphere grids, super-ensembles, and generative nudging to achieve state-of-the-art ensemble predictions that outperform baselines.
-
Explainable Load Forecasting with Covariate-Informed Time Series Foundation Models
Time series foundation models match the performance of specialized models for day-ahead load forecasting while providing explanations that match domain knowledge on weather and calendar effects.
-
SCARFACE: a harmonized spatio-temporal dataset integrating socio-economic, environmental, and agricultural indicators for the Po Valley (Italy), 2011--2024
SCARFACE provides a harmonized annual panel dataset of over 2,700 socio-economic, environmental, and agricultural indicators for 256 sub-regions in the Po Valley, Italy, spanning 2011 to 2024.
-
Smoothness Errors in Dynamics Models and How to Avoid Them
Relaxed unitary convolutions for GNNs on meshes balance smoothness preservation with natural smoothing in dynamics, outperforming unitary convolutions and other models on PDEs and weather tasks.
-
Generative Model Proposal based Particle Filtering for Data Assimilation
FPPF learns a conditional flow-matching proposal that approximates the optimal particle-filter proposal, retains exact importance weights, and with localization outperforms classical and generative DA baselines on chaotic systems.
-
Reconstructing GRACE Terrestrial Water Storage with Spatio-Temporal Graph Neural Networks: An Application to South America
MTGNN with hybrid adjacency matrix reconstructs GRACE-like TWS anomalies to 1940, reaching basin-mean correlation 0.94 and competitive performance with fewer predictors than baselines.
-
Rigorous uncertainty quantification of probabilistic AI weather forecasts with conformal prediction
Online conformal prediction post-processing guarantees calibrated uncertainty coverage for GenCast, NeuralGCM, and AIFS-ENS forecasts of temperature and precipitation including extremes.
-
Disentangling the effects of sea surface temperature and CO$_2$ in global machine learned weather-climate emulators
Training climate emulators on random-CO2 runs that break SST–CO2 correlation, plus an energy constraint, yields a data-efficient model that works on AMIP+4K and abrupt 4xCO2 cases prior ACE models mishandled.
-
Time series Foundation Models based on Physics-Informed Synthetic Histories for Cold-Start Photovoltaic Forecasting
A zero-shot pipeline with physics-informed synthetic histories lets time-series foundation models outperform baselines by 1.7-2x in cold-start PV forecasting on 440 sites across four datasets.
-
Scalable Uncertainty Quantification for Extreme Weather Forecasting via Empirical Neural Tangent Kernels
NTK-UQ produces 31-37% sharper 90% prediction intervals than split conformal prediction for extreme weather forecasts, with adaptive scaling via architecture-dependent eigenvalue truncation and ICA decomposition of last-layer features.
-
Probabilistic storyline attribution using machine learning
Distributional autoencoders trained on climate model simulations model full conditional distributions of European temperature fields to enable probabilistic storyline attribution, illustrated by higher intensities and probability ratios for a 2003-like heatwave in 2028 and 2053.
-
No Epoch Like the Present: Robust Climate Emulation Requires Out-of-Distribution Generalisation
ML climate emulators degrade under seasonal distribution shifts that proxy long-term climate change, but physically motivated compositional decompositions improve out-of-distribution performance with modest in-distribution trade-offs.
-
AxiomOcean: Forecasting the Three-Dimensional Structure of the Upper Ocean
AxiomOcean deploys a 3D encoder-backbone-decoder architecture that jointly predicts upper-ocean variables and outperforms prior AI models by 20-35% in day-1 RMSE while preserving eddy kinetic energy and vertical consistency.
-
Controlling Transient Amplification Improves Long-horizon Rollouts
Commutativity regularization mitigates transient error amplification in autoregressive neural simulators by penalizing non-normality and non-commutativity of Jacobians, yielding stable long-horizon rollouts.
-
Predictive and Prescriptive AI toward Optimizing Wildfire Suppression
A new optimization algorithm with double machine learning for wildfire spread estimation enables better crew assignments that reduce total area burned.
-
M-CaStLe: Uncovering Local Causal Structures in Multivariate Space-Time Gridded Data
M-CaStLe generalizes local stencil-based causal discovery to the multivariate case and decomposes resulting graphs into reaction and spatial components for interpretation in space-time gridded data.
-
Earth System Foundation Model (ESFM): A unified framework for heterogeneous data integration and forecasting
ESFM is a single open foundation model that unifies heterogeneous Earth data sources and forecasts missing regions while preserving inter-variable physical relationships.
-
A PMP-inspired Evaluation Framework for Assessing Deep-Learning Earth System Models
Standard PMP diagnostics show DL-ESMs match CMIP-class skill on many large-scale fields but still fail key precipitation, monsoon, and long-run stability tests.
-
Non-stationary time series attribution for heatwaves over Europe
European summer heatwave time series are attributed to anthropogenic forcing through a non-stationary Markov extreme-value likelihood; the attributable signal is essentially a mean shift, not a change in temporal dependence or tail shape.
-
Otter Weather: Skillful and Computationally Efficient Medium-Range Weather Forecasting
Otter Weather is a spatiotemporal model that outperforms NWP baselines by 9.6% at 24h lead with under 3.5 A100-days training and extends efficiency gains to probabilistic forecasting via CRPS.
-
Towards Fair Comparisons of AI- and Physics-Based Weather Models for Extreme Events via the Weighted Potential CRPS
Extends Potential CRPS with weights and IDR post-processing to enable fair comparisons of AIWP and NWP models on extreme weather, finding AI models more informative across most variables and thresholds.
-
CMIP-Forge: An Agentic System that Retrieves, Computes, and Self-Reviews Climate Science
CMIP-Forge presents a retrieval-augmented agentic system with automated guardrails and adversarial self-review for autonomous execution of climate research tasks on CMIP6 literature and ESGF data.
-
Water vapor buoyancy and the African easterly jet
The negative meridional moisture gradient reduces African easterly jet magnitude by 30% through vapor buoyancy counteracting the temperature gradient in thermal wind balance, with the effect strengthening under global warming in CMIP6 data.
-
Interpretable Neural Networks to Predict Momentum Fluxes of Orographic Gravity Waves
Neural networks predict orographic gravity wave momentum fluxes from coarse state variables with offline R² of 0.56-0.72, learn physically meaningful relationships via SHAP, and are compared to the Lott-Miller parameterization.
-
Prediction and Predictability of the Wet-Season Rainfall over Southeast India
Wet-season rainfall over southeast India is increasing in amount and variability but shows potential predictability up to 10 months ahead from tropical sea surface temperature networks.
-
Dynamics of East Atlantic seed vortex populations in global km-scale models
The explicit-convection km-scale simulation produces fewer and weaker Atlantic hurricanes than parameterized coarser runs because seed vortices fail to amplify after crossing the West African coast due to weaker top-heavy mass flux profiles and underestimated MCS stratiform components.
-
Mechanistic Interpretability Tool for AI Weather Models
An open-source tool is developed for mechanistic interpretability of AI weather models, demonstrated on GraphCast by identifying latent directions corresponding to interpretable weather features.
-
Enhancing AI and Dynamical Subseasonal Forecasts with Probabilistic Bias Correction
Probabilistic bias correction doubles ECMWF AI subseasonal skill and improves debiased dynamical forecasts for most pressure, temperature, and precipitation targets, winning ECMWF’s 2025 real-time competition.
-
CycloneMAE: A Scalable Multi-Task Learning Model for Global Tropical Cyclone Probabilistic Forecasting
CycloneMAE uses a TC structure-aware masked autoencoder with discrete probabilistic gridding and pre-train/fine-tune to deliver both deterministic and probabilistic forecasts, outperforming NWP systems in pressure and wind up to 120 hours and track up to 24 hours across five basins.
-
Structural Gating and Effect-aligned Lag-resolved Temporal Causal Discovery Framework with Application to Heat-Pollution Extremes
SGED-TCD is a lag-resolved causal discovery framework that uses structural gating and perturbation-effect alignment to infer interpretable weighted causal networks from complex time series, shown on heat-pollution extremes in China.
-
U-Cast: A Surprisingly Simple and Efficient Frontier Probabilistic AI Weather Forecaster
A standard U-Net with MAE pre-training plus short CRPS fine-tuning and MC Dropout matches GenCast and IFS ENS probabilistic skill at 1.5° while cutting training and inference cost by over 10×.
-
Sampling Parallelism for Fast and Efficient Bayesian Learning
Sampling parallelism distributes Bayesian sample evaluations across GPUs for near-perfect scaling, lower memory use, and faster convergence via per-GPU data augmentations, outperforming pure data parallelism in diversity.
-
Event-Aware Loss Design for Forecasting of Convective Precipitation and Lightning
A multi-task Patch-cGAN with lightning-derived spatial loss weighting improves post-processed forecasts of intense precipitation and lightning occurrence over the Korean Peninsula in summer 2025.
-
AI and physics-based weather forecasting: A comparative study
Raw IFS forecasts outperform raw AIFS for wind speed at all horizons, but post-processing with EMOS or QR reduces the gap, leaving IFS ahead mainly at short leads.
-
Achieving Skilled and Reliable Daily Probabilistic Forecasts of Wind Power at Subseasonal-to-Seasonal Timescales over France
A post-processing pipeline applied to ECMWF subseasonal ensembles produces calibrated daily wind power forecasts for France that improve on climatology by 5-15% in CRPS up to 16 days ahead.
-
Modelling convective cell occurrence in proximity to cold fronts using extreme gradient boosting
An XGBoost model reproduces convective cell frequency near cold fronts with high skill but underestimates counts at the surface front, depending most on CAPE and time of day.
-
Comparison of Two Operational Microphysics Schemes Across Various Regional-MPAS Simulations
Comparison of NSSL and TEMPO microphysics in MPAS-A shows structural differences in convection but both schemes produce less organized storms and poorer rainfall matches to observations than to each other.
-
Reviewing methods and assumptions for high-resolution large-scale onshore wind energy potential assessments
A critical review of methods for estimating onshore wind energy potentials at multiple levels, with an attempt to derive best practice recommendations.