STRATA is the first autoregressive transformer emulator for global 4.9-km storm-resolving atmospheric dynamics, achieving 50x better energy efficiency than the underlying physics model while producing realistic km-scale features in 24-hour forecasts.
hub
Brenner, and Stephan Hoyer
20 Pith papers cite this work, alongside 337 external citations. Polarity classification is still indexing.
20
Pith papers citing it
337
external citations · OpenAlex
hub tools
representative citing papers
Optimizing training data via a differentiable SCM yields climate emulators that outperform those trained on six standard ScenarioMIP pathways while using less data and isolating distinct forcing responses.
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.
Online conformal prediction post-processing guarantees calibrated uncertainty coverage for GenCast, NeuralGCM, and AIFS-ENS forecasts of temperature and precipitation including extremes.
PhysMetrics.Weather is an evaluation framework that quantifies physical realism of ML weather prediction models using conservation, spectral, and dynamical metrics.
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.
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.
Mechanism learning infers active local evolution rules via prototype-anchored descriptors to achieve more robust forecasting than direct state prediction on benchmarks like Burgers, WeatherBench2, and Lorenz96.
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.
HealDA supplies ML-based initial conditions for AI weather models that produce forecasts trailing ERA5-initialized runs by less than one day of effective lead time, with the skill gap arising mainly from initial error size.
EnScale emulates high-resolution regional climate model outputs from global circulation models for multiple variables using a two-step generative process with sparse local stochastic layers and energy score optimization, including a temporally consistent variant.
A physics-constrained consistency model downscales Greenland SMB and surface temperature by a factor of 32 while preserving coarse-scale sums and outperforming interpolation on test metrics.
Diffusion models recover known ENSO variability structure from synthetic LIM data when given enough samples, but require pre-training on CMIP6 plus fine-tuning to match observations with the ~700 samples available in ERSSTv5.
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.
Explicit 3D vertical coupling emerges as a key inductive bias for ML emulation of SSW dynamics in idealised Isca runs, though prediction skill does not guarantee physical fidelity in wave driving.
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×.
GraphCast shows regime-dependent skill versus ECMWF HRES in Brazil, underperforming on winter baroclinic systems in medium range but gaining in extended range and summer moisture transport.
A flow matching deep learning emulator trained on multiple SSPs generates forced climate responses for unseen scenarios and is validated against the MESMER-M statistical emulator.
GPU port of entropy-stable DG Euler solver with non-conservative buoyancy terms reaches nearly 70% of 64-bit peak on A100 volume kernels, delivers 10x speedup and 13x better energy efficiency versus CPU, and preserves symmetry-based flux savings.
Machine learning success in weather prediction will drive changes in development practices, data handling, verification, and service creation at weather centers.
citing papers explorer
-
Scaling Storm-Resolving Atmospheric AI Simulation to the Entire Planet
STRATA is the first autoregressive transformer emulator for global 4.9-km storm-resolving atmospheric dynamics, achieving 50x better energy efficiency than the underlying physics model while producing realistic km-scale features in 24-hour forecasts.
-
Optimal scenario design for climate emulation
Optimizing training data via a differentiable SCM yields climate emulators that outperform those trained on six standard ScenarioMIP pathways while using less data and isolating distinct forcing responses.
-
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.
-
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.
-
PhysMetrics.Weather: An Evaluation Framework for Physical Consistency in ML Weather Models
PhysMetrics.Weather is an evaluation framework that quantifies physical realism of ML weather prediction models using conservation, spectral, and dynamical metrics.
-
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.
-
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.
-
Mechanism Learning: Prototype-Anchored Mechanism Inference for Scientific Forecasting
Mechanism learning infers active local evolution rules via prototype-anchored descriptors to achieve more robust forecasting than direct state prediction on benchmarks like Burgers, WeatherBench2, and Lorenz96.
-
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.
-
HealDA: Highlighting the importance of initial errors in end-to-end AI weather forecasts
HealDA supplies ML-based initial conditions for AI weather models that produce forecasts trailing ERA5-initialized runs by less than one day of effective lead time, with the skill gap arising mainly from initial error size.
-
EnScale: Temporally-consistent multivariate generative downscaling via proper scoring rules
EnScale emulates high-resolution regional climate model outputs from global circulation models for multiple variables using a two-step generative process with sparse local stochastic layers and energy score optimization, including a temporally consistent variant.
-
Physics-constrained generative machine learning-based high-resolution downscaling of Greenland's surface mass balance and surface temperature
A physics-constrained consistency model downscales Greenland SMB and surface temperature by a factor of 32 while preserving coarse-scale sums and outperforming interpolation on test metrics.
-
Learning Climate Variability from Scarce Data with Diffusion Models: A Test Case for ENSO
Diffusion models recover known ENSO variability structure from synthetic LIM data when given enough samples, but require pre-training on CMIP6 plus fine-tuning to match observations with the ~700 samples available in ERSSTv5.
-
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.
-
Investigating Inductive Biases for Machine Learning Emulation of Sudden Stratospheric Warmings in Idealised Isca Simulations
Explicit 3D vertical coupling emerges as a key inductive bias for ML emulation of SSW dynamics in idealised Isca runs, though prediction skill does not guarantee physical fidelity in wave driving.
-
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×.
-
Performance Evaluation of GraphCast for Medium-Range Weather Forecasting over Brazil
GraphCast shows regime-dependent skill versus ECMWF HRES in Brazil, underperforming on winter baroclinic systems in medium range but gaining in extended range and summer moisture transport.
-
Emulating the Forced Response of Climate Models with Flow Matching
A flow matching deep learning emulator trained on multiple SSPs generates forced climate responses for unseen scenarios and is validated against the MESMER-M statistical emulator.
-
GPU Performance of an Entropy-Stable Discontinuous Galerkin Euler Solver with Non-Conservative Terms
GPU port of entropy-stable DG Euler solver with non-conservative buoyancy terms reaches nearly 70% of 64-bit peak on A100 volume kernels, delivers 10x speedup and 13x better energy efficiency versus CPU, and preserves symmetry-based flux savings.
-
Machine learning is revolutionizing weather forecasting -- the next step is a change in how we work
Machine learning success in weather prediction will drive changes in development practices, data handling, verification, and service creation at weather centers.