arXiv subjects

Mike Pritchard

Publications and source records attributed to Mike Pritchard.

18 recordsLinked to original sources

arXiv2026

Scaling Storm-Resolving Atmospheric AI Simulation to the Entire Planet

Kilometer-scale convection shapes precipitation extremes, tropical organization, and cloud feedbacks, but most global atmospheric models approximate these processes at 25-100 km resolution. Global storm-resolving physics models resolve convective systems explicitly, but at a cost -- roughly one MWh per simulated day on exascale supercomputers -- that limits long-duration simulation. We introduce STRATA (Storm-resolving Tile-based autoRegressive Atmosphere Transformer Architecture), the first autoregressive AI emulator for global storm-resolving atmospheric dynamics. STRATA is trained on the highest-resolution atmospheric dataset yet used for global AI emulation: 17 days of SCREAM physics-model output at 4.9-km resolution (~25 million grid cells) sampled every 10 minutes. Our central premise is that on 10-minute timescales atmospheric dynamics are predominantly local, so training on small spatial tiles trades scarce global temporal samples for abundant local spatial samples and enables global rollout via overlapping-tile blending. STRATA combines 3D patch embedding and local 3D neighborhood attention, a novel Stereographic Rotary Position Embedding (StereoRoPE) for grid-invariant encoding, and a pixel-space de-aliasing decoder that suppresses patch-scale rollout artifacts. An iso-FLOP scaling study reveals that km-scale emulation requires ~10x more FLOPs per grid point than coarse-resolution AI weather models, consistent with the higher information density of convective-scale dynamics. Trained on only 17 days of data, STRATA produces stable 24-hour global rollouts with realistic km-scale dynamics across diverse regimes, though large-scale biases develop with lead time. It achieves 48 simulation days per megawatt-hour -- about 50 times better energy efficiency than the SCREAM physics model -- and 741 simulated days per wall-clock day at 512 H100 GPUs. Code and dataset are publicly available.

physics.ao-ph↗

arXiv2026

ShardTensor: Domain Parallelism for Scientific Machine Learning

Scientific Machine Learning (SciML) faces unique challenges for extreme-resolution data, with mitigations that often fail to scale or degrade the accuracy of trained models. While some specialized methods have achieved remarkable results in training models or performing inference on massive spatial datasets with bespoke techniques, there is no generalized framework for parallelization over input data below batch size one per device. In this work we introduce ShardTensor: a novel paradigm of domain parallelism that enables flexible scaling of input data to arbitrary sizes. By decoupling the spatial dimensionality of input data from hardware constraints, ShardTensor enables scientific machine learning workloads to reach new levels of high fidelity training and inference. We demonstrate both strong and weak scaling of workloads during training and inference, showing improved latency with strong scaling and demonstrating the capacity to process higher data sizes with weak scaling. Additionally, we demonstrate multiple dimensions of parallelization, removing barriers to SciML on extreme-scale inputs.

cs.DC↗

arXiv2026

Towards accurate extreme event likelihoods from diffusion model climate emulators

ML climate model emulators are useful for scenario planning and adaptation, allowing for cost-efficient experimentation. Recently, the diffusion model Climate in a Bottle (cBottle) has been proposed for generation of atmospheric states compatible with boundary conditions of solar position and sea surface temperatures. Crucially, cBottle can be guided to generate extreme events such as Tropical Cyclones (TCs) over locations of interest. Diffusion models such as cBottle work by approximating the probability density of the training data. Here, we show use cases of the probability density estimates of atmospheric states obtained from this climate emulator. Most importantly, these estimates allow us to calculate likelihoods of extreme events under guidance. When guiding the model towards states including TCs, comparing the probability density under the guided and unguided model enables us to quantify how much more likely the guidance has made the TC. We show how these odds ratios allow us to importance-sample from the TC distribution, reducing the standard error of the probability estimate compared to simple Monte Carlo sampling. Furthermore, we discuss results and limitations of the application of model probability densities to extreme event attribution-like experiments. We present these early but encouraging results hoping they will spur more research into probabilistic information that can be gained from diffusion models of the atmosphere.

physics.ao-ph↗

arXiv2026

Demystifying Data-Driven Probabilistic Medium-Range Weather Forecasting

The recent revolution in data-driven methods for weather forecasting has lead to a fragmented landscape of complex, bespoke architectures and training strategies, obscuring the fundamental drivers of forecast accuracy. Here, we demonstrate that state-of-the-art probabilistic skill requires neither intricate architectural constraints nor specialized training heuristics. We introduce a scalable framework for learning multi-scale atmospheric dynamics by combining a directly downsampled latent space with a history-conditioned local projector that resolves high-resolution physics. We find that our framework design is robust to the choice of probabilistic estimator, seamlessly supporting stochastic interpolants, diffusion models, and CRPS-based ensemble training. Validated against the Integrated Forecasting System and the deep learning probabilistic model GenCast, our framework achieves statistically significant improvements on most of the variables. These results suggest scaling a general-purpose model is sufficient for state-of-the-art medium-range prediction, eliminating the need for tailored training recipes and proving effective across the full spectrum of probabilistic frameworks.

cs.LG↗

arXiv2026

Learning Accurate Storm-Scale Evolution from Observations

Accurate short-term prediction of clouds and precipitation is critical for severe weather warnings, aviation safety, and renewable energy operations. Forecasts at this timescale are provided by numerical weather models and extrapolation methods, both of which have limitations. Mesoscale numerical weather prediction models provide skillful forecasts at these scales but require significant modeling expertise and computational infrastructure, which limits their accessibility. Extrapolation-based methods are computationally lightweight but degrade rapidly beyond 1-2 hours. This presents an opportunity for data-driven forecasting directly from observations using geostationary satellites and ground-based radar, which provide high-frequency, high-resolution observations that capture mesoscale atmospheric evolution. We introduce Stormscope, a family of transformer-based generative diffusion models trained on high-resolution, multi-band geostationary satellite imagery and ground-based weather radar over the continental United States. Stormscope produces forecasts at a temporal resolution of 10 minutes and 6 kilometer spatial resolution, which are competitive with state-of-the-art mesoscale NWP models for lead times up to 6 hours. Its generative architecture enables large ensemble forecasts of explicit mesoscale dynamics for robust uncertainty quantification. Evaluated against extrapolation methods and operational mesoscale NWP models, Stormscope achieves leading performance on standard deterministic and probabilistic verification metrics across forecast horizons from 1 to 6 hours. As Stormscope relies on globally available satellite observations (and radar where available), it offers a pathway to extend skillful mesoscale forecasting to oceanic regions and countries without strong operational mesoscale modeling programs.

physics.ao-ph↗

arXiv2025

Crowdsourcing the Frontier: Advancing Hybrid Physics-ML Climate Simulation via a $50,000 Kaggle Competition

Subgrid machine-learning (ML) parameterizations have the potential to introduce a new generation of climate models that incorporate the effects of higher-resolution physics without incurring the prohibitive computational cost associated with more explicit physics-based simulations. However, important issues, ranging from online instability to inconsistent online performance, have limited their operational use for long-term climate projections. To more rapidly drive progress in solving these issues, domain scientists and machine learning researchers opened up the offline aspect of this problem to the broader machine learning and data science community with the release of ClimSim, a NeurIPS Datasets and Benchmarks publication, and an associated Kaggle competition. This paper reports on the downstream results of the Kaggle competition by coupling emulators inspired by the winning teams' architectures to an interactive climate model (including full cloud microphysics, a regime historically prone to online instability) and systematically evaluating their online performance. Our results demonstrate that online stability in the low-resolution, real-geography setting is reproducible across multiple diverse architectures, which we consider a key milestone. All tested architectures exhibit strikingly similar offline and online biases, though their responses to architecture-agnostic design choices (e.g., expanding the list of input variables) can differ significantly. Multiple Kaggle-inspired architectures achieve state-of-the-art (SOTA) results on certain metrics such as zonal mean bias patterns and global RMSE, indicating that crowdsourcing the essence of the offline problem is one path to improving online performance in hybrid physics-AI climate simulation.

physics.ao-ph↗

arXiv2025

Diagnosing Biases in Tropical Atlantic-Pacific Multi-Decadal Teleconnections Across CMIP6 and E3SM Models

Decadal-scale interactions between the tropical Atlantic and Pacific Oceans play a crucial role in global climate variability through bidirectional teleconnections. Current climate models show persistent biases in representing these basin interactions, particularly in simulating the Atlantic's influence on Pacific climate. Using historical simulations from 27 CMIP6 models and two configurations of the Energy Exascale Earth System Model (E3SM) during 1950-2015, we systematically evaluate tropical Atlantic-Pacific teleconnections through both Walker circulation and extratropical wave responses. Most models exhibit Pacific-dominated teleconnections, contradicting observational evidence of Atlantic control on Pacific variability during the past 40 years. By developing a performance metric that combines tropical circulation patterns and extratropical wave propagation, we identify two distinct model behaviors: high-skill models capture the bidirectional Atlantic-Pacific teleconnections with a secondary symptom of systematic 20-degree westward shifts in convective centers, while low-skill models display amplified Pacific dominance through reversed Walker circulation responses warming in both tropical basins. Comparative analysis between standard E3SMv2 and its multi-scale modeling framework configuration demonstrates that implementing more sophisticated cloud-scale processes alone, with limited model tuning, cannot resolve these teleconnection biases. Our results identify four CMIP6 models and E3SMv2 that effectively reproduce observed teleconnection pathways, offering a comprehensive diagnostic framework for evaluating decadal Atlantic-Pacific interactions in climate models.

physics.ao-ph↗

arXiv2024

Generative Data Assimilation of Sparse Weather Station Observations at Kilometer Scales

Data assimilation of observational data into full atmospheric states is essential for weather forecast model initialization. Recently, methods for deep generative data assimilation have been proposed which allow for using new input data without retraining the model. They could also dramatically accelerate the costly data assimilation process used in operational regional weather models. Here, in a central US testbed, we demonstrate the viability of score-based data assimilation in the context of realistically complex km-scale weather. We train an unconditional diffusion model to generate snapshots of a state-of-the-art km-scale analysis product, the High Resolution Rapid Refresh. Then, using score-based data assimilation to incorporate sparse weather station data, the model produces maps of precipitation and surface winds. The generated fields display physically plausible structures, such as gust fronts, and sensitivity tests confirm learnt physics through multivariate relationships. Preliminary skill analysis shows the approach already outperforms a naive baseline of the High-Resolution Rapid Refresh system itself. By incorporating observations from 40 weather stations, 10% lower RMSEs on left-out stations are attained. Despite some lingering imperfections such as insufficiently disperse ensemble DA estimates, we find the results overall an encouraging proof of concept, and the first at km-scale. It is a ripe time to explore extensions that combine increasingly ambitious regional state generators with an increasing set of in situ, ground-based, and satellite remote sensing data streams.

cs.LG↗

arXiv2024

DiffObs: Generative Diffusion for Global Forecasting of Satellite Observations

This work presents an autoregressive generative diffusion model (DiffObs) to predict the global evolution of daily precipitation, trained on a satellite observational product, and assessed with domain-specific diagnostics. The model is trained to probabilistically forecast day-ahead precipitation. Nonetheless, it is stable for multi-month rollouts, which reveal a qualitatively realistic superposition of convectively coupled wave modes in the tropics. Cross-spectral analysis confirms successful generation of low frequency variations associated with the Madden--Julian oscillation, which regulates most subseasonal to seasonal predictability in the observed atmosphere, and convectively coupled moist Kelvin waves with approximately correct dispersion relationships. Despite secondary issues and biases, the results affirm the potential for a next generation of global diffusion models trained on increasingly sparse, and increasingly direct and differentiated observations of the world, for practical applications in subseasonal and climate prediction.

physics.comp-ph↗

arXiv2023

Navigating the Noise: Bringing Clarity to ML Parameterization Design with O(100) Ensembles

Machine-learning (ML) parameterizations of subgrid processes (here of turbulence, convection, and radiation) may one day replace conventional parameterizations by emulating high-resolution physics without the cost of explicit simulation. However, uncertainty about the relationship between offline and online performance (i.e., when integrated with a large-scale general circulation model (GCM)) hinders their development. Much of this uncertainty stems from limited sampling of the noisy, emergent effects of upstream ML design decisions on downstream online hybrid simulation. Our work rectifies the sampling issue via the construction of a semi-automated, end-to-end pipeline for $\mathcal{O}(100)$ size ensembles of hybrid simulations, revealing important nuances in how systematic reductions in offline error manifest in changes to online error and online stability. For example, removing dropout and switching from a Mean Squared Error (MSE) to a Mean Absolute Error (MAE) loss both reduce offline error, but they have opposite effects on online error and online stability. Other design decisions, like incorporating memory, converting moisture input from specific humidity to relative humidity, using batch normalization, and training on multiple climates do not come with any such compromises. Finally, we show that ensemble sizes of $\mathcal{O}(100)$ may be necessary to reliably detect causally relevant differences online. By enabling rapid online experimentation at scale, we can empirically settle debates regarding subgrid ML parameterization design that would have otherwise remained unresolved in the noise.

physics.ao-ph↗

arXiv2023

Residual Corrective Diffusion Modeling for Km-scale Atmospheric Downscaling

The state of the art for physical hazard prediction from weather and climate requires expensive km-scale numerical simulations driven by coarser resolution global inputs. Here, a generative diffusion architecture is explored for downscaling such global inputs to km-scale, as a cost-effective machine learning alternative. The model is trained to predict 2km data from a regional weather model over Taiwan, conditioned on a 25km global reanalysis. To address the large resolution ratio, different physics involved at different scales and prediction of channels beyond those in the input data, we employ a two-step approach where a UNet predicts the mean and a corrector diffusion (CorrDiff) model predicts the residual. CorrDiff exhibits encouraging skill in bulk MAE and CRPS scores. The predicted spectra and distributions from CorrDiff faithfully recover important power law relationships in the target data. Case studies of coherent weather phenomena show that CorrDiff can help sharpen wind and temperature gradients that co-locate with intense rainfall in cold front, and can help intensify typhoons and synthesize rain band structures. Calibration of model uncertainty remains challenging. The prospect of unifying methods like CorrDiff with coarser resolution global weather models implies a potential for global-to-regional multi-scale machine learning simulation.

cs.LG↗

arXiv2023

Two-step hyperparameter optimization method: Accelerating hyperparameter search by using a fraction of a training dataset

Hyperparameter optimization (HPO) is an important step in machine learning (ML) model development, but common practices are archaic -- primarily relying on manual or grid searches. This is partly because adopting advanced HPO algorithms introduces added complexity to the workflow, leading to longer computation times. This poses a notable challenge to ML applications, as suboptimal hyperparameter selections curtail the potential of ML model performance, ultimately obstructing the full exploitation of ML techniques. In this article, we present a two-step HPO method as a strategic solution to curbing computational demands and wait times, gleaned from practical experiences in applied ML parameterization work. The initial phase involves a preliminary evaluation of hyperparameters on a small subset of the training dataset, followed by a re-evaluation of the top-performing candidate models post-retraining with the entire training dataset. This two-step HPO method is universally applicable across HPO search algorithms, and we argue it has attractive efficiency gains. As a case study, we present our recent application of the two-step HPO method to the development of neural network emulators for aerosol activation. Although our primary use case is a data-rich limit with many millions of samples, we also find that using up to 0.0025% of the data (a few thousand samples) in the initial step is sufficient to find optimal hyperparameter configurations from much more extensive sampling, achieving up to 135-times speedup. The benefits of this method materialize through an assessment of hyperparameters and model performance, revealing the minimal model complexity required to achieve the best performance. The assortment of top-performing models harvested from the HPO process allows us to choose a high-performing model with a low inference cost for efficient use in global climate models (GCMs).

cs.LG↗

arXiv2023

Using Neural Networks to Learn the Jet Stream Forced Response from Natural Variability

Two distinct features of anthropogenic climate change, warming in the tropical upper troposphere and warming at the Arctic surface, have competing effects on the mid-latitude jet stream's latitudinal position, often referred to as a "tug-of-war". Studies that investigate the jet's response to these thermal forcings show that it is sensitive to model type, season, initial atmospheric conditions, and the shape and magnitude of the forcing. Much of this past work focuses on studying a simulation's response to external manipulation. In contrast, we explore the potential to train a convolutional neural network (CNN) on internal variability alone and then use it to examine possible nonlinear responses of the jet to tropospheric thermal forcing that more closely resemble anthropogenic climate change. Our approach leverages the idea behind the fluctuation-dissipation theorem, which relates the internal variability of a system to its forced response but so far has been only used to quantify linear responses. We train a CNN on data from a long control run of the CESM dry dynamical core and show that it is able to skillfully predict the nonlinear response of the jet to sustained external forcing. The trained CNN provides a quick method for exploring the jet stream sensitivity to a wide range of tropospheric temperature tendencies and, considering that this method can likely be applied to any model with a long control run, could lend itself useful for early stage experiment design.

physics.ao-ph↗

arXiv2022

Understanding Extreme Precipitation Changes through Unsupervised Machine Learning

Despite the importance of quantifying how the spatial patterns of extreme precipitation will change with warming, we lack tools to objectively analyze the storm-scale outputs of modern climate models. To address this gap, we develop an unsupervised machine learning framework to quantify how storm dynamics affect changes in precipitation extremes, without sacrificing spatial information. For the upper precipitation quantiles (above the 80th percentile), we find that the spatial patterns of extreme precipitation changes are dominated by spatial shifts in storm dynamical regimes rather than changes in how these storm regimes produce precipitation. Our study shows how unsupervised machine learning, paired with domain knowledge, may allow us to better understand the physics of the atmosphere and anticipate the changes associated with a warming world.

physics.ao-ph↗

arXiv2022

Comparing Storm Resolving Models and Climates via Unsupervised Machine Learning

Global Storm-Resolving Models (GSRMs) have gained widespread interest because of the unprecedented detail with which they resolve the global climate. However, it remains difficult to quantify objective differences in how GSRMs resolve complex atmospheric formations. This lack of comprehensive tools for comparing model similarities is a problem in many disparate fields that involve simulation tools for complex data. To address this challenge we develop methods to estimate distributional distances based on both nonlinear dimensionality reduction and vector quantization. Our approach automatically learns physically meaningful notions of similarity from low-dimensional latent data representations that the different models produce. This enables an intercomparison of nine GSRMs based on their high-dimensional simulation data (2D vertical velocity snapshots) and reveals that only six are similar in their representation of atmospheric dynamics. Furthermore, we uncover signatures of the convective response to global warming in a fully unsupervised way. Our study provides a path toward evaluating future high-resolution simulation data more objectively.

physics.ao-ph↗

arXiv2021

Analyzing High-Resolution Clouds and Convection using Multi-Channel VAEs

Understanding the details of small-scale convection and storm formation is crucial to accurately represent the larger-scale planetary dynamics. Presently, atmospheric scientists run high-resolution, storm-resolving simulations to capture these kilometer-scale weather details. However, because they contain abundant information, these simulations can be overwhelming to analyze using conventional approaches. This paper takes a data-driven approach and jointly embeds spatial arrays of vertical wind velocities, temperatures, and water vapor information as three "channels" of a VAE architecture. Our "multi-channel VAE" results in more interpretable and robust latent structures than earlier work analyzing vertical velocities in isolation. Analyzing and clustering the VAE's latent space identifies weather patterns and their geographical manifestations in a fully unsupervised fashion. Our approach shows that VAEs can play essential roles in analyzing high-dimensional simulation data and extracting critical weather and climate characteristics.

physics.ao-ph↗

arXiv2021

Assessing the Potential of Deep Learning for Emulating Cloud Superparameterization in Climate Models with Real-Geography Boundary Conditions

We explore the potential of feed-forward deep neural networks (DNNs) for emulating cloud superparameterization in realistic geography, using offline fits to data from the Super Parameterized Community Atmospheric Model. To identify the network architecture of greatest skill, we formally optimize hyperparameters using ~250 trials. Our DNN explains over 70 percent of the temporal variance at the 15-minute sampling scale throughout the mid-to-upper troposphere. Autocorrelation timescale analysis compared against DNN skill suggests the less good fit in the tropical, marine boundary layer is driven by neural network difficulty emulating fast, stochastic signals in convection. However, spectral analysis in the temporal domain indicates skillful emulation of signals on diurnal to synoptic scales. A close look at the diurnal cycle reveals correct emulation of land-sea contrasts and vertical structure in the heating and moistening fields, but some distortion of precipitation. Sensitivity tests targeting precipitation skill reveal complementary effects of adding positive constraints vs. hyperparameter tuning, motivating the use of both in the future. A first attempt to force an offline land model with DNN emulated atmospheric fields produces reassuring results further supporting neural network emulation viability in real-geography settings. Overall, the fit skill is competitive with recent attempts by sophisticated Residual and Convolutional Neural Network architectures trained on added information, including memory of past states. Our results confirm the parameterizability of superparameterized convection with continents through machine learning and we highlight advantages of casting this problem locally in space and time for accurate emulation and hopefully quick implementation of hybrid climate models.

physics.ao-ph↗

arXiv2020

A Fortran-Keras Deep Learning Bridge for Scientific Computing

Implementing artificial neural networks is commonly achieved via high-level programming languages like Python and easy-to-use deep learning libraries like Keras. These software libraries come pre-loaded with a variety of network architectures, provide autodifferentiation, and support GPUs for fast and efficient computation. As a result, a deep learning practitioner will favor training a neural network model in Python, where these tools are readily available. However, many large-scale scientific computation projects are written in Fortran, making it difficult to integrate with modern deep learning methods. To alleviate this problem, we introduce a software library, the Fortran-Keras Bridge (FKB). This two-way bridge connects environments where deep learning resources are plentiful, with those where they are scarce. The paper describes several unique features offered by FKB, such as customizable layers, loss functions, and network ensembles. The paper concludes with a case study that applies FKB to address open questions about the robustness of an experimental approach to global climate simulation, in which subgrid physics are outsourced to deep neural network emulators. In this context, FKB enables a hyperparameter search of one hundred plus candidate models of subgrid cloud and radiation physics, initially implemented in Keras, to be transferred and used in Fortran. Such a process allows the model's emergent behavior to be assessed, i.e. when fit imperfections are coupled to explicit planetary-scale fluid dynamics. The results reveal a previously unrecognized strong relationship between offline validation error and online performance, in which the choice of optimizer proves unexpectedly critical. This reveals many neural network architectures that produce considerable improvements in stability including some with reduced error, for an especially challenging training dataset.

cs.LG↗