Peter Ukkonen PhD

Abstract:

Abstract Subgrid machine‐learning (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 ML researchers opened up the offline aspect of this problem to the broader ML and data science community with the release of ClimSim, a NeurIPS Data sets 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 results on certain metrics such as zonal mean bias patterns and global Root Mean Squared Error, indicating that crowdsourcing the essence of the offline problem is one path to improving online performance in hybrid physics‐AI climate simulation. Plain Language Summary Future climate models may use machine learning (ML) to replace small‐scale physical processes that are otherwise too costly to simulate directly over long timescales. Such “hybrid” physics–ML models could improve predictions by reducing uncertainties from current approximations. But making them run reliably in full climate simulations has been a major challenge. To speed progress, scientists created an open data set, benchmarking framework, and global competition to drive improvement for these ML components. This paper follows up on that competition by testing ideas from the winning teams within hybrid climate models. For the first time, we show that stable hybrid simulation is now reproducible across a range of diverse ML architectures. We find that different architectures share similar patterns of errors both before and after coupling, although their responses to added training inputs can differ. Finally, some competition‐inspired designs achieve state‐of‐the‐art scores on individual performance measures, but no single approach beats the previous benchmark (Hu et al., 2025, https://doi.org/10.1029/2024ms004618 ) on every metric. Key Points Online stability in the low‐resolution real‐geography setting is reproducibly achievable across diverse architectures Offline and online zonal mean biases are near‐identical across architectures; online runs underestimate tropical precipitable water An expanded variable list is universally beneficial offline but has diverging, architecture‐dependent effects online

Abstract:

Machine learning hold the promise of unlocking more accurate and realistic parameterizations of atmospheric processes, but brings its own set of challenges and drawbacks. Among top issues are generalization, stability and interpretability. Here we present a parameter-efficient neural network parameterization which aims to address these issues by incorporating physical knowledge to a high degree. By predicting fluxes and microphysical process rates instead of total tendencies, the conservation of water can be hardcoded, which is shown to improve online performance. Furthermore, a physically motivated architecture based on vertically recurrent neural networks enables high computational efficiency and a low number of parameters. The models are trained and evaluated using a superparameterization setup with real orography. The impact of incorporating stochasticity is also discussed.

Abstract:

Machine learning has the potential to improve the physical realism and/or computational efficiency of parameterizations. A typical approach has been to feed concatenated vertical profiles to a dense neural network. However, feed‐forward networks lack the connections to propagate information sequentially through the vertical column. Here we examine if predictions can be improved by instead traversing the column with recurrent neural networks (RNNs) such as Long Short‐Term Memory (LSTMs). This method encodes physical priors (locality) and uses parameters more efficiently. Firstly, we test RNN‐based radiation emulators in the Integrated Forecasting System. We achieve near‐perfect offline accuracy, and the forecast skill of a suite of global weather simulations using the emulator are for the most part statistically indistinguishable from reference runs. But can radiation emulators provide both high accuracy and a speed‐up? We find optimized, state‐of‐the‐art radiation code on CPU generally faster than RNN‐based emulators on GPU, although the latter can be more energy efficient. To test the method more broadly, and explore recent challenges in parameterization, we also adapt it to data sets from other studies. RNNs outperform reference feed‐forward networks in emulating gravity waves, and when combined with horizontal convolutions, for non‐local unified parameterization. In emulation of moist physics with memory, the RNNs have similar offline accuracy as ResNets, the previous state‐of‐the‐art. However, the RNNs are more efficient, and more stable in autoregressive semi‐prognostic tests. Multi‐step autoregressive training improves performance in these tests and enables a latent representation of convective memory. Recently proposed linearly recurrent models achieve similar performance to LSTMs.

Abstract:

The aim of this technical note is to describe the Cycle 46 reference configuration of the HARMONIE-AROME convection-permitting numerical weather prediction model. HARMONIE-AROME is one of the canonical system configurations that is developed, maintained, and validated in the ACCORD consortium, a collaboration of 26 countries in Europe and northern Africa on short-range mesoscale numerical weather prediction. This technical note describes updates to the physical parametrizations, both upper-air and surface, configuration choices such as lateral boundary conditions, model levels, horizontal resolution, model time step, and databases associated with the model, such as for physiography and aerosols. Much of the physics developments are related to improving the representation of clouds in the model, including developments in the turbulence, shallow convection, and statistical cloud scheme, as well as changes in radiation and cloud microphysics concerning cloud droplet number concentration and longwave cloud liquid optical properties. Near real-time aerosols and the ICE-T microphysics scheme, which improves the representation of supercooled liquid, and a wind farm parametrization have been added as options. Surface-wise, one of the main advances is the implementation of the lake model FLake. An outlook on upcoming developments is also included.