Unmasking Compensating Errors in Numerical Weather Models: A Process-Additive Neural Network Approach
Abstract
This work introduces a Process Additive Neural Network (PANN) framework to disentangle compensating biases in atmospheric models, where errors in different physical processes can offset each other and obscure their origins. Using a single-column model, each PANN branch represents the tendency error associated with a separate physical process, and the summed branch outputs approximate the total error relative to a reference tendency. Once trained successfully, the branch outputs are expected to provide process-level error attributions and, when averaged, estimates of process-level bias. Initial experiments focus on cloud-related processes in OpenIFS SCM case studies, where cloud and convection tendencies are strongly coupled and compensation is difficult to diagnose using standard approaches. The aim is not to replace human interpretation of physical processes, but to extend it to regimes where model complexity, process interaction, and compensating errors make direct diagnosis difficult. By providing a physically structured form of machine-learning interpretability, PANN can help connect complex model behaviour with process-level explanations. This work supports both improved understanding of compensating errors and the development of more physically consistent weather and climate models.
Related
- Probabilistic AI models outperform numerical weather prediction models on record breaking heat extremes
- Comparison of data-driven and numerical weather prediction models for rainfall nowcasting in major cities of Bangladesh
- Impact of Two Radio Occultation Constellations on Numerical Weather Prediction: Polar‐Orbits Only and Mixed Polar‐Equatorial Orbits
- WRF-JAX v1.1.28: a differentiable regional weather model with fully coupled physics, numerically consistent with WRF and matching 128-process CPU performance on a single GPU
- Fragile against what? Imagining climate change resilience by unmasking governance fragility.
BibTeX
@misc{ichikawa2026unmasking,
title = {{Unmasking Compensating Errors in Numerical Weather Models: A Process-Additive Neural Network Approach}},
author = {Yuiko Ichikawa and Edward Groot and Jack Atkinson},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.19772},
eprint = {2610.19772},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.19772},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Unmasking Compensating Errors in Numerical Weather Models: A Process-Additive Neural Network Approach
AU - Yuiko Ichikawa
AU - Edward Groot
AU - Jack Atkinson
PY - 2026
DA - 2026/10/14
PB - arXiv
DO - 10.48550/arxiv.2610.19772
UR - https://arxiv.org/abs/2610.19772
AB - This work introduces a Process Additive Neural Network (PANN) framework to disentangle compensating biases in atmospheric models, where errors in different physical processes can offset each other and obscure their origins. Using a single-column model, each PANN branch represents the tendency error associated with a separate physical process, and the summed branch outputs approximate the total error relative to a reference tendency. Once trained successfully, the branch outputs are expected to provide process-level error attributions and, when averaged, estimates of process-level bias. Initial experiments focus on cloud-related processes in OpenIFS SCM case studies, where cloud and convection tendencies are strongly coupled and compensation is difficult to diagnose using standard approaches. The aim is not to replace human interpretation of physical processes, but to extend it to regimes where model complexity, process interaction, and compensating errors make direct diagnosis difficult. By providing a physically structured form of machine-learning interpretability, PANN can help connect complex model behaviour with process-level explanations. This work supports both improved understanding of compensating errors and the development of more physically consistent weather and climate models.
ER -