---
title: "Scale-aware and spatially-localised optimisation of kilometre-scale MLWP models"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2610.37325v1"
permalink: "https://feed.astralor.com/items/itm_dfe9070232e2795c35a9c586"
published: "2026-10-27T14:35:19.000Z"
score: 52
channel: "ai"
tags: ["推理效率","Scale-aware","MLWP","Preserving","FastNet","Relative"]
---

# Scale-aware and spatially-localised optimisation of kilometre-scale MLWP models

## At a glance

Preserving fine-scale structure without degrading spatial accuracy is a central challenge for kilometre-scale machine-learning weather prediction (MLWP). We introduce a loss based on the dual-tree complex wavelet transf…

## Key points

- Preserving fine-scale structure without degrading spatial accuracy is a central challenge for kilometre-scale machine-learning weather prediction (MLWP).
- We introduce a loss based on the dual-tree complex wavelet transform that constrains atmospheric fields jointly by scale, location, and orientation.
- We evaluate its performance using an experimental joint global--regional FastNet configuration, which produces \SI{1.5}{km} forecasts over the UK alongside $\sim$\SI{31}{km} global forecasts.

## Full text

Preserving fine-scale structure without degrading spatial accuracy is a central challenge for kilometre-scale machine-learning weather prediction (MLWP). We introduce a loss based on the dual-tree complex wavelet transform that constrains atmospheric fields jointly by scale, location, and orientation. We evaluate its performance using an experimental joint global--regional FastNet configuration, which produces \SI{1.5}{km} forecasts over the UK alongside $\sim$\SI{31}{km} global forecasts. Relative to both mean-squared error (MSE) and an FFT-based spectral objective, the wavelet loss substantially improves spectral fidelity. Compared with MSE training, it maintains similar pointwise performance, improves neighbourhood-based verification, and produces sharper precipitation fields with more realistic intensity distributions. These results demonstrate that spatially localised wavelet constraints can preserve multiscale atmospheric structure while largely avoiding the loss of pointwise accuracy associated with globally supported spectral objectives.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_dfe9070232e2795c35a9c586