Saved in:
Bibliographic Details
Main Authors: Redondo, Yago del Valle Inclan, Arriaga-Varela, Enrique, Lyamzin, Dmitry, Cervantes, Pablo, Ramalho, Tiago
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.20228
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911228084879360
author Redondo, Yago del Valle Inclan
Arriaga-Varela, Enrique
Lyamzin, Dmitry
Cervantes, Pablo
Ramalho, Tiago
author_facet Redondo, Yago del Valle Inclan
Arriaga-Varela, Enrique
Lyamzin, Dmitry
Cervantes, Pablo
Ramalho, Tiago
contents We introduce SpLIIF to generate implicit neural representations and enable arbitrary downscaling of weather variables. We train a model from sparse weather stations and topography over Japan and evaluate in- and out-of-distribution accuracy predicting temperature and wind, comparing it to both an interpolation baseline and CorrDiff. We find the model to be up to 50% better than both CorrDiff and the baseline at downscaling temperature, and around 10-20% better for wind.
format Preprint
id arxiv_https___arxiv_org_abs_2510_20228
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Sparse Local Implicit Image Function for sub-km Weather Downscaling
Redondo, Yago del Valle Inclan
Arriaga-Varela, Enrique
Lyamzin, Dmitry
Cervantes, Pablo
Ramalho, Tiago
Machine Learning
We introduce SpLIIF to generate implicit neural representations and enable arbitrary downscaling of weather variables. We train a model from sparse weather stations and topography over Japan and evaluate in- and out-of-distribution accuracy predicting temperature and wind, comparing it to both an interpolation baseline and CorrDiff. We find the model to be up to 50% better than both CorrDiff and the baseline at downscaling temperature, and around 10-20% better for wind.
title Sparse Local Implicit Image Function for sub-km Weather Downscaling
topic Machine Learning
url https://arxiv.org/abs/2510.20228