Detecting Weather Fronts from Satellite Observations: What Does Reanalysis Add?
Abstract
Weather fronts are transition regions between air masses with contrasting thermodynamic properties and are associated with mid-latitude precipitation and severe weather. We investigate machine learning (ML) methods for detecting weather fronts and quantify the contribution of reanalysis data to performance. We compare two configurations of the same ML model for pixel-wise frontal classification over the North Atlantic and Europe: (1) using only geostationary infrared satellite observations, and (2) additionally using 28 ERA5 reanalysis fields. The satellite-only model achieves a neighbourhood critical success index at a 150 km tolerance (CSI@150) of $0.402$, which lies within the range of agreement between analyses from different meteorological services. The satellite+ERA5 model achieves $0.575$, an improvement of $0.173$. Our analysis shows that low-level winds are the least predictable from the satellite representation (R2 = 0.23, compared with $0.88$ for temperature), while removing winds or moisture causes the largest decreases in detection performance.
Related
- SATFRONT: Detecting Atmospheric Fronts with Satellite Data using Machine Learning
- Improving global precipitation forecasts with an AI weather model trained on satellite observations
- All-weather lake-ice phenology from Sentinel-1 and reanalysis thermodynamics: a validated, transferable method applied to 131 Adirondack lakes
- Satellite observations and energy balance constrain the thermal footprint of a hyperscale data-centre cluster
- Spatial variation of aerosol properties over Europe derived from satellite observations and comparison with model calculations
BibTeX
@misc{dravilas2026detecting,
title = {{Detecting Weather Fronts from Satellite Observations: What Does Reanalysis Add?}},
author = {Ioannis Dravilas and Miruna Stoicescu and Roope Tervo and Carlos Horn and Oriol Hinojo Comellas and Sina Montazeri and Andrea Meraner and Ana Lucic},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.08148},
eprint = {2611.08148},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.08148},
}
RIS
TY - UNPB
TI - Detecting Weather Fronts from Satellite Observations: What Does Reanalysis Add?
AU - Ioannis Dravilas
AU - Miruna Stoicescu
AU - Roope Tervo
AU - Carlos Horn
AU - Oriol Hinojo Comellas
AU - Sina Montazeri
AU - Andrea Meraner
AU - Ana Lucic
PY - 2026
DA - 2026/11/05
PB - arXiv
DO - 10.48550/arxiv.2611.08148
UR - https://arxiv.org/abs/2611.08148
AB - Weather fronts are transition regions between air masses with contrasting thermodynamic properties and are associated with mid-latitude precipitation and severe weather. We investigate machine learning (ML) methods for detecting weather fronts and quantify the contribution of reanalysis data to performance. We compare two configurations of the same ML model for pixel-wise frontal classification over the North Atlantic and Europe: (1) using only geostationary infrared satellite observations, and (2) additionally using 28 ERA5 reanalysis fields. The satellite-only model achieves a neighbourhood critical success index at a 150 km tolerance (CSI@150) of $0.402$, which lies within the range of agreement between analyses from different meteorological services. The satellite+ERA5 model achieves $0.575$, an improvement of $0.173$. Our analysis shows that low-level winds are the least predictable from the satellite representation (R^2 = 0.23, compared with $0.88$ for temperature), while removing winds or moisture causes the largest decreases in detection performance.
ER -