[2604.04306] HighFM: Towards a Foundation Model for Learning Representations from High-Frequency Earth Observation Data

Computer Science > Computer Vision and Pattern Recognition arXiv:2604.04306 (cs) [Submitted on 5 Apr 2026] Title:HighFM: Towards a Foundation Model for Learning Representations from High-Frequency Earth Observation Data Authors:Stella Girtsou, Konstantinos Alexis, Giorgos Giannopoulos, Harris Kontoes View a PDF of the paper titled HighFM: Towards a Foundation Model for Learning Representations from High-Frequency Earth Observation Data, by Stella Girtsou and 3 other authors View PDF HTML (experimental) Abstract:The increasing frequency and severity of climate related disasters have intensified the need for real time monitoring, early warning, and informed decision-making. Earth Observation (EO), powered by satellite data and Machine Learning (ML), offers powerful tools to meet these challenges. Foundation Models (FMs) have revolutionized EO ML by enabling general-purpose pretraining on large scale remote sensing datasets. However most existing models rely on high-resolution satellite imagery with low revisit rates limiting their suitability for fast-evolving phenomena and time critical emergency response. In this work, we present HighFM, a first cut approach towards a FM for high temporal resolution, multispectral EO data. Leveraging over 2 TB of SEVIRI imagery from the Meteosat Second Generation (MSG) platform, we adapt the SatMAE masked autoencoding framework to learn robust spatiotemporal representations. To support real time monitoring, we enhance the original architec...