Satellite-based remote sensing is a valuable source of spatial data for monitoring vast territories, yet the availability of high-resolution (HR) imagery remains a major bottleneck. Image super-resolution (SR) offers an alternative to upgrading onboard sensors, but reconstruction quality is often degraded by atmospheric artifacts such as clouds, which are typically handled by pre- or post-processing rather than being integrated into the SR pipeline. In this study, we investigate whether geospatial priors derived from AlphaEarth Foundation (AEF) model embeddings can simultaneously enhance SR reconstruction and facilitate cloud-affected region recovery. We explore both single-image (SISR) and multi-image (MISR) configurations of ESRGAN for RGB and multispectral (MSI) imagery, conditioned on AEF embeddings. For cloud removal, we evaluate Progressive Multi-scale Attention Autoencoder (PMAA) and Conditional Tabular GAN (CTGAN) models, also assessing the effect of AEF conditioning. Using a modified Cloud-S2-NAIP dataset, we demonstrate that AEF conditioning consistently improves SSIM, LPIPS, CLIP-based similarity, and SAM across both SR and cloud-restoration tasks. The AEF-conditioned MISR ESRGAN achieves the best overall LPIPS (0.264) and CLIP similarity (0.952), outperforming both standalone SR configurations and PMAA-/CTGAN-based inpainting. Crucially, we show that AEF conditioning provides an implicit cloud-inpainting effect within the MISR framework, enabling simultaneous spatial enhancement and cloud-damage recovery. This approach offers a promising pipeline for cloud-aware SR in real-world applications such as environmental monitoring and infrastructure analysis.
E. Burnaev, S. Illarionova, Usman Tasuev et al.· IEEE Access· 0 citations
Effective flood monitoring is critical for minimizing the impacts of flood disasters on populations and infrastructure. Yet reliable remote sensing across extensive and environmentally diverse regions remains challenging, as most segmentation algorithms lack the generalisation capacity required for large-scale application, while annotated flood data are scarce and unevenly distributed. This study presents an end-to-end multimodal framework for Russian Federation territories sustainable flood monitoring and damage assessment based on synthetic aperture radar data, multispectral imagery, and digital elevation models with their derivatives, forming a 21-channel input. Using a self-collected multimodal dataset covering seven Russian regions, two strategies for water surface detection under limited data conditions were compared: a supervised U-Net++ model and the self-supervised AnySat architecture pre-trained and fine-tuned for the segmentation task. Under the data conditions of this study, supervised learning proved more effective, while the AnySat-based approach offered greater stability and retains advantages for settings where larger unlabelled data or missing modalities at inference are expected. The best flood area predictions were used to estimate flood impact in urban areas in terms of the area affected, material damage, casualties, and ecological and agricultural impact. The estimations were conducted following the official methodology of the Russian Ministry of Emergency Situations. Applied to the 2019 Tulun flood, the obtained results closely matched official assessments, except for material damage, due to the open-source databases usage. The results demonstrate the potential of deep learning and multimodal satellite data integration for scalable, reliable flood monitoring across diverse environmental and data-limited conditions.
I. Novikov, S. Illarionova, Ruslan B. Dzharkinov et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
