ESA StormSurgeCastNet Dataset

This is the dataset for the storm surge forecasting work of Ebel et al (2024), featuring the curation of a global and multi-decadal dataset of extreme weather induced storm surges as well as the implementation of neural networks for addressing the associated forecasting task. The provided dataset is multi-modal and features preprocessed in-situ tide gauge records as well as atmospheric reanalysis products and ocean state simulations. The models applied in this work learn to fuse the sparse yet accurate in-situ measurements with the global ocean and atmosphere state products. This way, more accurate storm surge forecasts are achieved, with predictions broadcasted to sites missing well-maintained tidal gauge infrastructure.

• The publication is available in the proceedings https://openaccess.thecvf.com/content/CVPR2024W/EarthVision/html/Ebel_Implicit_Assimilation_of_Sparse_In_Situ_Data_for_Dense__CVPRW_2024_paper.html
• For the associated code, please see https://github.com/PatrickESA/StormSurgeCastNet
• For any further questions, please reach out to me here or via the credentials on my website.

Reference:

P. Ebel, B. Victor, P. Naylor, G. Meoni, F. Serva, R. Schneider Implicit Assimilation of Sparse In Situ Data for Dense & Global Storm Surge Forecasting. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2024.