An adaptive spatiotemporal clustering framework boosts deep learning reconstruction of global ocean subsurface temperature fields from surface data, delivering 12.4% to 27.2% RMSE improvements when paired with models such as DP-CNN, Attention U-Net, and ViT.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
physics.ao-ph 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
An Adaptive Spatiotemporal Clustering Framework for 3D Ocean Subsurface Temperature Reconstruction
An adaptive spatiotemporal clustering framework boosts deep learning reconstruction of global ocean subsurface temperature fields from surface data, delivering 12.4% to 27.2% RMSE improvements when paired with models such as DP-CNN, Attention U-Net, and ViT.