Graph-Based Deep Learning for Sea Surface Temperature Forecasts
This work addresses the problem of resource-intensive SST forecasts for marine ecosystem and aquaculture management, but it is incremental as it applies emerging graph-based methods to a known domain.
The paper tackled sea surface temperature forecasting by exploring graph neural networks as an alternative to grid-based deep learning, finding that GNNs achieved better one-month-ahead predictions than a persistence model in most oceans based on root mean square errors.
Sea surface temperature (SST) forecasts help with managing the marine ecosystem and the aquaculture impacted by anthropogenic climate change. Numerical dynamical models are resource intensive for SST forecasts; machine learning (ML) models could reduce high computational requirements and have been in the focus of the research community recently. ML models normally require a large amount of data for training. Environmental data are collected on regularly-spaced grids, so early work mainly used grid-based deep learning (DL) for prediction. However, both grid data and the corresponding DL approaches have inherent problems. As geometric DL has emerged, graphs as a more generalized data structure and graph neural networks (GNNs) have been introduced to the spatiotemporal domains. In this work, we preliminarily explored graph re-sampling and GNNs for global SST forecasts, and GNNs show better one month ahead SST prediction than the persistence model in most oceans in terms of root mean square errors.