BROWSE
Detail
Journal Article
Spatiotemporal neural network with attention mechanism for El Niño forecasts
- Authors
- Jinah Kim, Minho Kwon, Sung-Dae Kim, Jong-Seong Kug, Joon-Gyu Ryu, Jaeil Kim
- Issue Date
- 2022-05
- Citation
- Scientific Reports, v.12, pp.1-15
- ISSN
- 2045-2322
- Publisher
- Nature Research
- Language
- English
- Type
- Journal Article
- Abstract
- To learn spatiotemporal representations and anomaly predictions from geophysical data, we propose STANet, a spatiotemporal neural network with a trainable attention mechanism, and apply it to El Ni챰o predictions for long-lead forecasts. The STANet makes two critical architectural improvements: it learns spatial features globally by expanding the network's receptive field and encodes long-term sequential features with visual attention using a stateful long-short term memory network. The STANet conducts multitask learning of Nino3.4 index prediction and calendar month classification for predicted indices. In a comparison of the proposed STANet performance with the state-of-the-art model, the accuracy of the 12-month forecast lead correlation coefficient was improved by 5.8% and 13% for Nino3.4 index prediction and corresponding temporal classification, respectively. Furthermore, the spatially attentive regions for the strong El혻Ni챰o events displayed spatial relationships consistent with the revealed precursor for El혻Ni챰o occurrence, indicating that the proposed STANet provides good understanding of the spatiotemporal behavior of global sea surface temperature and oceanic heat content for El혻Ni챰o evolution.
- KSP Keywords
- Attention mechanism, Correlation Coefficient, Geophysical Data, Heat content, Long-short term memory(LSTM), Memory network, Receptive field, Sea Surface Temperature, Sequential features, Spatial relationships, Temporal classification
This work is distributed under the term of Creative Commons License (CCL)
(CC BY)
(CC BY)
