Physics-informed graph attention LSTM with a novel spatial season-aware GPD claims to outperform baselines and SEAS5 for extreme rainfall prediction across Thailand gauge stations.
The MJO can be characterized as an eastward-moving pulse of cloud and rainfall near the equator that typically recurs every 30 to 60 days
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Leveraging Teleconnections with Physics-Informed Graph Attention Networks for Long-Range Extreme Rainfall Forecasting in Thailand
Physics-informed graph attention LSTM with a novel spatial season-aware GPD claims to outperform baselines and SEAS5 for extreme rainfall prediction across Thailand gauge stations.