GraphTransformers for Geospatial Forecasting of Hurricane Trajectories
This work addresses hurricane forecasting for disaster management, representing an incremental advance by applying a novel method to a known bottleneck in geospatial sequence modeling.
The paper tackles hurricane trajectory prediction by introducing GraphTransformers to leverage emergent graph structures between geospatial points, resulting in significant improvements over state-of-the-art Transformer baselines on the HURDAT dataset.
In this paper we introduce a novel framework for trajectory prediction of geospatial sequences using GraphTransformers. When viewed across several sequences, we observed that a graph structure automatically emerges between different geospatial points that is often not taken into account for such sequence modeling tasks. We show that by leveraging this graph structure explicitly, geospatial trajectory prediction can be significantly improved. Our GraphTransformer approach improves upon state-of-the-art Transformer based baseline significantly on HURDAT, a dataset where we are interested in predicting the trajectory of a hurricane on a 6 hourly basis.