Novel Intelligent Spatiotemporal Grid Earthquake Early-Warning Model

The integration analysis of multi-type geospatial information poses challenges to existing spatiotemporal data organization models and based on deep learning. For earthquake early warning, this study proposes a novel intelligent grid model GeoSOT (SGMG-EEW) for feature fusion data. This includes seismic sample (SGSM) three-dimensional group convolution neural network (3DGCNN-SGM). SGSM solves the problem concerning that layers different types cannot form an ensemble with consistent structure transforms representation into samples 3DGCNN-SGM is first application in learning multi-source geographic It avoids direct superposition calculation between layers, which may negatively affect results. In study, taking atmospheric temperature anomaly historical precursory from Japan as example, warning verification experiment was conducted proposed SGMG-EEW. Five groups control experiments were designed, namely use only, non-group group, support vector machine statistical group. results showed not only compatible expression single type but can also multiple data, forming deep-learning-oriented structure. Compared traditional model, more suitable