2026-10-14T04:56:36Z https://www.scidb.cn/oai

10.57760/sciencedb.39556 2026-06-11T14:37:37Z

2026-06-11 PINN-STEC doi:10.57760/sciencedb.39556 en A High-Precision Ionosphere Slant Total Electron Content Forecast Model Based on Physical Information Neural Network • BDS-3 single-frequency observational data can generate high-quality benchmark STEC via GIM spatiotemporal interpolation and thin-shell geometric mapping. • Physics-Informed Neural Network constrains network training with ionospheric non-negative rules and thin-shell projection principles embedded within composite loss functions. • Latitude-dependent geomagnetic disturbance and geometric path effects can be fully learned by the feature-attention PINN for robust multi-step STEC prediction under geomagnetic storms.; • Physics-Informed Neural Network constrains network training with ionospheric non-negative rules and thin-shell projection principles embedded within composite loss functions.; • Latitude-dependent geomagnetic disturbance and geometric path effects can be fully learned by the feature-attention PINN for robust multi-step STEC prediction under geomagnetic storms. zhanghuan PUBLIC https://creativecommons.org/publicdomain/zero/1.0/ dataset Science Data Bank