Yu Xuexiang, Yuan Jiajia, Yang Xu, Zhu Mingfei, Han Yuchen, Wei Min, Guo Zhongchen
School of Earth and Environment, Anhui University of Science & Technology, Huainan 232001, China.
Anhui Provincial Joint Laboratory of Urban 3D Real-Scene and Intelligent Security Monitoring, Huainan 232001, China.
iScience. 2025 Mar 31;28(5):112328. doi: 10.1016/j.isci.2025.112328. eCollection 2025 May 16.
Traditional Zenith Tropospheric Delay (ZTD) models often face difficulties in maintaining prediction accuracy under complex meteorological conditions and data loss. To address this, we propose the transformer-xLSTM (TransXLT) model, which integrates spatial-temporal information from global navigation satellite system (GNSS) stations, ERA5 (global atmospheric reanalysis), and GPT3 (empirical ZTD estimation). Missing data are reconstructed using a sparse attention-based time series reconstruction (SASR) method. Experimental results show: (1) under a 120-h data loss, SASR reduces mean absolute error (MAE) by 24.5% compared to cubic Hermite interpolation; (2) SASR lowers training root mean square error (RMSE) by 15.1% versus direct data deletion; and (3) TransXLT achieves an average RMSE of 8.13 mm across six sites, reducing RMSE by up to 76.54% compared to benchmarks like CNN-LSTM and ERA5. Demonstrating robustness across varying latitudes, altitudes, and seasons, the model significantly advances ZTD estimation accuracy for GNSS applications.
传统的天顶对流层延迟（ZTD）模型在复杂气象条件和数据丢失情况下，往往难以保持预测精度。为解决这一问题，我们提出了变压器长短期记忆网络（Transformer-xLSTM，简称TransXLT）模型，该模型整合了来自全球导航卫星系统（GNSS）站、ERA5（全球大气再分析）和GPT3（经验ZTD估计）的时空信息。使用基于稀疏注意力的时间序列重建（SASR）方法对缺失数据进行重建。实验结果表明：（1）在120小时数据丢失情况下，与三次埃尔米特插值相比，SASR将平均绝对误差（MAE）降低了24.5%；（2）与直接删除数据相比，SASR将训练均方根误差（RMSE）降低了15.1%；（3）TransXLT在六个站点实现了平均RMSE为8.13毫米，与CNN-LSTM和ERA5等基准相比，RMSE降低了高达76.54%。该模型在不同纬度、海拔和季节均表现出稳健性，显著提高了GNSS应用中ZTD估计的精度。
