Regional Zenith Tropospheric Delay Modeling Based on Least Squares Support Vector Machine Using GNSS and ERA5 Data

The meteorological reanalysis data has been widely applied to derive zenith tropospheric delay (ZTD) with a high spatial and temporal resolution. With the rapid development of artificial intelligence, machine learning also begins as high-efficiency tool be employed in modeling predicting ZTD. In this paper, we develop three new regional ZTD models based on least squares support vector (LSSVM), using both International GNSS Service (IGS)-ZTD products European Centre for Medium-Range Weather Forecasts Reanalysis 5 (ERA5) over Europe throughout 2018. Among them, ERA5 is extended ERA5S-ZTD ERA5P-ZTD background by model method integral method, respectively. Depending different data, schemes are designed construct LSSVM algorithm, including without ERA5S-ZTD, ERA5P-ZTD. To investigate advantage feasibility proposed models, evaluate accuracy two segmental comparison IGS-ZTD 85 IGS stations Europe. results show that overall average Root Mean Square Errors (RMSE) value all sites 30.1 mm 10.7 RMSE 25.8 mm, 22.9 9 schemes, Moreover, improvement rate 19.1% 1.6% ERA5P-ZTD, order explore reason lower loop verification performed estimating values each available station. actuality, monthly estimated positive most stations, biggest can even reach about 40%. negative mainly comes from specific these located edge region, near coast, well similarity between individual verified station training stations.