Bamahry, Fikri ; Legrand, Juliette ; Pottiaux, Eric ; Bruyninx, Carine ; Fabian, Andras

Talk presented at EUREF Symposium 2023, Gothenburg, Sweden on 2023-05-26

Abstract: Data cleaning in GNSS position time series analysis is a critical step that can affect the accuracy and reliability of daily GNSS position and velocity estimates. GNSS time series cleaning methods often involve identifying positions that differ statistically from other positions without knowing the cause of the position outlier. However, data quality degradation is a crucial factor affecting the quality of GNSS position estimates. In this investigation, we implemented a supervised machine-learning algorithm to automatically identify possible position outliers caused by degraded data quality. Our approach investigated the correlation between GPS data quality indicators and outliers in daily position time series to construct a predictive model that can identify possible outliers in daily position time series. Our algorithm was trained using the position time series of EPN stations along with six GPS daily data quality indicators: the number of observed versus expected observations in dual frequency, the lowest elevation cut-off observed, the number of missing epochs, the number of satellites, the number of observations, and the number of cycle slips. Through this process, we identified the most important GPS data quality indicators explaining outliers in the GPS position time series. In this presentation, we will present the preliminary results of our work.

Keyword(s): GNSS quality metrics ; GNSS position time series ; Machine learning
Funding: ESFRIFED/ EF/211/SERVE / SERVE