Determination of realistic uplift rate and noise assessment using GNSS coordinate time series

The Earth deformation generally is divided into horizontal and vertical components. The vertical component has different behaviour and complexity caused by the tidal and non-tidal loading and other effects. In this contribution, the vertical motion rate of the Earth crust is considered and studied based on GNSS coordinate time series. The focus of this presentation is the signal processing algorithm applied to the time series. One of the challenges in signal extraction and data processing in the sequential steps is the error budget. The strategy which we considered in the uplift determination process after data-cleaning and outlier detection is to firstly detect the significant signals in the time series and the change points. Then, after removing the aforementioned signals e.g. annual, semi-annual, diurnal, tidal etc., the rate of the up component of the GNSS time series and the related uncertainty is estimated. To achieve this, we employed Monte-Carlo Singular Spectrum Analysis i.e. Monte-Carlo SSA for the signal detection process, Bayesian Estimator of Abrupt change, Seasonal change, and Trend (BEAST) for the change point detection, variance estimation algorithm e.g. LS-VCE, Allan variance etc. for noise characteristic determination and Least-Squares estimation for the estimation of uplift rate and the related uncertainty. In an overall view, in this contribution, the importance of realistic errors is highlighted to estimate the uplift rate and the related uncertainty e.g. as geophysical boundary value for mantle convection models.