Synthetic GNSS time series dataset for benchmarking velocity estimation and horizontal strain analysis algorithms

This dataset was generated for a controlled numerical experiment aimed at evaluating the stability of GNSS data processing algorithms used for station velocity estimation and subsequent horizontal strain analysis. It is intended to support a reproducible comparison of processing approaches under different types of time-series disturbances. The dataset includes 16 virtual GNSS stations. For each station, planar coordinates, predefined true horizontal velocities in the east and north directions, and formal uncertainties of the displacement components are provided. The synthetic observation dataset contains monthly east and north displacement time series covering 72 epochs over an approximately six-year period. Eight time-series scenarios are included: a reference scenario without additional disturbances; white noise; temporally correlated colored noise; seasonal variations; data gaps; coordinate jumps; outliers; and a combined scenario incorporating several types of disturbances. In total, the dataset contains 9,216 station-epoch observations. The predefined true station velocities provide reference values for direct assessment of velocity estimation errors. The dataset can also be used as input for subsequent calculation of horizontal strain parameters, including strain tensor components, dilation, shear strain, principal strains, and rotation rates within triangular finite elements. This makes it suitable for testing how errors introduced during GNSS time-series processing propagate into deformation estimates and how the resulting uncertainties depend on the selected processing chain and the spatial geometry of the GNSS network. The dataset is intended for methodological studies, algorithm validation, educational applications, and reproducible benchmarking of GNSS-based deformation analysis workflows.