This paper employs time-frequency analysis on time series with a sampling interval of 10 s of the E, N, H components of a single permanent GNSS station in Krakow which uses three different Ground-Based Augmentation Systems (GBAS). Time-frequency analysis was conducted using the Fourier transform band-pass filter, which segments the time series into frequency components. Through the examination of these coordinates and their discrepancies, it was observed that the correlation between results from different GBAS solutions is negligible. Additionally, the analysis of power-law noise in the E, N and H components reveals that the spectrum index of these time series, computed through robust estimation and a linear regression model, ranges from α = 0.90 to 1.22. This suggests their resemblance to pink noise or flicker noise