The treatment of time series data of GPS permanent stations for extracting geodetic signals

It is well known that the coordinates and the zenith total delays, estimated by the GPS, provide a big amount of information. Currently a lot of stations with a very long data time series are present all over the world which make now feasible to perform a critical analysis of the coordinates helpful for the detection of signals, such as: height displacements, velocities and strain rate suitable for geophysical interpretations. In the present work a method to assess the reliability of the time series is described and applied to some stations with a period of observations long 5 years at least. The method essentially build a "fingerprint" which give information how long time a signal can emerge from the time series. Furthermore with the method, harmonic signatures and jumps on the time series are detected and analysed. The method does not only detects the value of the jumps but also where they are placed along the time series. The good efficiency in the detection and evaluation of jumps is of utmost importance because they introduce spurious signatures which could dramatically delay the enhancement of the signals. So our method allows to detect jumps that could be neglected or not assessed or be themselves a signal to detect.