Temporal and spatial stochastic behaviour of high-frequency slant tropospheric delays from simulations and real GPS data

In this paper, a turbulence theory-based simulation procedure for slant tropospheric delay variations is presented. Based on this procedure tropospheric delay variations are simulated for three different geometric scenarios. The stochastic behaviour of the generated time series is assessed in terms of temporal structure functions. It is shown that the temporal structure functions – in general – follow a 5/3 to 2/3 power-law behaviour. Deviations from this behaviour due to the complex interaction between varying observation geometry and atmospheric/turbulent conditions are discussed.In addition to simulated slant delay variations, zenith tropospheric delays of a specially designed Global Positioning System (GPS) network are estimated by ‘Precise Point Positioning’. Both their temporal and spatial structure functions are computed and compared with their theoretically predicted behaviour. For the temporal behaviour of real tropospheric delay time series, we found an initial 5/3 power-law behaviour and correlation lengths of approximately 2000 s. The spatial behaviour of ZTDs of the 16 km straight line network showed a 2/3 power-law behaviour, indicating two-dimensional turbulence processes.