Workshop 2018
Alvaro Santamaria-Gomez - On the spatial correlation of colored noise in GPS position time series - October 29, 2018 • 167 Views
Variability in GPS residual position time series (trend, offsets and seasonal removed) is dominated by colored noise from a few months of observation. The colored noise is generally well represented by flicker noise and its origin has not been unveiled yet. We investigate the spatial extent or radial footprint of the colored noise in residual position time series from the IGS combined solution. Using differenced residual position time series between IGS station pairs separated up to 8 000 km, we build the colored noise semi-variogram from a power-law noise process fitted to each component. We also build the semi-variogram corresponding to deformation series at the same set of stations obtained from the GFZ’s atmospheric, oceanic and hydrospheric loading models. Our results show the colored noise is spatially correlated from distances of ~ 750 km up to ~ 4 000 km, depending on the component. It is dominated by flicker noise in the horizontal components, but it is whiter in the vertical component. At very short distances (< 250 km), the colored noise suddenly increases producing a banana-like shape in the semi-variogram that may be explained by the impact of local tectonic activity. Colored noise in modeled loading deformation is spatially correlated up to ~ 1 500 km and its amplitude is also half that seen in the GPS series with a slightly whiter spectral index. When correcting the GPS residual series for the modelled loading deformation, the spatial correlation is reduced down to ~ 2 000 km for the vertical component. In terms of colored noise, this is the major benefit of correcting the GPS series with loading models. For the horizontal components, there is no significant change, except the colored noise becoming slightly whiter for distances shorter than ~ 3 000 km.

View PDF
 
 
 
 



Please include the IGS logo and citation in all websites, presentations, publications, or posters that use IGS data, products, or services. IGS is a service of IAG.