Error analysis of continuous GPS position time series

摘要: [1] A total of 954 continuous GPS position time series from 414 individual sites in nine different solutions were analyzed for noise content using maximum likelihood estimation (MLE). The lengths the varied around 16 months to over 10 years. MLE was used analyze data two ways. In first analysis assumed be white only, a combination plus flicker noise, or random walk noise. For second spectral index and amplitude power law estimated simultaneously with where globally distributed, can best described by Both components show latitude dependence their amplitudes (higher at equatorial sites) together bias larger values Southern Hemisphere. regional solutions, spatially correlated (common mode) signal has been removed, is significantly lower. more than global probably reflects mixture local effects. significant reduction seen since networks began recording early 1990s. comparison monument types California Integrated Network suggests that deep drill braced preferred stability.