ULR Re-analysed Global GPS Solution for Vertical Land Motion Correction at Tide Gauges

The presentation will review the recent results published by Wöppelmann et al. (2007) in Global and Planetary Change. Geocentric sea-level trend estimates were derived from the global GPS analyses conducted at ULR consortium to correct a set of relevant tide gauges from the vertical motion of the land upon which they are settled. The exercise proved worthwhile. The results showed a reduced dispersion of the estimated sea level trends, either regionally or globally, after application of the GPS corrections compared to the corrections derived from the glacio-isostatic adjustment models of Peltier (2004). Here we will focus on two important issues that were not addressed in Wöppelmann et al. (2007). The first issue concerns the noise content of our GPS solutions. Previous works have shown that GPS coordinate time series are subject to significant time-correlated (coloured) noise, with a large predominance of flicker noise (Zhang et al. 1997, Mao et al. 1999, Williams et al. 2004). The presence of coloured noise in a time series has a significant effect on the rate uncertainty, which may otherwise be underestimated by as much as an order of magnitude. We therefore carefully investigate the now 10-year long data set of reanalysed GPS solutions for noise content using the Allan variance technique (Feissel et al. 2007). Preliminary results show that the reanalysed solutions at ULR exhibit far less flicker noise than any other solution published so far in the literature available to us. The percentage of stations with flicker noise drops to only about 20%. These encouraging results advocate for a comprehensive reanalysis strategy with full coherent models over the entire observation data span. Moreover, the noise level reaches the best levels of other geodetic results recently published, namely the VLBI level in the horizontal component and the SLR level in the vertical component (Feissel et al. 2007). The second issue that we would like to address in the presentation relates to the reference frame realisation. This is indeed a long standing key issue in achieving the accuracy goal required by long term sea level studies from tide gauges. We investigate the impact of the recent transition from ITRF2000 to ITRF2005 in our previous and current sea level trend estimates.