HR: 1330h
AN: G52B-0041    [PDF]
TI: GPS observations of Ocean Tide Loading in the British Isles
AU: * Allinson, C R
EM: c.r.allinson@ncl.ac.uk
AF: University Of Newcastle upon Tyne, School of Civil Engineering and Geosciences Bedson Building, Newcastle upon Tyne, NE1 7RU United Kingdom
AU: Clarke, P J
EM: peter.clarke@ncl.ac.uk
AF: University Of Newcastle upon Tyne, School of Civil Engineering and Geosciences Bedson Building, Newcastle upon Tyne, NE1 7RU United Kingdom
AU: King, M A
EM: m.a.king@ncl.ac.uk
AF: University Of Newcastle upon Tyne, School of Civil Engineering and Geosciences Bedson Building, Newcastle upon Tyne, NE1 7RU United Kingdom
AU: Edwards, S J
EM: S.J.Edwards@ncl.ac.uk
AF: University Of Newcastle upon Tyne, School of Civil Engineering and Geosciences Bedson Building, Newcastle upon Tyne, NE1 7RU United Kingdom
AU: Cruddace, P
EM: Paul.Cruddace@ordnancesurvey.co.uk
AF: Ordnance Survey, Romsey Road, Southampton, SO16 4GU United Kingdom
AU: Baker, T F
EM: tfb@pol.ac.uk
AF: Proudman Oceanographic Laboratory, Bidston Observatory, Birkenhead, CH43 7RA United Kingdom
AB: Model predictions show that ocean tide loading (OTL) in the British Isles is particularly large (up to several cms in the S. W. of Britain). The primary cause of uncertainty in OTL models is the ocean tide models on which they depend. Interaction of the tides in the open oceans and the more restricted coastal regions, along with the intricate shape of the coastline, make accurate modelling of OTL in the British Isles difficult. Therefore an approach to directly measure these effects has been devised, using data gathered from a network of continuously operating GPS receivers (COGRs), in an attempt to validate or improve existing OTL models. We observe OTL at diurnal and semi-diurnal periods by directly estimating fixed-period harmonic motions within individual daily GIPSY/OASIS II GPS analyses. This approach enables us to solve separately for the tropospheric zenith wet delay (ZWD) as a random-walk parameter. An iterative Kalman Filter approach to combine the multiple daily solutions enables us to isolate the principal near-diurnal (K$_{1}$, O$_{1}$, P$_{1}$, Q$_{1}$) and near-semi-diurnal (M$_{2}$, S$_{2}$, N$_{2}$, K$_{2}$) OTL components. A preliminary test shows that data from six UK sites produce estimates which are in good agreement with OTL predicted by the FES99 model, where values vary from 3-43 mm in amplitude (M$_{2}$) at the six sites. Our final estimates are generated using approximately 1000 days of data in the solution, resulting in amplitude standard deviations of approximately 1 mm per component (compared with an RMS difference between OTL models of 2 mm per component). However, the phase estimates of the OTL components take much longer to converge. Phase standard deviations for M$_{2}$ remain approximately around 10 degrees after 1000 days and 20-40 degrees for the remaining components (compared to the phase RMS between OTL models of 1-17 degrees for the larger amplitude signals). We also show that our estimates are stable for the majority of OTL components when data from at least 90 days are stacked, with an amplitude standard deviation of approximately 2 mm. Exceptionally, the K$_{1}$ component requires at least 200 days of data before similar confidence levels are achieved.
UR: http://www.staff.ncl.ac.uk/m.a.king/UK_otl.htm
DE: 1243 Space geodetic surveys
DE: 1249 Tides--Earth
DE: 1255 Tides--ocean (4560)
SC: Geodesy [G]
MN: 2003 Fall Meeting