[unav_all] DOGEx: Detections of Offsets in GPS time series
Experiment [update]
Matt King
m.a.king at newcastle.ac.uk
Tue Jul 26 10:23:51 MDT 2011
Author: Matt King and Simon Williams
This email is an update on the previously announced community experiment DOGEx (please note the web link to the dataset has been updated - see below). A number of groups are now working on the dataset and we have shown updated results at IUGG in Melbourne. The IUGG poster may be seen at ftp://stella.ncl.ac.uk/pub/matt/DOGEx_IUGG.pdf
If you would like to participate in the experiment, then please see the email below.
** Please also note, we are now welcoming in a separate category *non-automated* solutions of the same dataset - that is, ones performed by an expert user.
***********************************************
DOGEx is designed to test automated offset detection algorithms and expert users against a known truth in the presence of realistic GPS timeseries signal, noise, offset frequency and gaps. DOGEx is a one-way blind experiment in that the actual offset times and magnitudes are not released with the site timeseries. The data set we supply is a simulated time series of daily NEU coordinates and the formal errors for 50 "sites".
The experiment is being run jointly by Matt King (Newcastle University, Newcastle upon Tyne, UK; m.a.king at ncl.ac.uk) and Simon Williams (National Oceanography Centre, Liverpool, UK; sdwil at pol.ac.uk).
The data set and further information may be found at http://www.cost-es0701.geoenvi.org/working-groups/working-group-3
We look forward to hearing from you with solutions! Please circulate this email to other interested colleagues.
Matt King
Reader in Polar Geodesy & RCUK Academic Fellow
School of Civil Engineering and Geosciences Email: m.a.king at ncl.ac.uk
Cassie Building Ph: 0191 222 7833 Int: + 44 191 222 7833
Newcastle University Fax: 0191 222 6502 Int: + 44 191 222 6502
Newcastle upon Tyne Web: http://www.staff.ncl.ac.uk/m.a.king/
United Kingdom, NE1 7RU http://www.ceg.ncl.ac.uk/profiles2/m.a.king
More information about the unav_all
mailing list