No, these are final products. If you transformed them realy correctly into NEU components, there shouln't be any problem with publication.

Velocities don't need postprocessing unless you want to change the frame in which they are provided or to compute their formal uncertainty with a different stochastic model. Note that EUREF provides velocities in two frames (ETRF2000 and IGb08) and also in two realizations of the same frame (IGb08) with small differences.

The problem that Alvaro mentioned is much more complicated. Since we now know that ITRF2008 origin rate information is more accurate than ITRF2000 (and ETRF2000(R08) in consequence, because for transformation it uses ITRF2000 by 14-parameter tranfsformation) for proper vertical interpretation of the motion you should always use ITRF2008. Or IGb08 if available. The difference is about -1.8 * sin(phi) + 0.5, details you can read in a report of working group by X. Collilieux et al.; 2013, IUGG2011 proceedings. ETRF should be used only for purposes related to reference frame maintenance in Europe.

Concerning the basic question I always prefare to use time series from SINEX files, make the pre-processing by myself (outliers, offsets and seasonals influence velocity a lot) and then determine velocity using Least Squares Estimation. It is time consuming, but I am sure about reference frame and I see the time series to verify the velocity.

You can also transform Vx,Vy,Vz into Vn,Ve,Vu  by multiplying the vector for the velocities of X,Y, and Z with the transformation matrix G.

Hi D.Ugur Sanil,

Please send me the transformation matrix G or please refer the relevant book. I want to convert GPS velocities from earth to navigation frame.