Processing SG data

1. Here is data from J9 Strasbourg, from the GGP database, at 1 minute sampling. It is the same data referred to in the paper by Hinderer et al. (2002) which can be found here. The data format is called plotp.

gravity: g1997.m1, and pressure: p1997.m1

2. The fixed pressure is p1997.m3, and the plot of the two pressure files is p1997a.pdf

3. After subtracting local tides and nominal atmospheric pressure, the gravity series becomes g1997.plm. A plot of the former two gravity files is g1997a.pdf.

3. After a first fix (fix2) of the major problems, we have the series g199702.plm. After a second pass, where I take care of many smaller problems (fix3), the series becomes g199703.plm.

4. A plot of the last two fixed gravity series is g1997b.pdf.

5. A comparison of the spectra of the fix2 and fix3 solutions is shown in st970203s.pdf.

Comments

1. The pressure must be fixed first if it is going to be used to remove pressure before fixing files.

2. We should always remove tides and pressure before fixing problems, then gaps can be interpolated linearly.

3. Notice the dramatic improvement of the noise level when going from fix2 to fix3 at periods of the the terdiurnal and quardiurnal tides and shorter. The long periods are almost unaffected by these improvements, and there is little evidence of the gain in signal to noise just in the overall look of the time series (see g1997b.pdf).

David Crossley

August 21, 2008