6 Apr 2004

Giant drift rates look like incorrect earth tides; notice SBNM that
shows the opposite drift curve, unless you realize it starts at the
high.  All stations show a diurnal-looking signal of the appropriate
size for ETC.

Check GMT offset and times in the data file.

The low drift rate at the beginning makes it look like the second day
and beyond have the issues, and the earlier data are residual linear
drift of ~30 uGal/day.

7 Apr 2004

Looking at data for station 0 (WBB before moving to Nebraska), the
minimal ETC residual is for GMT-8!  Why?  Perhaps the clock was set
wrong in Utah?  ETC for drift data prior to trip also has minimal
residual at GMT-8, although residual is still 20 uGal peak-trough.

So, reset alldata.txt to GMT-8 and reprocess with reduce.py....

Drift not improved by going to GMT-8; drift curves now look more
recurring.

So, reset to -7 hrs, and analyse each day as it's own survey; the drift
function for all days ends up with big residuals.  Still need to work
on drift function for multi-day surveys.


Split reduction into subdirectories for each day: 07-23 and 07-24

Reprocess and check MICROTEL and SBNM values.  Differences could well
be changes in Earth tides, which are _NOT_ being computed correctly for
these stations/times/dates/something....Why?


8 Apr 2004

Checking the sign of the correction with the Aug 03 NE data, it is
determined that Tamura corrections are added, not subtracted.  Results
from yesterday were with Tamura subtracted, which added a huge ETC
signal to the data.

Rerun the days with the corrected (reverted) reduce.py code, and get
much better drifts.

Differences between MICTROTEL and SBNM are -18 and -20 uGal,
respectively.  So, apply +20 uGal to all stations on the 24th when
interpreting....


9 Apr 2004

Found problem with ETC - dates set to 2004, not 2003!

Use SBNM as reference, so still apply -20 uGal shifts.
