[mojoB]

Dam monitoring systems use DGPS, one receiver on the wall, the other nearby.

Without a local reference station, atmospheric and ionospheric disturbances would totally overwhelm the effect they are trying to measure.

This [1] covers it quite well.

[1] http://pasadena.wr.usgs.gov/office/hudnut/SRL/

[eps]

Sure thing, that's the simple way :)

It was in mid 90s, but I can guarantee you that I sat next to a math PhD guy who worked exactly on what I described, including the dam deployment. Single receiver, 3mm precison from 3-4 hours of data, no DGPS. This wasn't in the US though.

[xemoka]

And before selective availability[0] was turned off? So did he have military GPS equipment and not consumer grade?

[0] http://www.gps.gov/systems/gps/modernization/sa/

[jevinskie]

Very interesting data. It was cool to notice some graphs had much higher variability and then notice that they were all sites that are far from the equator. Then the article helpfully explained precisely why they had more variance!