Jim,

The first task when considering strainmeters or tiltmeters is just
what range of strain or displacement is expected over what period of
time. If you have an active fault crossing a paved road, a linear
line of nails makes a great creepmeter for movements of centimeters
per year. For continuous measurements a taught wire (often of Invar
or carbon fiber) is stretched across the fault at an oblique angle
inside a large (eg 6") ABS pipe between buried concrete vaults (pre-
cast septic tanks are used). Often several fiber creepmeters are installed
in parallel using the same pipe, vaults, and monuments. Extra fibers
are set aside as spares, and using different wires or fibers allows 
temperature compensation.
One end of the wire or fiber is anchored to a large a monument in as deep
a hole as you can afford, and the free end is coupled to a displacement
transducer anchored to a similar monument. In hard rock sites, earth-
tide resolution has been achieved at 10^-7 strain. Otherwise surface
hydrology limits useful long term measurements, unless large movements
are expected, such as 2 to 3 cm/year at Parkfield and Hollister in CA.
Much work has been done trying to get reproducible data, especially
in surface installations, over time periods of tectonic interest, like
YEARS. After a number of wishful claims of success, surface strain
or creep measurements have faded away except in locations with very
active tectonics. There are no cheap or easy ways to do this.

Among the best strainmeters are the 750 meter baseline laser interferometer
systems at Pinon Flat that operate in evacuated tubes and have massive
end monuments with "optical anchors" thru the decomposed granite of the
surface to virgin rock 20 meters down. They have achieved stability of
better than 10^-6 / year but at great expense and effort.

GPS measurements are also used, but over longer distances, and expected
tectonic deformations and fault creep have been seen. But even here
the wide area data does not agree with the on-fault creep data because
the fault is at depth and its movement is better integrated over a large
area of the surface rather than being localized at a surface fault.

A classical reference on the subject can be found in:

Agnew, D.C., "Strainmeters and Tiltmeters",
Reviews of Geophysics,
vol 24, No 3, 579-634, 1986
_______________

Regards,
Sean-Thomas
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)