This statement tells what we used to do before funding was terminated in about 2000.
PURPOSE AND SCOPE
The long-term, fixed purpose of this investigation was to search for and monitor the spatial and temporal nature of nearfield displacement across active and potentially active faults. Thus, we documented pre-, co- and post-seismic displacement and aseismic creep, if any, especially where seismographic, paleoseismic, and geomorphic evidence indicated current or recent fault activity. The geodetic arrays range in length from 300 m to 7000 m and are intermediate in scale, therefore, between the regional geodetic surveys and point measurements by continually recording instruments such as creepmeters, tiltmeters, and strainmeters.
The surveying array measurements yield data on the amount of surficial preseismic, coseismic and post-seismic horizontal displacement at a scale intermediate between that obtained by GPS and by existing USGS geodimeter networks and what might be gained from creepmeters or by study of offsett road stripes, tire tracks, stream gulches, and similar imprecise and ephemeral markers. Leveled alignment arrays give detailed information on the distribution of horizontal and vertical displacement, which, in turn, provides information on fault slip. This kind of information is needed to solve more basic problems such as the relation between individual faults and regional tectonic processes in southern California, as well as the physical processes that occur in the region of failure on a fault at the time of an earthquake (Wesson and Wallace, 1985). In the long run, the most reliable basis for earthquake prediction and hazard reduction will come from a thorough understanding of all these processes (Page et al, 1992).
METHODS AND PROCEDURES
We surveyed and maintained 63 short leveling arrays in California ranging in length from 250 m to 7000 m and ranging in geometry from straight lines to L-, Z-, W-, and box-shapes. Choices of array locations and geometry were dictated by how well-located is the fault, by limitations imposed by local topography, access and property ownership, and by the presence, or anticipated presence of instrumental installations.
All of our leveling was done according to First Order, Class II standards (Federal Geodetic Commission, 1984) with an accepted uncertainty of 1.0 mm (L1/2) where L equals the one-way line length in kilometers, but we commonly achieved an observed sensitivity of 0.5 mm (L1/2) (Sylvester et al., 1993; Sylvester, 1995a). Bench marks are Class B rod marks (Floyd, 1978) or stainless steel plugs cemented in large boulders embedded deeply in alluvium or moraine (Sylvester, 1984), spaced no more than 50 m apart. Most are 40 m apart. Where the arrays cross faults, bench marks may be spaced as closely as 3 m to determine details of vertical displacement within the fault zone itself. The data were adjusted for closure, but we made no corrections for temperature, gravity, or magnetic effects, because over line lengths and shot lengths as short as these, the corrections are small enough to be neglected (Stein and others, 1986).
All leveling data collected during the period of study were obtained with a shaded Wild NAK2 precision automatic level with optical micrometer. We used strut-supported, double-sided GPL-3 precise invar leveling rods, serial numbers 6477A/B. The leveling rods were calibrated in March 1995 by the U.S. Navy Gage and Standards Laboratory, Pomona. During the last 5 years, some of the leveling was been done with a Leica NA3000 digital level with 3 m-long bar-coded rods, serial numbers 9505 and 9511.
We used a Wild TC2000 Total Station Distance Meter for our trilateration surveys and acquired, thereby, horizontal lengths determined by the instrument from slope lengths. Bench marks are Class B rod marks (Floyd, 1978) or bronze tablets set in concrete by the USGS or NGS.
Typically the arrays have apertures of a few hundred meters and were placed across a fault according to the topography, access, and trespass permission. By means of repetitive surveys with the Wild TC2000 over several years, we achieved a precision of 1 ppm in horizontal measurements of the quadrilaterals, but only about 5-10 ppm in vertical measurements. That precision is sufficient to measure coseismic slip and rapid horizontal creep and strain, but does not supplant precise leveling to determine vertical displacement.
We had two each Ashtech Z12 GPS receivers and choke ring antennas. We had an RTK package so that we could operate one receiver as a base station and the other as a rover out to distances as great as 8-10 km. We used this system to determine the XY positions of bench marks in our fault-crossing leveling and trilateration arrays. This work turned each of the leveling arrays into a leveled alignment array, capable of determining XYZ ground displacement to about 3 mm. Post processing of the field data was done by JPL. Results of the GPS surveys accompanied the level line data when available. Observation data and Summary Data Sheets were archived at UCSD.
Several straight line arrays of from 10 to 20 nails were established across the San Andreas fault in 1967 to document creep along the fault in central California (Rogers and Nason, 1971). We relocated those at Cholame and San Juan Bautista and resurveyed them intermittently until they were obliterated. The line at San Juan Bautista was paved over in 1995, but we reestablished a new line in June, 1995, using a Wild T2 theodolite to establish a straight line through the nails and precision calipers to measure the deviation of the nails from the straight line. We can measure the deviation to ± 1 mm. We inspected the line in December 1995 and observed no cracks in the asphalt at that time. We did observe cracks in the asphalt in summer 1996 and in December 1996. A nice set of en echelon cracks was extant in September 1997. The driveway and nails were paved over again after November 2008, and the nails have not been reset, so monitoring no longer occurs there or on any of the nail lines.
Return to the Title Page.