Three creepmeters of new design were operated on the Superstition Hills fault following the 24 November 1987 earthquake. Each creepmeter consists of a digital displacement sensor interrogated periodically by a computer. At the end of each hour, the computer determines a recording rate appropriate to the velocity of the slip on the fault in that hour. No data are recorded if the fault is inactive. Data are recorded at a maximum rate of one sample per minute if slip on the fault is detected. The resolution and range of the creepmeters are 10 μm and 15 cm, respectively.
Fault activity is observed to be confined to periods of episodic slip (creep events) with mean amplitudes 3 to 6 mm and durations less than 4 hr. A total of 80 creep events were recorded. During the first week following the earthquake, creep events occurred several times each day. After 6 months, events occurred every 3 weeks. This increase in the mean time interval between creep events results in the observed decay in afterslip rate. Between creep events, the fault is inactive although more than 200 microstrain of fault-zone deformation may accumulate before each event.
Although creep events were observed to propagate unilaterally in both directions along the fault, the most impulsive events appeared to propagate bilaterally from a dilatant offset separating the northern and southern halves of the 24-km-long fault zone. Maximum propagation velocities of 30 m/sec were observed. Aftershocks (2 < M < 4.7) were observed to occur near the creepmeters following 6 to 9 cm of surface slip on the fault. A future aftershock in this region is anticipated in mid-1989 based on afterslip rates in the year following the main shock.