Analysis of near-source static and dynamic measurements from the 1979 Imperial Valley earthquake

Abstract

abstractGross features of the rupture mechanism of the 1979 Imperial Valley earthquake (ML = 6.6) are inferred from qualitative analysis of near-source ground motion data and observed surface rupture. A lower bound on the event's seismic moment of 2.5 × 1025 dyne-cm is obtained by assuming that the average slip over the whole fault plane equals the average surface rupture, 40.5 cm. Far-field estimates of moment suggest an average slip over the fault plane of 105 cm, from which a static stress drop of 11 bars is obtained. An alternative slip model, consistent with the far-field moment, has 40.5 cm of slip in the upper 5 km of the fault and 120 cm of slip in the lower 5 km. This model suggests a static stress drop of 39 bars. From the lower estimate of 11 bars, an average strain drop of 32 µstrain is derived. This strain drop is four times greater than the strain that could have accumulated since the 1940 El Centro earthquake based on measured strain rates for the region. Hence, a major portion of the strain released in the 1979 main shock had been accumulated prior to 1940.Unusually large amplitude (500 to 600 cm/sec2) vertical accelerations were recorded at stations E05, E06, E07, E08, EDA of the EI Centro array, and the five stations of the differential array near EDA. Although the peak acceleration of 1705 cm/sec2 at E06 is probably amplified by a factor of 3 due to local site conditions, these large amplitude vertical accelerations are unusual in that they are evident on only a few stations, all of which are near the fault trace and at about the same epicentral range. Two possible explanations are considered: first, that they are due to a direct P wave generated from a region about 17 km north of the hypocenter, or second, that they are due to a PP phase that is unusually strong in the Imperial Valley due to the large P-wave velocity gradient in the upper 5 km of the Imperial Valley.Based on the distribution of both the horizontal and vertical offsets, it is likely that the rupture went beyond stations E06 and E07 during the main shock. By exploiting the antisymmetry of the parallel components of particle velocity between E06 and E07 and by examining polarization diagrams of the particle velocity at E06 and E07, an average rupture velocity in the basement of 2.5 to 2.6 km/sec between the hypocenter and station E06 is obtained. In addition, several lines of evidence suggest that the Imperial fault dips about 75° to the NE.