Height changes in the epicentral region preceding the January 17, 1994 Northridge earthquake

Abstract

Analysis of the results of repeated levelings through the epicentral region of the Mw6.7, 1994 Northridge earthquake has disclosed the occurrence of differential uplift in this area that preceded the earthquake. Although the distribution of the relevant vertical-control data is somewhat sparse, in both space and time, those data that we have recovered indicate that this uplift exceeded 0.10 m and peaked 20-25 km west of the 1994 epicenter. While our data also indicate that this deformational event must have occurred during the period 1978-1989, evidence based on the character and magnitude of misclosures developed from 1987 and 1989 surveys argue that the deformation occurred largely during the period 1987-1989. The preseismic vertical-displacement field that preceded the Northridge earthquake is similar in form and, less certainly, magnitude to that which preceded the Mw6.7, 1971 San Fernando earthquake; other possible, but less significant southern California analogues include the deformational events that preceded the Mw5.3, 1973 Point Mugu and Mw5.9, 1987 Whittier Narrows earthquakes. The small but growing number of recognized aseismic deformational episodes that preceded small to moderate magnitude earthquakes in southern California suggests that the deployment of the dense array of continuously recording GPS receivers planned for southern California can be expected to detect and more accurately describe such events than was heretofore possible. Moreover, if the relation between the duration of these deformational anomalies and the magnitudes of any ensuing earthquakes can be much more clearly established, the near perfect temporal control on position afforded by GPS suggests that we may be on the threshold of a realistic earthquake warning system. I n t r o d u c t i o n Owing to severe funding constraints experienced by a number of local jurisdictions, geodetic control work had virtually ceased in southern California by 1979. It was, in fact, not until the late 1980s, and owing largely to the efforts of the National Geodetic Survey to bring to completion the new North American Vertical Datum, that any significant number of higher order vertical-control surveys were again carried out in this area. Several of these vertical-control lines fell within the subsequently defined epicentral region of the Mw6.7 Northridge earthquake. Analysis and comparisons of these surveys, though sparsely distributed, against the results of still earlier levelings has disclosed compelling evidence of an intriguing preseismic deformational event, especially when viewed as an indication of what can be expected to emerge with the anticipated installation of a dense array of GPS observatories in southern California, where this array is integrated in turn with interferometric synthetic aperature radar (InSAR) imagery in this region. While a fairly clearly defined association between a regional deformational event and subsequent earthquake activity has been recognized in southern California (Castle and Bernknopf, 1996), localized preseismic deformational anomalies, such as the one described here, are seemingly rare, whether in southern California or elsewhere. However, their detection has previously depended on the spatial and temporal distribution of those surveys that led to their recognition, surveys not generally deployed with geophysical needs in mind. Thus uncertainty surrounding the frequency of occurrence of these preseismic events, as well, of course, as similar events tha t show no demonstrable associations with subsequent seismicity, remains unresolved. Discrimination between these separable types of deformational events is especially perplexing and probably will remain so until our data reservoir is vastly enlarged. Nevertheless, current uncertainties should