High-resolution seismic reflection profiling is increasingly being used to map faults beneath urban areas and important facilities. To illustrate this use, we give examples of marine profiling efforts on the Columbia and Willamette Rivers in the city of Portland, Oregon, and along the Panama Canal Zone in the Republic of Panama. In both of these examples the profiles were designed to cross suspected fault zones, but new features were also discovered while profiling away from known faults.

In Portland, we concurrently used single-channel Uniboom and bubble-pulser marine seismic systems to collect ~130 kilometers of profile data along the Willamette and Columbia Rivers, crossing four suspected fault zones. The data showed a remarkably strong reflection from the unconformity at the base of Late Pleistocene flood deposits at 0 to 85 m depth. The amplitude of this reflector suggests it will produce resonances that will have a strong influence on earthquake-induced ground shaking in the 1 to 10 Hz range. The unconformity deepens to almost 85 m depth within a 1.5-km-wide, prehistoric river channel lying beneath and south of the Columbia River. Strata within this ancient channel appear to be broken by the East Bank fault running just north of downtown Portland, indicating at least one active fault may lie directly beneath the city.

In Panama we collected over 90 km of marine seismic profiles near the north end of the Panama Canal. Although the profiles crossed one known fault (the Rio Gatún fault), the dense vegetation on land limits detection of faults in the canal region. The seismic profiling was used to determine whether other faults are present beneath the canal and whether any such faults are active. The profiles from Limón Bay at the north end of the canal showed a sequence of middle to late Miocene strata (Gatún Formation) containing prominent, parallel reflectors broken in at least thirteen locations by obvious faulting. The parallel bedding between the faults indicates the faulting post-dates deposition of these late Miocene strata, but younger strata are not present along the seismic profiles to determine whether there has been Pliocene or Quaternary motion on the faults. Because of this uncertainty in age of motion, subsequent hazard analyses have treated these faults in a probabilistic rather than deterministic manner.