A field program was conducted during the summer of 1977 in the Appalachian Mountains to study some of the effects of topography on seismic-wave motions. The study areas included Powell Mountain (394 m of relief) in Lee County, Virginia, Bays and River Mountains (260 m of relief) in Sullivan County, Tennessee, and Gap and Brush Mountains (236 m of relief) in Montgomery County, Virginia. The majority of the data were recorded using 1 Hz horizontal seismometers aligned parallel to the long dimension of the northeast trending ridges. Some vertical (1 Hz) data were also collected. The signal sources were quarry and mine blasts located mostly to the west and northwest of the study areas. The Tennessee earthquake of July 27, 1977 [mb (Lg) = 3.5] was recorded at Bays and River Mountains by both horizontal and vertical sensors.
The basic data consist of maximum amplitudes and their associated periods measured from the Sg-Lg coda (some Rayleigh-wave and P-wave data were also obtained) at various sites on the ridges and in the valleys. In total, 137 events were considered from which 444 measurements of trace amplitude and wave period were made. These were cast in the form of 382 amplitude ratios between the individual sites. Four of the five ridges (Powell Mountain, Gap Mountain, Bays Mountain, and River Mountain) show amplification at the mountaintops with respect to the valleys. However, one ridge, with a less pronounced crest than the others (Brush Mountain), exhibited some wave-amplitude suppression. The vertical data obtained display a lesser degree of amplification than do the horizontal data as might be expected from other published observations.
To reduce the scatter found in the individual amplitude ratios, average amplitude ratios between mountaintop and valley sites were determined. These average ratios showed that the seismic wave amplitudes at the mountaintops were amplified by factors from 1.7 to 3.4. Employing the theory for scattering of SH waves in a half-space with an irregular free surface, the predicted ground motions for each of the study areas were obtained. The comparison of the observed average ratios with these theoretical calculations indicates that a topographic effect is present but inadequately modeled by the theory.