Scattering from near-source topography: Teleseismic observations and numerical simulations

Abstract

Two-dimensional linear elastic finite-difference simulations of teleseismic P waveforms from point sources in two-dimensional structure have been used to explore the variations that may be induced in event magnitude-yield determination by the emplacement of explosive sources under mountainous topographic features. The southern Sahara French test site in Algeria, at Taourirt Tan Afella Massif on the Ahaggar plateau has been used as a case study. The topography of this test site is extreme, and the event locations permit a test of the hypothesis that topography influences short-period magnitudes of contained nuclear explosions. The maximum expected variation is plus or minus 0.15 magnitude units from the network mean. The magnitude variations are expected to change rapidly with take-off angle and azimuth. Teleseismic observations of the explosions at the southern Sahara test site are compared to predictions made from two-dimensional linear finite-difference simulations. Waveform data from the arrays EKA and YKA as well as LRSM data have been deconvolved to broadband displacement for inspection of the apparent far-field P -wave source. Qualitative comparisons are favorable that the topography above the explosions RUBIS and SAPHIR defocused teleseismic pP at certain take-off angles and azimuths. Long-period positive-polarity pulses can be seen at several sites that may indicate Rayleigh-to- P scattering from topography near the source. WWSSN maximum-likelihood magnitude data for the “ a ,” “ ab ,” and “ max ” P phases have been used to estimate that the magnitude variation due to topographic scattering is no more than 0.15 rms magnitude units across the WWSSN for the southern Sahara test site.