Velocity structure of the shallow lunar crust

The data from the Apollo-14 and Apollo-16 Active Seismic Experiments have been reanalyzed and show that a power-law velocity variation with depth,v(z)110z ^1/6 m s^–1 (0<z<10 m), is consistent with both the travel times and amplitudes of the first arrivals for source-to-geophone separations up to 32 m. The data were improved by removing spurious glitches, by filtering and stacking. While this improved the signal-to-noise ratios, it was not possible to measure the arrival times or amplitudes of the first arrivals beyond 32 m. The data quality precludes a definitive distinction between the power-law velocity variation and the layered-velocity model proposed previously. However, the physical evidence that the shallow lunar regolith is made up of fine particles adds weight to the 1/6-power velocity model because this is the variation predicted theoretically for self-compacting spheres.The 1/6-power law predicts the travel time,t(x), varies with separation,x, ast(x)=t ₀(x/x ₀)^5/6 and, using a first-order theory, the amplitude,A(x), varies asA(x)=A ₀(x/x ₀)^{–(13–m)/12},m>1; the layervelocity model predictst(x)=t ₀(x/x ₀) andA(x)=A ₀(x/x ₀)^–2, respectively. The measured exponents for the arrival times were between 0.63 and 0.84 while those for the amplitudes were between –1.5 and –2.2. The large variability in the amplitude exponent is due, in part, to the coarseness with which the amplitudes are measured (only five bits are used per amplitude measurement) and the variability in geophone sensitivity and thumper-shot strengths.A least-squares analysis was devised which uses redundancy in the amplitude data to extract the geophone sensitivities, shot strengths and amplitude exponent. The method was used on the Apollo-16 ASE data and it indicates there may be as much as 30 to 40% variation in geophone sensitivities (due to siting and coupling effects) and 15 to 20% variability in the thumper-shot strengths. However, because of the low signal-to-noise ratios in the data, there is not sufficient accuracy or redundancy in the data to allow high confidence in these results.