A search for the minimum number of stations needed for seismic networking on Mars

The efficiency of a seismic network in providing information on the rate of seismicity, and on the inner structure of Mars, is estimated through a sta tistical analysis which takes into account the possible existence of a liqu id core, the expected low rate of seismicity of Mars when compared to the E arth's, and the attenuating properties of the mantle. The tests are perform ed for two frequency ranges (0.1-1.0 Hz and 0.5-2.5 Hz), for three instrume ntal noise amplitude densities ranging from 5 to 500 x 10(-10) m s(-2) Hz(- 1/2), and for three network configurations consisting of 4, 12 and 16 stati ons. Travel lime tables are computed for P, S, PcP, ScS, and PKP phases usi ng a simplified three layer model. Present-day estimates of liquid core rad ius induce a 25 degrees wide shadow zone beginning at epicentral distances larger than 110 degrees. Consequently, the best detection efficiency which can be expected from any network is of the order of 60% for mantle body wav es. The detection efficiency is primarily controlled by the instrumental no ise level. Since the amplitude of mantle body waves rapidly decreases with epicentral distance, high noise level instruments can only detect local eve nts. Therefore, the detection score attained by 4 highly sensitive stations can be up to 30 and 7 times better than the score attained by 12 high nois e level sensors, for mantle P and S waves, respectively. If crustal scatter ing is negligible, the record of mantle P waves on a network consisting of four low noise level instruments would permit to sample Mars' mantle down t o the coremantle boundary. Conversely, the deepest penetration of rays reco rded by a network of 12 high noise level sensors would hardly reach 300 km. In fact, strong crustal scattering might be the most important difficulty to be encountered in a seismic exploration of Mars. A possibility to deal w ith this problem would be to associate each of the four low noise instrumen ts with three medium noise level sensors. This network strategy might permi t to sample P and S mantle waves travelling down to 400-600 km, even if a l ot of seismic energy is lost through crustal scattering. (C) 1999 Elsevier Science Ltd. All rights reserved.