ON ACH, OR HOW RELIABLE IS REGIONAL TELESEISMIC DELAY-TIME TOMOGRAPHY

ACH (named after Aki et al., 1976, Bull. Seismol. Sec. Am., 66: 501-52 4; Aki et al., 1977, J. Geophys. Res., 82: 277-296) is a standard, wid ely used, method for three-dimensional seismic imaging of the Earth. T he fundamental hypothesis which underlies the method is that the time residuals generated outside the given target volume (from the seismic source to the bottom of the modelled zone) are approximately constant across the seismic array. The main purpose of this study is to check t his assumption. We computed travel times for a given station and event distribution using a three-dimensional global Earth model taken from seismic tomography. We found that the relative residuals generated out side the target volume are not negligible and that the fundamental hyp othesis underlying ACH is thus not verified. These deviations are gene rated in the lower and/or upper mantle and the corresponding proportio ns are entirely dependent on the ray paths. The bias in the inverted m odel is statistically similar to the input model outside the target vo lume. We thus recommend caution in any interpretations involving AGH-g enerated models. A secondary, somewhat independent, outcome of this st udy is that Fermat's principle, used to linearize the inverse problem in ray theory based tomography, seems to be valid without any restrict ions (given our input model is representative for the true Earth) for rays with turning points in the lower mantle. For rays with turning po ints in the upper mantle, the constant ray path approximation is proba bly not true. This applies to global tomography as well. (C) 1997 Else vier Science B.V. (C) 1997 Elsevier Science B.V.