GNGTS 2024 - Atti del 42° Convegno Nazionale

Session 3.3 GNGTS 2024 Amplitude and traveltme inversion for mono- channel Boomer surveys A. Vesnaver, L. Baradello Department of Geophysics, Isttuto Nazionale di Oceanografa e Geofsica Sperimentale - OGS, Trieste, Italy INTRODUCTION Mono-channel recording systems with a Boomer seismic source are very cheap and can be easily deployed in sensitive environments such as lagoons or busy harbours (Zecchin et al. 2008). The price paid for these advantages is the lack of signal redundancy typical of multi-channel records, which makes it possible to estimate wave propagation velocity and angle-dependent reflectivity, and to improve the signal-to-noise ratio by stacking or migration. In this paper, we show that some of this information can be obtained by inverting the amplitudes and traveltimes of shallow primary reflections and their multiples, using a single offset in a Boomer survey. Amplitudes and traveltimes can in principle be inverted separately, but doing so we do not use the information redundancy embedded in the velocity: it determines both the traveltimes along the ray paths and the amplitude of primaries and multiple reflections via the acoustic impedance contrasts at the layer interfaces. Therefore, the coupling of these two inversion algorithms can extract more information from our minimal data set. The possible ambiguities of one inversion can be limited by constraints coming from the other inversion, so improving the stability of both. AMPLITUDE AND TRAVELTIME INVERSION The simplest object function we can create for a joint inversion of amplitudes and traveltimes is the sum of the squared differences between measured and modeled data, minimizing it as a function of the Earth model parameters: Object ( T j , A j , V, L, ρ ) = + , (1) where T j and A j are the measured traveltimes and amplitudes of primaries and multiples in a single trace. We assume a 1D Earth model, with the data compensated for the geometrical spreading – (e.g., by a t 2 gain function). We note that the modeled traveltimes t ( V, L ) depend on the layer velocity V and the layer thickness L , but not on the density ρ . Similarly, the modeled amplitudes a ( V , ρ ) do not depend on the thickness L . Therefore, a separate inversion of amplitudes and traveltimes can avoid cross-talk between density and thickness. On the other hand, the velocity V influences both the amplitude (via the acoustic ∑ j [ T j − t ( V , L )] 2 ∑ j [ A j − a ( V , ρ ) ] 2