Revealing 4D Subsidence with 3D Water-Bottom Traveltime Inversion

The quality of time-lapse analysis depends highly on the repeatability of the acquisition. However, in practice, it is almost impossible to perfectly mimic a base survey due to environmental conditions and inaccurate measurements. Lack of repeatability often results in 4D noise, which may compromise the 4D signal. In the presence of subsidence, caused by the depletion of the reservoir, 4D signal exists outside of the reservoir area, and its extraction from noisy 4D data can be challenging without a priori information. Water layer tomography has already been proposed to recover uncertain parameters from the acquisition in order to address the non-repeatability effects in the data but not with as many parameters as presented in this paper: water velocity, source position, top of the water layer and start of data time. Unlike most water layer tomography, our method not only relies on the inversion of the water-bottom primary but also of the first-order multiple travel times picked in the data. An application to a 3D deep-water survey offshore Angola is presented. The flow is applied independently to all vintages of a 4D project resulting in significant reduction of the 4D noise and a clear visibility of the subsidence.