We present an adaptive double-focusing method for applying source-receiver Marchenko redatuming to field data. Receiver redatuming is achieved by a first focusing step, where the coupled Marchenko equations are iteratively solved for the oneway Green’s functions. Next, source redatuming is typically performed by a multi-dimensional deconvolution of these Green’s functions. Instead, we propose a second focusing step for source Marchenko redatuming, using the upgoing Green’s function and the downgoing focusing function to obtain a redatumed reflection response in the physical medium. This method makes adaptive processing more straight-forward, making it less sensitive to imperfections in the data and the acquisition geometry and more suitable for the application to field data. In addition, it is cheaper and can be parallelized by pair of focal points.

We propose an adaptive approach for the removal of internal multiples caused by an overburden using source-receiver Marchenko redatuming. Typically, a multi-dimensional deconvolution using the one-way Green’s functions is performed to achieve redatuming, after these functions have been retrieved by solving the coupled Marchenko equations. However, this processing step is sensitive to imperfections in the data and the acquisition geometry as well as computationally expensive. We propose an adaptive redatuming method that is less sensitive to such imperfections and as such should be beneficial when attempting source-receiver Marchenko redatuming with field data. We show this using 2D synthetic data from the Santos basin offshore Brazil. As an added bonus, it is also computationally less expensive. A disadvantage of the proposed method is that the obtained reflection response exists in the physical medium, causing some interactions with the overburden to remain.