# Virtual sources and receivers in the real Earth -- considerations for practical applications

## TL;DR

This paper discusses a method to accurately retrieve homogeneous Green's functions from field seismic data using a single-sided scheme, improving subsurface monitoring and source localization capabilities.

## Contribution

It introduces a practical single-sided Green's function retrieval method applicable to real data, overcoming classical limitations requiring an enclosing surface.

## Key findings

- Single-sided Green's function retrieval works on synthetic and real data.
- Recording limitations significantly affect the accuracy of the retrieval.
- The method has potential for future applications like source localization.

## Abstract

To enhance monitoring of the subsurface, virtual sources and receivers inside the subsurface can be created from seismic reflection data at the surface of the Earth using the Marchenko method. The response between these virtual sources and receivers can be obtained through the use of homogeneous Green's function retrieval. A homogeneous Green's function is a superposition of a Green's function and its time-reversal. The main aim of this paper is to obtain accurate homogeneous Green's functions from field data. Classical homogeneous Green's function retrieval requires an unrealistic enclosing recording surface, however, by using a recently proposed single-sided retrieval scheme, this requirement can be avoided. We first demonstrate the principles of using the single-sided representation on synthetic data and show that different source signatures can be taken into account. Because the Marchenko method is sensitive to recording limitations of the reflection data, we study five cases of recording limitations with synthetic data and demonstrate their effects on the final result. Finally, the method is demonstrated on a pre-processed field dataset which fulfills the requirements for applying the single-sided Green's function retrieval scheme. The scheme has the potential to be used in future applications, such as source localization.