Ultra-low memory seismic inversion with randomized trace estimation

TL;DR

This paper introduces a low-memory seismic inversion method using randomized linear algebra, enabling efficient and competitive results with minimal memory, suitable for GPU acceleration.

Contribution

The paper presents a novel seismic inversion technique that significantly reduces memory requirements by employing randomized trace estimation, facilitating large-scale wavefield matrix computations.

Findings

Achieves competitive inversion results with reduced memory usage

Demonstrates effectiveness on synthetic seismic data

Enables GPU-accelerated seismic inversion

Abstract

Inspired by recent work on extended image volumes that lays the ground for randomized probing of extremely large seismic wavefield matrices, we present a memory frugal and computationally efficient inversion methodology that uses techniques from randomized linear algebra. By means of a carefully selected realistic synthetic example, we demonstrate that we are capable of achieving competitive inversion results at a fraction of the memory cost of conventional full-waveform inversion with limited computational overhead. By exchanging memory for negligible computational overhead, we open with the presented technology the door towards the use of low-memory accelerators such as GPUs.