Welcome to LASIF 2.0


LASIF (LArge-scale Seismic Inversion Framework 2.0) is a data-driven end-to-end workflow tool to perform adjoint full seismic waveform inversions.

Github: The code is developed at and available from its Github repository.

For the initial version of LASIF, we refer to its repository and its corresponding documentation.

If you use LASIF in your project, please consider citing our paper(s). If you use this version of LASIF, please refer to this paper:

Latest paper

Solvi Thrastarson, Dirk-Philip van Herwaarden, Lion Krischer and Andreas Fichtner (2021)
LASIF: LArge-scale Seismic Inversion Framework, an updated version,

    doi = {10.31223/x5nc84},
    url = {https://doi.org/10.31223%2Fx5nc84},
    year = 2021,
    month = {jan},
    publisher = {California Digital Library ({CDL})},
    author = {Solvi Thrastarson and Dirk-Philip van Herwaarden and Lion Krischer and Andreas Fichtner},
    title = {{LASIF}: {LArge}-scale Seismic Inversion Framework, an updated version}

For more details, we refer to the paper associated with the original version of LASIF.

Original LASIF paper

Lion Krischer, Andreas Fichtner, Saule Zukauskaite, and Heiner Igel (2015),
Large‐Scale Seismic Inversion Framework,
Seismological Research Letters, 86(4), 1198–1207.

    title={Large-scale seismic inversion framework},
    author={Krischer, Lion and Fichtner, Andreas and Zukauskaite, Saule and Igel, Heiner},
    journal={Seismological Research Letters},
    publisher={Seismological Society of America}

Dealing with the large amounts of data present in modern full seismic waveform inversions in an organized, reproducible and shareable way continues to be a major difficulty potentially even hindering actual research. LASIF improves the speed, reliability, and ease with which such inversion can be carried out.

Full seismic waveform inversion using adjoint methods evolved into a well established tool in recent years that has seen many applications. While the procedures employed are (to a certain extent) well understood, large scale applications to real-world problems are often hindered by practical issues.

The inversions use an iterative approach and thus by their very nature encompass many repetitive, arduous, and error-prone tasks. Amongst these are data acquisition and management, quality checks, preprocessing, selecting time windows suitable for misfit calculations, the derivation of adjoint sources, model updates, and interfacing with numerical wave propagation codes.

The LASIF workflow framework is designed to tackle these problems. One major focus of the package is to handle vast amount of data in an organized way while also efficiently utilizing modern HPC systems. The use of a unified framework enables reproducibility and an efficient collaboration on and exchange of tomographic images.