SigMT

SigMT is an open-source Python package developed for the processing of magnetotelluric (MT) time series data to estimate MT impedance and tipper responses. The package enables data processing with minimal manual intervention by utilizing a Mahalanobis distance based data selection tool. Final estimation is performed using a robust statistical approach. In addition, SigMT includes noise rejection tools such as coherency thresholding and polarization direction analysis to enhance the reliability of the estimated MT responses. The package currently supports Metronix and Phoenix (MTU-5C) time series formats through a graphical user interface.

The foundations of SigMT began at the end of 2019—before the project had a name—as a collection of experimental codes written out of curiosity to explore magnetotelluric data processing. At that time, I found manual time series editing very time consuming and wanted to develop a smarter, more automated approach. During the 2020 COVID-19 lockdown, the availability of additional free time allowed these initial ideas to evolve into a complete and structured processing pipeline. In particular, the use of Mahalanobis distance based data selection, as described by Platz and Weckmann (2019), significantly improved the results. By the end of 2022, the project was released on GitHub as a script-based package with core functionalities, marking its first public version. This effort culminated in the 2023 publication describing the methodology and implementation. SigMT continues to be maintained and developed as a personal project, with ongoing improvements driven by user feedback.