I am a Geophysicist and AI researcher specializing in the application of deep learning, machine learning and high performance computing in Geosciences. As the leader of an independent research group, I develop data-driven methods to analyze large-scale seismic datasets, with applications in earthquake source processes, Early Warning Systems (EWS), and Seismic Signal Analysis. My work includes creating open-source tools such as AWESAM and SAIPy in collaboration with my team and securing approximately €2 million in research funding to advance AI-driven seismic analysis.

Institute Webpage

email: srivastava@fias.uni-frankfurt.de

nishtha.vns@gmail.com

Linkedin • Google Scholar • Twitter • Bluesky

Research Position

• Independent Research Group Leader, Frankfurt Institute for Advanced Studies (FIAS), Germany (August 2020 – till date)
• Postdoctoral Researcher, Frankfurt Institute for Advanced Studies (FIAS), Germany (October 2018 – July 2020)

Funding Acquired (around €2 million)

• Seismology and AI Project funded by the Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung – BMBF).
• EQUATE Project funded by the 2023 edition of the Partenariat Hubert Curien (PHC) Procope Program between Germany and France in collaboration with CentraleSupélec, Université Paris-Saclay.
• GANANA Project for Europe-India Partnership for Scientific High-Performance Computing funded by Horizon Europe funding programme.

Teaching Experience

• Guest Lecturer in Faculty of Geosciences and Geography, Goethe University, 60438 Frankfurt am Main, (April 2022 – till date)
• Trained the scientist of Institute of Seismological Research (ISR) Gandhinagar and National Centre of Seismology (NCS), Ministry of Earth Sciences (MoES), New Delhi on ‘Introduction to Deep Learning’ through an Intensive 4-day workshop at ISR (February 2023)

MoU signed

• Established five years Research co-operation between Frankfurt Institute for Advanced Studies, Germany and Institute of Seismological Research Gandhinagar India

Research Events organized

• Co-convener in European Geosciences Union 2024 in the session titled Interfacing machine learning and numerical modelling - challenges, successes, and lessons learned
• Organized Seismology and Artificial Intelligence workshop funded by the Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung - BMBF) from 13th September to 15 September 2023.
• Co-convener in European Geosciences Union 2023 in the session titled Hybrid modeling of natural hazards: blending deep-learning, data-driven approaches and physics-based simulations
• Convener in Seismological Society of America 2023 in the session titled Opportunities and Challenges for Machine Learning Applications in Seismology
• Organized the Kickoff SAI workshop funded by the Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung - BMBF) from 27th September to 1st October 2021.

Research Team

Research Projects

SAIPy - An open source python package for Earthquake Monitoring using deep learning

We have developed an open-source Python package - SAIPy for fast seismic waveform data processing by implementing deep learning. SAIPy is an open-source python library where we strive to combine our previously published models into an automated pipeline for monitoring continuous seismic data so that it can be easily implemented by seismologists. It offers solutions for multiple seismological tasks such as earthquake detection, magnitude estimation, seismic phase picking, and polarity identification.

Earthquake Forecasting

Reliable earthquake forecasting methods have long been sought after, and so the rise of modern data science techniques raises a new question: does deep learning have the potential to learn this pattern? In this study, we leverage the large amount of earthquakes reported via good seismic station coverage in the subduction zone of Japan. We pose earthquake forecasting as a classification problem and train a Deep Learning Network to decide, whether a timeseries of length ≥ 2 years will end in an earthquake on the following day with magnitude ≥ 5 or not.

High magnitude event detection in HR-GNSS data using deep learning

High-rate Global Navigation Satellite System (HR-GNSS) data can be highly useful for earthquake analysis as it provides continuous high-rate measurements of ground motion. In this project, we are attempting to identify high magnitude event and estimate the magnitude using HR-GNSS displacement time series.

AWESAM - Seismo-Volcanic event Analysis

Many active volcanoes exhibit Strombolian activity, which is typically characterized by relatively frequent mild volcanic explosions and also by rare and much more destructive major explosions and paroxysms. Catalogs of these eruptions and, specifically, seismo-volcanic events may be generated using continuous seismic recordings at stations in the proximity of volcanoes. We developed an automated user-friendly, time-saving, automated approach via a python package labelled as: the Adaptive-Window Volcanic Event Selection Analysis Module (AWESAM). This strategy of creating seismo-volcanic event catalogs consists of three main steps: 1) identification of potential volcanic events based on squared ground-velocity amplitudes, an adaptive MaxFilter, and a prominence threshold. 2) catalog consolidation by comparing and verifying the initial detections based on recordings from two different seismic stations. 3) identification and exclusion of signals from regional tectonic earthquakes. The strength of the python package is the reliable detection of very small and frequent events as well as major explosions and paroxysms. We applied AWESAM to seismic data from Stromboli (Italy), Mount Etna (Italy), Yasur (Vanuatu) and Whakaari (New Zealand) and performed an inter-event time analysis to identify characteristic patterns in the events’ recurrence time and the volcanic activity. We also derive a new amplitude-frequency relationship from seismo-volcanic events. With this relation, we can confirm a change in slope for large events at Stromboli, which is based on ten years of data.