Title: Senior CNRS Research Scientist
Country/Region: France
Period: 2025/6/20 - 2025/7/31
Theme: Aseismic Slip and Aftershocks During the Very Early POSTseismic Deformation (VE-POST): Intensive High-Rate GNSS Position Time
Host: Yuji ITOH
Introduction: Mathilde Vergnolle, Ph.D. in geophysics, has been a CNRS researcher since 2006 and works at the Géoazur laboratory (UniCA, OCA, CNRS, IRD). Her research focuses on transient aseismic deformation of the Earth's crust following or between earthquakes, using GNSS geodesy and modeling to understand the seismic cycle. She co-led the development of the GNSS European infrastructure within EPOS (2015–2019) and currently leads both the EPOS European GNSS Data Portal and the French GNSS data distribution service. Since 2015, she has led a project on very early postseismic deformation (funded by the French National Research Agency), enhancing our ability to observe and analyze the transition from fast to slow slip. In collaboration with Dr. Y. Itoh, this work now aims to expand observation of this early phase beyond a few case studies.
Research Report:
Report (June, 21-July, 30 2025)
Large earthquakes release and redistribute accumulated stress to the surrounding medium and on the fault, which induces postseismic aseismic transient deformation. It is well recognized that postseismic deformation significantly contributes to the deformation budget over a seismic cycle and redistributes stresses after the mainshock. Aseismic postseismic creep on the fault (afterslip) is considered one of the primary drivers of aftershocks. Large aftershocks can cause further damage to buildings already weakened by the mainshock, posing a major risk to vulnerable societies. Therefore, precisely documenting the postseismic deformation is essential.
The applicant has developed a dedicated high-rate kinematic GNSS processing scheme to capture the very early stage of the postseismic phase with high precision, referred to as VE-POST (i.e. from 5 minutes after the mainshock to the first standard daily positioning; Twardzik, Vergnolle et al., 2019). The study of VE-POST deformation following three major subduction earthquakes (Twardzik, Vergnolle et al., 2019) revealed that approximately 30% of the postseismic deformation within the first day after the mainshock is underestimated when VE-POST is not accounted for. In collaboration with Dr. Y. Itoh (ERI) and C. Twardzik (Géoazur), this work aims to expand observations of this early phase beyond isolated case studies.
A systematic global study is currently underway to evaluate VE-POST and its effect on fault slip budgets for large earthquakes worldwide. This novel and unique study, led by C. Twardzik, will provide statistically robust estimates of VE-POST’s contribution to the overall coseismic slip budget. and first results will be presented this fall at the Earth Science French conference (Twardzik, C. Y. Itoh, M. Vergnolle, 2025, RST, France). The analysis is based on 5-minute position time series provided by the Nevada Geodetic Laboratory (NGL), using their automated standard kinematic processing. The main advantage of using the NGL solution is the high density of station coverage, which makes such a global-scale study feasible. However, a custom kinematic processing pipeline, specifically tailored to capture VE-POST, has been shown to improve the quality of position time series. Yet, the software originally used for this (GIPSY-OASIS II, JPL) has been deprecated.
As planned during the applicant’s stay at ERI, the GNSS processing scheme is being revised and adapted for use with GipsyX (versions 2.3 and 2.4, JPL). The revised workflow requires fewer processing steps—only one (possibly two)—compared to the previous five-step procedure. It is no longer necessary to process pre- and post-earthquake data separately. A new software feature now enables the Kalman filter to be tuned with different noise models and levels depending on the epoch, allowing the use of white noise during the mainshock and random walk noise elsewhere. This improves the ability to account for large displacements during the earthquake relative to those expected during pre- and postseismic periods. First results from the updated processing will be presented at the "Slow-to-Fast Earthquake" workshop in September 2025 by Y. Itoh, who leads the study ("Revisiting the Early Postseismic Deformation of the 2003 Tokachi-Oki Earthquake," Itoh, Y., C. Twardzik, M. Vergnolle, L. Maubant, 2025). Ongoing testing is still needed to improve both the processing and post-processing of position time series—especially regarding discontinuities at UTC day transitions and the integration of multi-GNSS capabilities—before routine processing can begin.
In parallel, to better understand seismic–aseismic interactions, the applicant supervised a 5-week research internship, with C. Twardzik and participation of Y. Itoh. The project focused on assessing the ability of the rate-and-state friction law to explain aseismic slip evolution, particularly at the seismic-to-aseismic transition (Fukuda et al., 2009, 2013, 2021). M. Blacher, intern, analyzed jointly the global earthquake catalog and the NGL geodetic database to extract all magnitude ≥7.5 earthquakes for which high-rate position time series are available. She also implemented the full rate-and-state friction law and conducted initial optimizations using the custom-developed code. This work will continue and will serve as base for testing the sensitivity of this mechanical model—and its optimization procedures—to detect postseismic transient acceleration at the fast-to-slow slip transition (Ampuero and Perfettini, 2008). The methodology will be applied to real datasets, including newly processed VE-POST data, and to synthetic datasets, which still need to be generated.
As part of this ongoing project, two training research topics, are ready to be proposed by the applicant, in collaboration with Y. Itoh, for the start of the next academic year in France (September 2025). One will be aimed at Geodesy engineering students, and the other at Earth Science Master’s students. These topics will involve routine GNSS data processing, analysis of post-processed time series via slip inversion, generation of synthetic datasets with representative noise, and application of rate-and-state (and alternative) models to interpret VE-POST observations.
During her stay at ERI, the applicant also actively participated in meetings related to Louise Maubant’s postdoctoral project with Yuji Itoh. Discussions included GNSS processing strategies, post-processing methods and their limitations, and techniques used in seismology, all aimed at identifying subtle slow slip events (SSE) hidden at the day scale or velocity variation during SSEs. L. Maubant has joined the VE-POST group and attends all regular meetings. The applicant also participated in weekly group meetings organized by Dr. Y. Aoki and attended ERI seminars.
Report (June, 21-July, 30 2025) - Mathilde Vergnolle
Large earthquakes release and redistribute accumulated stress to the surrounding medium and on the fault, which induces postseismic aseismic transient deformation. It is well recognized that postseismic deformation significantly contributes to the deformation budget over a seismic cycle and redistributes stresses after the mainshock. Aseismic postseismic creep on the fault (afterslip) is considered one of the primary drivers of aftershocks. Large aftershocks can cause further damage to buildings already weakened by the mainshock, posing a major risk to vulnerable societies. Therefore, precisely documenting the postseismic deformation is essential.
The applicant has developed a dedicated high-rate kinematic GNSS processing scheme to capture the very early stage of the postseismic phase with high precision, referred to as VE-POST (i.e. from 5 minutes after the mainshock to the first standard daily positioning; Twardzik, Vergnolle et al., 2019). The study of VE-POST deformation following three major subduction earthquakes (Twardzik, Vergnolle et al., 2019) revealed that approximately 30% of the postseismic deformation within the first day after the mainshock is underestimated when VE-POST is not accounted for. In collaboration with Dr. Y. Itoh (ERI) and C. Twardzik (Géoazur), this work aims to expand observations of this early phase beyond isolated case studies.
A systematic global study is currently underway to evaluate VE-POST and its effect on fault slip budgets for large earthquakes worldwide. This novel and unique study, led by C. Twardzik, will provide statistically robust estimates of VE-POST’s contribution to the overall coseismic slip budget. and first results will be presented this fall at the Earth Science French conference (Twardzik, C. Y. Itoh, M. Vergnolle, 2025, RST, France). The analysis is based on 5-minute position time series provided by the Nevada Geodetic Laboratory (NGL), using their automated standard kinematic processing. The main advantage of using the NGL solution is the high density of station coverage, which makes such a global-scale study feasible. However, a custom kinematic processing pipeline, specifically tailored to capture VE-POST, has been shown to improve the quality of position time series. Yet, the software originally used for this (GIPSY-OASIS II, JPL) has been deprecated.
As planned during the applicant’s stay at ERI, the GNSS processing scheme is being revised and adapted for use with GipsyX (versions 2.3 and 2.4, JPL). The revised workflow requires fewer processing steps—only one (possibly two)—compared to the previous five-step procedure. It is no longer necessary to process pre- and post-earthquake data separately. A new software feature now enables the Kalman filter to be tuned with different noise models and levels depending on the epoch, allowing the use of white noise during the mainshock and random walk noise elsewhere. This improves the ability to account for large displacements during the earthquake relative to those expected during pre- and postseismic periods. First results from the updated processing will be presented at the "Slow-to-Fast Earthquake" workshop in September 2025 by Y. Itoh, who leads the study ("Revisiting the Early Postseismic Deformation of the 2003 Tokachi-Oki Earthquake," Itoh, Y., C. Twardzik, M. Vergnolle, L. Maubant, 2025). Ongoing testing is still needed to improve both the processing and post-processing of position time series—especially regarding discontinuities at UTC day transitions and the integration of multi-GNSS capabilities—before routine processing can begin.
In parallel, to better understand seismic–aseismic interactions, the applicant supervised a 5-week research internship, with C. Twardzik and participation of Y. Itoh. The project focused on assessing the ability of the rate-and-state friction law to explain aseismic slip evolution, particularly at the seismic-to-aseismic transition (Fukuda et al., 2009, 2013, 2021). M. Blacher, intern, analyzed jointly the global earthquake catalog and the NGL geodetic database to extract all magnitude ≥7.5 earthquakes for which high-rate position time series are available. She also implemented the full rate-and-state friction law and conducted initial optimizations using the custom-developed code. This work will continue and will serve as base for testing the sensitivity of this mechanical model—and its optimization procedures—to detect postseismic transient acceleration at the fast-to-slow slip transition (Ampuero and Perfettini, 2008). The methodology will be applied to real datasets, including newly processed VE-POST data, and to synthetic datasets, which still need to be generated.
As part of this ongoing project, two training research topics, are ready to be proposed by the applicant, in collaboration with Y. Itoh, for the start of the next academic year in France (September 2025). One will be aimed at Geodesy engineering students, and the other at Earth Science Master’s students. These topics will involve routine GNSS data processing, analysis of post-processed time series via slip inversion, generation of synthetic datasets with representative noise, and application of rate-and-state (and alternative) models to interpret VE-POST observations.
During her stay at ERI, the applicant also actively participated in meetings related to Louise Maubant’s postdoctoral project with Yuji Itoh. Discussions included GNSS processing strategies, post-processing methods and their limitations, and techniques used in seismology, all aimed at identifying subtle slow slip events (SSE) hidden at the day scale or velocity variation during SSEs. L. Maubant has joined the VE-POST group and attends all regular meetings. The applicant also participated in weekly group meetings organized by Dr. Y. Aoki and attended ERI seminars.
Report (June, 21-July, 30 2025) - Mathilde Vergnolle