← Back to Sessions

Characterizing Surface Fault Displacement Uncertainty and Its Effects on Probabilistic Fault Displacement Hazard. Example From the 2023 m7.8 Pazarcık, Türkiye Earthquake

Description: 

Surface fault displacement poses a significant threat to distributed infrastructure systems, and as such, they need to withstand long return period events. Thus, properly characterizing the upper percentiles of the fault displacement distribution has a large impact on the design requirements. Using the 2023 M7.8 Pazarcık earthquake as an example, we developed an open-source software to characterize the measurement uncertainty in surface fault displacement of Global navigation satellite system (GNSS) fault-perpendicular profiles of offset features and rank the contribution of each component. For every iteration, both sides of the surveyed feature are extended into the rupture zone based on a principal component projection, and the displacement is measured parallel to the strike of the rupture. Considered components of uncertainty include: (1) location error in the horizontal and vertical direction, (2) rupture location uncertainty, (3) rupture azimuth uncertainty, and (4) interpretation uncertainty on which survey points are used to project the feature into the fault zone. We compared the distribution of estimated displacements against field-based methods and evaluated the impact of displacement uncertainty in the development of fault displacement models (FDMs). Ongoing analyses indicate that field-based displacement values are generally in agreement with the mean of the modeled displacement distributions and that reported minimum and maximum displacements correspond to the modeled 2nd and 98th percentiles. To quantify the impact of displacement uncertainty on FDMs, we organized the collected profiles into good, mediocre, and poor quality, estimated the uncertainty for each class, and generated synthetic datasets based on the previous rules. Regressing FDMs with and without considering the uncertainty of displacements showed that omitting the uncertainty of displacement during the regression process inflates the aleatory variability, which overestimates design values at large return periods.

Session: Learning Across Geological, Geophysical & Model-Derived Observations to Constrain Earthquake Behavior - III

Type: Oral

Date: 5/1/2024

Presentation Time: 02:30 PM (local time)

Presenting Author: Grigorios

Student Presenter: No

Invited Presentation: 

Authors