255-11 Efficient Sensitivity Analysis in Thermal-Kinematic Modeling Using the Taguchi Method

Session: Broad Applications of Thermochronology to Understanding Geologic Rates and Processes Through the Sedimentary Record

Presenting Author:

Authors:

Abstract:

Thermochronologic data are commonly used to reconstruct exhumation histories in orogenic systems through inverse thermal-kinematic modeling. However, the uncertainties reported by such models typically reflect only the statistical spread of acceptable solutions for a particular dataset and do not account for the broader range of plausible outcomes resulting from uncertainties in all model inputs. In this study, we introduce the use of a popular method in engineering circles, the Taguchi design of experiments, to conduct more comprehensive sensitivity analyses. A worked example is based on a previously published model that inverts a dataset containing multiple chronometric ages derived from a single sample in the central Nepalese Himalaya. We independently vary thermal, deformational, and kinetic parameters and evaluate their influence on modeled exhumation rates, finding that the diffusion activation energies for the utilized chronometric systems are the most influential parameters. Our analysis also reveals that, while some parameters (e.g., deformational properties) influence the entire exhumation history, others (e.g., kinetic parameters) significantly impact only specific time slices. This effect is compounded in multichronometer models. Compared with formal uncertainties from inversion alone, the expanded uncertainty envelope derived from input sensitivity is substantially larger. For example, a 10% variation in activation energy produces a ~5% change in the absolute percentage difference in exhumation rates between perturbed models and the reference, over 100 times the influence of the least sensitive parameter. These results suggest that the topology of model-derived exhumation histories is likely more robust than the absolute rates derived from the models. Thermochronologic interpretations should account for unquantifiable uncertainty in input parameters to avoid overconfidence in modeled exhumation rate estimates.

Geological Society of America Abstracts with Program. Vol. 57, No. 6, 2025

doi: 10.1130/abs/2025AM-11116

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.