Bayesian Analysis of Parameter Correlarions in the Cole-Cole Model for Spectral Induced Polarization Tomography

This research explores the interdependence and correlation of Cole-Cole model (CCM) parameters by applying Bayesian inference to invert spectral induced polarization (SIP) data. The goal is to enhance the interpretation of subsurface properties by closely examining parameter relationships. We introduce an innovative ۲.۵D inversion algorithm tailored for SIP data, built with Python-based tools and advanced statistical methods. Using synthetic models and Markov chain Monte Carlo (McMC) inversion, we test our approach across various subsurface configurations, including a homogeneous earth model, a two-layer setup, and a scenario with two anomalies within an homogeneous background. Visualizations, including McMC chains and corner plots, reveal interdependencies among CCM parameters, highlighting the convergence and robustness of our estimates. By validating against synthetic models, we demonstrate the precision and effectiveness of our method. Ultimately, this study underscores the potential of Bayesian inversion to improve geophysical data interpretation and deepen our understanding of the relationships between CCM parameters in estimated models across diverse geological environments.

کلیدواژه ها:

Bayesian inference ، Cole-Cole model ، McMC sampling ، Spectral Induced Polarization ، Uncertainty analysis