Abstract:Iterative geostatistical seismic inversion integrates seismic and well data to infer the spatial distribution of subsurface elastic properties. These methods provide limited assessment to the spatial uncertainty of the inverted elastic properties, overlooking alternative sources of uncertainty such as those associated with poor well-log data, upscaling and noise within the seismic data. We express uncertain well-log samples, due to bad logging reads and upscaling, in terms of local probability distribution functions (PDFs). Local PDFs are used as conditioning data to a stochastic sequential simulation algorithm, included as the model perturbation within the inversion. The problem of noisy seismic and narrow exploration of the model parameter space, particularly at early steps of the inversion, is tackled by the introduction of a cap on local correlation coefficients responsible for the generation of the new set of models during the next iteration. A single geostatistical framework is proposed and illustrated with its application to a real case study. When compared against a conventional iterative geostatistical seismic inversion, the integration of additional sources of uncertainty increases the match between real and inverted seismic traces and the variability within the ensemble of models inverted at the last iteration. The selection of the local PDFs plays a central role in the reliability of the inverted elastic models. Avoiding high local correlation coefficients at early stages of the inversion increases convergence in terms of global correlation between synthetic and real seismic reflection data at the end of the inversion.

What problem does this paper attempt to address?

This paper attempts to solve the problem of integrating uncertainty in iterative geostatistical seismic inversion. Specifically, existing iterative geostatistical seismic inversion methods often overlook uncertainties from multiple sources when evaluating the spatial distribution of subsurface elastic properties, such as poor well - log data quality, uncertainties in the up - scaling process of data, and noise in seismic data. These problems may reduce the reliability of inversion results, especially in the early iteration process. Due to the overly high local correlation coefficient, over - fitting may occur, reducing the exploration range of the model parameter space. To overcome these challenges, the author proposes a new geostatistical framework that can integrate the above - mentioned types of uncertainty. The main methods include: 1. **Uncertain well - log data**: Represent the uncertainty of well - log data caused by wellbore collapse or measurement error by introducing a local probability distribution function (PDF). These local PDFs can be derived based on prior knowledge, pseudo - well - log values, or other geological models. 2. **Uncertainty in the up - scaling process of well - log data**: When up - scaling high - resolution well - log data to the seismic grid, use local PDFs instead of a single value to represent the distribution of high - resolution measurements within each grid cell. This can avoid the problem of choosing a single value to characterize the cell distribution, thereby better preserving the statistical characteristics of the original data. 3. **Uncertainty in the recorded seismic reflection data**: Control the convergence speed of the inversion process by limiting the maximum value of the local correlation coefficient to prevent over - fitting in the early iterations. This method is similar to the temperature control mechanism in the simulated annealing algorithm, allowing the acceptance of poorer solutions in the initial iterations and finally converging to better solutions. Through these improvements, the author aims to improve the reliability and accuracy of the inversion results and provide a more comprehensive uncertainty assessment. The paper also verifies the effectiveness of the proposed method through a practical case study, demonstrating the advantages of the new method in matching real and inverted seismic traces and increasing the variability of the model ensemble compared with traditional iterative geostatistical seismic inversion.

Strategies for integrating uncertainty in iterative geostatistical seismic inversion

Multi-scale uncertainty quantification in geostatistical seismic inversion

Uncertainty Quantification in Seismic Inversion Through Integrated Importance Sampling and Ensemble Methods

Geostatistical Rock Physics AVA Inversion

Geostatistical inversion method based on seismic waveform similarity

Efficiency and robustness in Monte Carlo sampling of 3-D geophysical inversions with Obsidian v0.1.2: Setting up for success

Geostatistical Seismic Inversion Constrained by Reservoir Sedimentary Structural Features

Improving Seismic Inversion Robustness via Deformed Jackson Gaussian

Full-waveform earthquake source inversion using simulation-based inference

Bayesian joint inversion of seismic and electromagnetic data for reservoir litho-fluid facies including geophysical and petrophysical rock properties

Coupling geologically consistent geostatistical history matching with parameter uncertainty quantification

Enhancement of Bayesian Seismic Inversion Using Machine Learning and Sparse Spike Wavelet: Case Study Norne Field Dataset

Stochastic full waveform inversion with deep generative prior for uncertainty quantification

Local uncertainty quantification for 3D time-domain full waveform inversion with ensemble Kalman filters: application to a North sea OBC dataset

Applying Spatial Bootstrap and Bayesian Update in uncertainty assessment at oil reservoir appraisal stages

Annealed Stein Variational Gradient Descent for Improved Uncertainty Estimation in Full-Waveform Inversion

Stochastic multi-fidelity joint hydrogeophysical inversion of consistent geological models

Bayesian linearized amplitude variation with offset and azimuth inversion and uncertainty analysis in horizontal transversely isotropic media

A Workflow to Integrate Numerical Simulation, Machine Learning Regression and Bayesian Inversion for Induced Seismicity Study: Principles and a Case Study

Joint Elastic and Petrophysical Inversion Using Prestack Seismic and Well Log Data

Nonstationary Prestack Linear Bayesian Stochastic Inversion