Title: Integrated Seismic and Magnetotelluric Inversion for High-Resolution Subsurface Reservoir Characterization in Geophysics

Project Abstract

Integrated Seismic and Magnetotelluric Inversion for High-Resolution Subsurface Reservoir Characterization in Geophysics presents a multidisciplinary framework that leverages the complementary sensitivities of seismic and magnetotelluric (MT) methods to refine subsurface reservoir models with unprecedented resolution and reliability. The study develops a joint inversion workflow that merges high-frequency seismic reflections, which constrain acoustic impedance and stratigraphic layering, with MT responses that are sensitive to electrical resistivity and pore-fluid content, thereby enhancing discrimination between lithologies, fluid types, and fracture networks. We implement a scalable, data-assimilation-driven inversion engine that propagates uncertainty through each stage, enabling probabilistic assessment of reservoir properties such as porosity, permeability, fluid saturation, and fracture density. The methodology integrates advanced preconditioning, multi-physics forward modeling, and cross-modal regularization to reconcile discrepancies between seismic and MT datasets arising from anisotropy, heterogeneity, and non-uniqueness in inversion. Key methodological advances include (i) simultaneous velocity–fracture–resistivity estimation using a multi-parameter Bayesian framework; (ii) stochastic optimization techniques that balance data misfit against geophysically plausible priors to avoid overfitting in areas of sparse coverage; (iii) joint rock-physics constraints that relate acoustic impedance, electrical conductivity, and formation damage indicators to rock properties; (iv) incorporation of time-lapse (4D) data to monitor dynamic reservoir processes such as pressure buildup, water flooding, or hydrocarbon migration; (v) robust treatment of MT impedance data with sensitivity-weighted misfit to mitigate near-surface and galvanic distortion effects; and (vi) a scalable computational architecture leveraging high-performance computing and model-order reduction to enable practical deployment on field-scale problems. The project demonstrates the approach on synthetic benchmarks and field datasets from mature and developing reservoirs, demonstrating improvements in delineating thin interbeds, inversion of fractured zones, and resolving fluid boundaries that are ambiguously defined by seismic alone. Quantitative metrics include reduced uncertainty in porosity and permeability estimates, enhanced delineation of permeability-contrast channels, and more accurate prediction of producible reserves when integrated with well-log and production data. The research also addresses practical considerations such as data quality weighting, survey design optimization for joint surveys, and computational costs, providing guidelines for practitioners to implement integrated seismic-MT inversion in exploration and production workflows. The anticipated outcome is a robust, transferable methodology that increases confidence in subsurface reservoir characterization, enabling more informed decision-making for field development, enhanced oil recovery planning, and hydrogeological risk assessment.

Project Overview

What This Project Is About

---

This project combines two geophysical methods, seismic and magnetotelluric (MT), to create a clearer picture of what lies beneath the ground. Seismic methods use sound-like waves to map rock structures, while MT measures natural electric and magnetic fields to infer how rocks conduct electricity. By using both, students learn how to better identify features like oil and gas reservoirs, water, and rock types in a single, high-resolution model.

---

---

The Problem It Addresses

---

Relying on one method can miss important details or give ambiguous results. Seismic data is great for shapes and boundaries but not always for rock properties, while MT helps with electrical properties but has lower resolution. The project tackles how to integrate these datasets to reduce uncertainty and improve decision-making in resource exploration and groundwater studies.

---

---

Objectives of the Project

---

1. Learn the basics of how seismic and MT data are collected and processed.
2. Understand how to combine different data types into a unified subsurface model.
3. Develop a simple workflow to estimate rock properties and reservoir characteristics more reliably.
4. Evaluate the benefits and limitations of integrated inversion approaches.

---

---

What You Will Do Step by Step

---

1. Review foundational concepts in seismic and MT methods.
2. Obtain or simulate a small dataset for both seismic and MT measurements.
3. Preprocess data to remove noise and align measurement scales.
4. Build a basic joint inversion model combining seismic and MT information.
5. Test the model against known benchmarks or simple synthetic scenarios.
6. Interpret the resulting subsurface properties in terms of potential reservoirs or fluids.
7. Discuss uncertainties and potential improvements for real-world use.

---

---

---

---

Expected Outcome

A clear, integrated subsurface model that highlights key features such as boundaries and rock properties, with an understandable assessment of confidence and practical implications for exploration or water resources, ready to guide further study or fieldwork.