Abstract:This study investigates the integration of machine learning (ML) and data assimilation (DA) techniques, focusing on implementing surrogate models for Geological Carbon Storage (GCS) projects while maintaining high fidelity physical results in posterior states. Initially, we evaluate the surrogate modeling capability of two distinct machine learning models, Fourier Neural Operators (FNOs) and Transformer UNet (T-UNet), in the context of CO$_2$ injection simulations within channelized reservoirs. We introduce the Surrogate-based hybrid ESMDA (SH-ESMDA), an adaptation of the traditional Ensemble Smoother with Multiple Data Assimilation (ESMDA). This method uses FNOs and T-UNet as surrogate models and has the potential to make the standard ESMDA process at least 50% faster or more, depending on the number of assimilation steps. Additionally, we introduce Surrogate-based Hybrid RML (SH-RML), a variational data assimilation approach that relies on the randomized maximum likelihood (RML) where both the FNO and the T-UNet enable the computation of gradients for the optimization of the objective function, and a high-fidelity model is employed for the computation of the posterior states. Our comparative analyses show that SH-RML offers better uncertainty quantification compared to conventional ESMDA for the case study.

What problem does this paper attempt to address?

### What problems does this paper attempt to solve? This paper aims to solve the problems of data assimilation (DA) and uncertainty quantification (UQ) in Geological Carbon Storage (GCS) projects. Specifically, the paper focuses on the following key issues: 1. **Improving computational efficiency**: - Traditional DA methods such as Ensemble Smoother with Multiple Data Assimilation (ESMDA) are computationally expensive when dealing with complex geological structures. The paper introduces machine - learning - based surrogate models, such as Fourier Neural Operators (FNOs) and Transformer UNet (T - UNet), to accelerate the DA process. For example, the Surrogate - based hybrid ESMDA (SH - ESMDA) method can increase the speed of the standard ESMDA by at least 50%. 2. **Enhancing prediction accuracy and uncertainty quantification**: - The paper proposes two hybrid methods: SH - ESMDA and Surrogate - based Hybrid RML (SH - RML). Among them, SH - RML calculates the posterior state by using the high - fidelity physical simulator DARTS and uses ML surrogate models for gradient calculation, thereby achieving better history matching and uncertainty quantification in the variational framework. Experimental results show that SH - RML is superior to the traditional ESMDA method in terms of uncertainty quantification. 3. **Dealing with complex geological structures and nonlinear problems**: - GCS projects usually involve complex geological structures, such as channelized reservoirs. These structures lead to highly nonlinear and non - Gaussian distribution parameter problems, which pose challenges to traditional DA methods. By combining ML surrogate models and high - fidelity physical simulators, the paper can more accurately capture these complexity and nonlinear characteristics. 4. **Reducing the dependence on a large amount of high - fidelity simulation data**: - Training ML surrogate models usually requires a large amount of high - fidelity simulation data, which may be difficult to obtain in actual GCS projects. By introducing new ML architectures such as FNOs and T - UNet, the paper attempts to improve the performance of surrogate models under limited data conditions, especially on small data sets. ### Summary The main goal of this paper is to develop efficient surrogate models and hybrid methods by combining machine learning and data assimilation techniques, in order to improve the computational efficiency, prediction accuracy and uncertainty quantification ability of GCS projects, while dealing with the challenges brought by complex geological structures.

AI enhanced data assimilation and uncertainty quantification applied to Geological Carbon Storage

Likelihood-Free Inference and Hierarchical Data Assimilation for Geological Carbon Storage

A deep learning-accelerated data assimilation and forecasting workflow for commercial-scale geologic carbon storage

Uncertainty Quantification Based on Deep-Learning Approach Integrating Time-Lapse Seismic Data for Geological Carbon Storage

Surrogate Model for Geological CO2 Storage and Its Use in Hierarchical MCMC History Matching

Efficient machine-learning surrogates for large-scale geological carbon and energy storage

An uncertainty-aware Digital Shadow for underground multimodal CO2 storage monitoring

Accelerated training of deep learning surrogate models for surface displacement and flow, with application to MCMC-based history matching of CO2 storage operations

A deep-learning-based surrogate model for data assimilation in dynamic subsurface flow problems

Using Deep Learning to Improve Ensemble Smoother: Applications to Subsurface Characterization

An adaptive Kriging surrogate method for efficient uncertainty quantification with an application to geological carbon sequestration modeling

A deep learning-based surrogate model for seismic data assimilation in fault activation modeling

Uncertainty Quantification in CO2 Trapping Mechanisms: A Case Study of PUNQ-S3 Reservoir Model Using Representative Geological Realizations and Unsupervised Machine Learning

Assimilation of Geophysics-Derived Spatial Data for Model Calibration in Geologic CO2 Sequestration

A Digital Twin for Geological Carbon Storage with Controlled Injectivity

Surrogate-Assisted Evolutionary Optimization of CO 2 -ESGR and Storage

Towards efficient and accurate CO 2 sequestration optimization: Integrating hierarchical spatio-temporal information into deep-learning-based surrogate models

Machine Learning Application for CCUS Carbon Storage: Fracture Analysis and Mapping in the Illinois Basin

Rock Physics-Based Data Assimilation of Integrated Continuous Active-Source Seismic and Pressure Monitoring Data during Geological Carbon Storage

Deep Learning Framework for History Matching CO2 Storage with 4D Seismic and Monitoring Well Data

CO2 Leakage Rate Forecasting Using Optimized Deep Learning