Search papers, labs, and topics across Lattice.
This study systematically compares data assimilation (DA) algorithms for subsurface flow using latent diffusion models (LDMs) to address the challenges of parameter calibration in the presence of geological uncertainty. The research reveals a critical trade-off between model-space updates, which significantly reduce uncertainty but yield geologically unrealistic models, and latent-space updates, which preserve geological realism but offer limited uncertainty reduction. By employing rigorous Markov chain Monte Carlo (MCMC) and Sequential Monte Carlo (SMC) techniques within the LDM latent space, the authors demonstrate that these methods outperform traditional ensemble Kalman approaches in terms of data mismatch and uncertainty reduction while maintaining geological plausibility.
MCMC and SMC techniques in latent space achieve lower data mismatch and better uncertainty reduction than traditional ensemble Kalman methods, all while preserving geological realism.
Data assimilation (DA) in subsurface flow entails calibrating model parameters to match observed data, typically at wells, while preserving geological realism. Latent diffusion models (LDMs) provide efficient mappings from high-dimensional geological model space to a low-dimensional latent variable, reducing the dimensionality of the inverse problem while maintaining plausibility in posterior geomodels. However, the high nonlinearity in the LDM mapping may degrade the performance of Kalman-gain-based ensemble updates. We present a systematic comparison of DA algorithms applied to large-scale 3D channelized geomodels with hierarchical geological uncertainty. We compare model-space and latent-space DA using the ensemble smoother with multiple data assimilation (ESMDA), and demonstrate a key trade-off: model-space updates achieve significant uncertainty reduction but produce geologically unrealistic posterior models, while latent-space updates preserve realism but exhibit limited uncertainty reduction. Motivated by this, we explore rigorous Markov chain Monte Carlo (MCMC) and Sequential Monte Carlo (SMC) algorithms in the 3D-LDM latent space. To accommodate their high computational demands, we develop a fast surrogate flow model that approximates well-rate responses. MCMC and SMC are evaluated against ESMDA across three synthetic test cases, with DA performed in the LDM latent space. All models maintain geological realism due to the LDM parameterization. MCMC and SMC are consistent with one another and achieve lower data mismatch and more uncertainty reduction than latent-space ESMDA. Our overall results demonstrate that ensemble Kalman methods may provide overestimated posterior uncertainty with highly nonlinear parameterizations, while rigorous Monte Carlo sampling, enabled by fast surrogate models, can provide a more reliable alternative.
Stanford HAI