A Planetary Inverse Modeling Framework for Sedimentary Basin Evolution

Thesis Abstract

Advances in planetary geology and sedimentology increasingly demand robust frameworks to interpret basin evolution across diverse planetary contexts. This study addresses the challenge of disentangling the multifactorial controls on sedimentary basin development by proposing a Planetary Inverse Modeling Framework (PIMF) that integrates stratigraphic, structural, climatic, and geochemical data within a unified probabilistic inversion approach. The aim is to reconstruct time–variant boundary conditions and interior processes that shape basin architecture on terrestrial planets, with explicit attention to uncertainty propagation and model identifiability. Specific objectives are (i) to formulate a modular inversion architecture that couples stratigraphic forward models with stress–strain, climate forcing, and diagenetic alteration modules; (ii) to implement a Bayesian hierarchical inference scheme that estimates latent state variables such as paleo-subsidence rates, paleotemperature, sediment supply, and diagenetic alteration indices; (iii) to calibrate the framework against well-documented terrestrial analogues (e.g., the Western Interior Seaway, North Sea basins) and extendable to Mars and Venus by incorporating planet-specific priors; (iv) to perform sensitivity and identifiability analyses to determine parameter regimes where basin signals are robust against data sparsity; and (v) to provide a decision-support tool for prospecting and mission planning in planetary exploration. The methodology adopts a mixed-methods, theory-driven research design grounded in geostatistical inversion, Bayesian data assimilation, and numerical basin modeling. The population comprises published and archived stratigraphic, structural, and geochemical datasets from Earth analogue basins and simulated datasets representing Martian and lunar basins. A stratified sampling approach is used to select 12 terrestrial analog basins with high-resolution sequence stratigraphy and well-constrained tectono-climatic histories, supplemented by 5 synthetic planetary scenarios to test transferability. Data collection instruments include high-resolution well logs, core samples, seismic cross-sections, regional thermochronology compilations, remote-sensing basemap products, and climate proxy records. Analytical techniques encompass (i) Bayesian Markov Chain Monte Carlo (MCMC) for posterior parameter estimation, (ii) Gaussian process emulators to accelerate computationally intensive forward models, (iii) regression analysis to identify covariate relationships between subsidence, sediment supply, and climate indices, (iv) entropy-based model selection to quantify information gain, and (v) sensitivity analysis using Sobol indices to assess parameter identifiability. The framework integrates a probabilistic forward model of basin evolution with a modular inversion layer that accommodates planet-specific priors derived from crater counts, atmospheric modeling, and plausible lithofacies assemblies. Model specification comprises a hierarchical structure a global prior that encodes planetary context (gravity, crustal thickness, tectonic regime) and local priors drawn from Earth analogues; a forward model linking boundary conditions (subsidence, accommodation space, sediment supply, diagenesis) to observable stratigraphic geometries and geochemical signatures; and a likelihood function combining seismic and outcrop observations, thermochronology dates, and remote-sensing derived facies maps. Ethical considerations center on data provenance and credit for analogue datasets. Expected findings include (i) quantified relationships between subsidence evolution and climate-driven sediment supply across analogue basins, (ii) identifiability maps delineating parameter regimes where inferences remain robust with sparse data, and (iii) a transferable set of priors enabling planetary basin inference with limited in-situ measurements. The study is anticipated to demonstrate that the PIMF can recover plausible paleoenvironmental histories and basin geometries from incomplete datasets, while providing calibrated uncertainties that inform mission targeting and sample selection. The contribution to knowledge lies in delivering a formalized, transferable inverse modeling framework capable of integrating multi-disciplinary data to reconstruct planet-wide sedimentary basin evolution, bridging Earth-based theory with planetary exploration needs. The study concludes with recommendations to refine modular components (e.g., diagenetic models for porous rocks on non-Earth planets), enhance data assimilation efficiency, and expand analogue databases to improve prior specifications for future planetary missions.