Simulation of the Permian Source-to-Sink System in the Junggar Basin

To overcome the long-standing limitations of source-to-sink (S2S) studies of the Permian in the Junggar Basin—namely an overemphasis on static characterization and a lack of constraints from numerical sedimentary modeling—this study aims to develop an integrated, basin–mountain coupled forward-modeling workflow for the S2S system of the Lower Permian Wuerhe Formation. The goal is to achieve a dynamic, quantitative reconstruction of source-area surface processes, sediment supply, and basin depositional responses, and to predict sandbody distribution. The research includes: (1) within a unified spatial framework, characterizing accommodation-space evolution controlled by source-area tectonic evolution, rainfall and erosion-driven sediment supply, as well as depositional-area subsidence and lake-level variations; (2) deriving key surface-process and paleogeomorphic parameters, including paleoflow directions, time-varying runoff and sediment fluxes, and background geomorphic attributes (paleoslope, paleo-elevation, and paleowater depth); and (3) simulating sediment transport and deposition within the lacustrine basin to establish spatiotemporal evolution of geologically interpretable products—lithology, water depth, facies belts, sandbody distribution, depositional thickness, and stratigraphic architecture and sequence-filling styles—and constraining these results with geological observations.

Methodologically, we first prescribe initial topography and uplift rates in the source area, the spatiotemporal distribution of rainfall intensity, erosion rates of the source rocks, and a lake-level curve, while assigning a basement subsidence rate in the depositional area to jointly constrain the temporal evolution of accommodation space. We then run Badlands to obtain key outputs from topographic evolution and drainage/flow-routing calculations, and use these outputs as boundary conditions for Sedsim to perform depositional forward modeling and generate sedimentary results directly comparable to geological interpretation. Finally, the forward-model outputs are calibrated against well, seismic, and outcrop data; sensitivity analyses and iterative updates are conducted for critical parameters (uplift, erosion, rainfall, lake level, and subsidence) to obtain an optimal parameter set that is both process-consistent and consistent with observations.

The results indicate that the coupled Badlands–Sedsim forward-modeling workflow effectively transfers quantitative signals of source-area surface processes into basin-scale depositional responses, enabling a shift in S2S studies from “static description” to “process-based dynamic constraint.” Through data calibration and sensitivity-driven iteration, the workflow significantly improves the geological consistency and interpretability of the simulation results, providing a reproducible quantitative approach for understanding sedimentary evolution, sequence-filling mechanisms, and predicting favorable sandbody fairways in the Permian Junggar Basin, particularly for the Lower Permian Wuerhe Formation.