Answers are grounded in this paper and cite their sources.
Seawater intrusion (SWI) remains a major challenge in the management of coastal aquifers. Although high-fidelity variably-density flow and transport models can accurately represent SWI processes, their computational cost limits their use in multi-scenario optimization, uncertainty quantification, and risk assessment. To address this limitation, surrogate modeling approaches have increasingly been adopted, aiming to approximate the input-output relationships of complex SWI models at substantially lower cost. This paper reviews recent advances in surrogate models for SWI simulation and management and proposes a classification framework based on the degree of physics integration and the adopted dimensionality-reduction strategy. Existing approaches are categorized into five groups: data-driven models, physics-integrated models, multi-fidelity models, reduced-order models, and neural operators. Furthermore, the coupling strategies between surrogate models and pumping-optimization frameworks (e.g., genetic algorithms, particle swarm optimization, Bayesian optimization) are examined, together with their roles in supporting multi-objective pumping management and uncertainty analysis.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: