The Earth's surface forms the interface between the solid Earth, atmosphere, and hydrosphere, serving not only as the fundamental environment for human activities but also as a key zone where tectonic processes, climate change, and surface processes interact. Surface processes, including erosion, sedimentation, and sediment transport, not only shape topography but also influence crustal deformation and lithospheric evolution through mass redistribution, topographic loading and unloading, and thermal structure modifications. Consequently, the coupling between surface processes and deep Earth dynamics has become a major frontier in Earth system science. This review provides a systematic overview of surface process-lithosphere coupling from the perspective of quantitative numerical modeling. We first summarize the governing equations and numerical implementations of major surface processes, including fluvial incision, hillslope diffusion, sediment transport, and isostatic adjustment, and compare geomorphic models with different levels of complexity in terms of process representation, spatial scale, and computational efficiency. We then synthesize the principal feedback mechanisms through which surface processes influence the thermo-mechanical evolution of the lithosphere, including erosional unloading, sedimentary loading, stress redistribution, thermal blanketing, and sediment property variations. These mechanisms demonstrate that surface processes are not merely passive responses to tectonic forcing but can actively modify deep geodynamic processes through multiple feedback pathways. Furthermore, we review the development of coupled tectonic-surface process models and classify existing studies into four categories: landscape evolution models, geodynamic models, one-way coupled models, and two-way coupled models. By comparing their representations of feedback mechanisms, interactions between surface and deep Earth processes, and numerical coupling strategies, we highlight the evolution of the field from unidirectional forcing frameworks toward fully coupled two-way feedback systems. Recent developments indicate a transition from simplified two-dimensional approaches to three-dimensional, multi-process, and Earth system coupling. Despite substantial progress, current models remain largely based on fluvial-dominated landscape evolution frameworks. The integrated representation of sedimentary processes, glacial dynamics, weathering, impact-related processes, and their interactions remains limited, and a unified physical framework capable of consistently describing multiple feedback mechanisms is still lacking. These limitations constrain the application of coupled models to long-term Earth system evolution and to planetary environments such as Mars and other terrestrial planets. Future advances will require tighter coupling between surface processes and lithospheric dynamics within unified numerical frameworks, together with stronger constraints from multi-source observations and planetary analog studies, to improve model robustness, testability, and predictive capability. Such efforts will provide a quantitative framework into the co-evolution of surface and deep Earth systems on both Earth and planetary scales.
The Earth's tectonic evolution involves immense spatial and temporal scales, posing fundamental challenges for direct observation of its dynamic mechanisms. Analog modeling of tectonic processes provides a key approach for quantitative analysis in the laboratory by constructing scaled models that satisfy similarity principles. However, conventional normal-gravity analog modeling is limited by insufficient self-weight stress of the model and scale effects of material strength, making it difficult to faithfully reproduce the stress state and long-term evolution of deep geological environments. Hypergravity tectonic analog modeling has emerged as a solution to these limitations. By using a centrifuge to generate a high-gravity field, this method achieves stress equivalence between the scaled model and the natural prototype, while the "time-scaling effect" significantly accelerates slow geological flow processes. This paper systematically reviews the development and application of hypergravity analog modeling. It first outlines the similarity theory and the inherent limitations of normal-gravity experiments. It then focuses on the core principles of hypergravity experiments: by increasing the body force field, it becomes possible to use materials with higher strength and a wider rheological spectrum, thereby constructing more realistic models and enabling non-destructive, successive observations of progressive deformation. The article reviews in detail the application of hypergravity technology in key tectonic fields, including plate subduction initiation, fold-and-thrust belt formation, regional extensional tectonics, and evaporite/magmatic diapirism. It highlights unique insights gained from hypergravity simulations that are difficult to obtain from normal-gravity experiments, such as the rotation mechanism of low-angle subduction, the evolution of detachment faults, the driving mechanisms of salt tectonics, and the behavior of deep ductile flow. Compared with normal-gravity simulations, the core advantages of hypergravity modeling are: accurate matching of the stress field, effective compression of experimental time, compatibility with materials having widely varying rheological properties, and detailed quantitative characterization of deformation processes. Although hypergravity simulations have achieved substantial results using drum centrifuges, the potential of large-scale models in beam centrifuges remains to be further explored. Looking forward, the deep integration of hypergravity technology with modern observation techniques such as particle image velocimetry (PIV), digital image correlation (DIC), and X-ray computed tomography (CT) will greatly enhance the spatial resolution and quantitative analysis dimensions of experimental data. Hypergravity tectonic analog modeling will continue to provide irreplaceable experimental constraints on major frontier geological problems, including plate interaction, three-dimensional evolution of orogenic belts, and deep material cycling, thereby driving innovation in geodynamic theory.
Artificial intelligence (AI) technologies have been extensively applied across a diverse range of disciplinary fields. In the realm of Earth sciences research, traditional data processing methods are frequently confronted with formidable challenges due to the intricate physical processes and massive datasets inherent to the field, making AI an indispensable tool for elevating research efficiency and predictive accuracy. With the rapid advancement of deep learning and neural network technologies, AI methodologies are capable of providing feasible solutions for scenarios featuring data scarcity or high-dimensional complex systems, which remarkably enhances the predictive accuracy of Earth sciences models and accelerates computational processes. In subfields including seismology, geodynamics, climate modeling and mineral resource exploration, AI technologies enable scientists to extract valuable information from complex datasets and generate more precise predictions in a significantly shortened timeframe through the advanced processing of large-scale geological, geophysical and environmental data. Notably, the proposition of physics-informed neural networks (PINN) has further propelled the application of AI in Earth sciences: by integrating data-driven approaches with fundamental physical laws, PINNs offer a novel paradigm for addressing complex problems that cannot be efficiently tackled by traditional simulation methods. Although AI technologies have yielded promising preliminary results in Earth sciences applications, numerous challenges remain to be addressed, such as high-dimensional data processing, AI model interpretability, and the modeling of multi-physics coupling problems. This paper provides a comprehensive review of the applications of AI—especially deep learning and PINN—in Earth sciences, discusses their respective advantages and existing challenges across various application domains, and prospects the future research directions in this field.
Topography and gravity anomalies are two of the most fundamental observations of a planetary surface. Their magnitude and pattern reflect both the shallow lithospheric flexural isostasy and the deep mantle dynamics. Mantle dynamics primarily influence medium- to long-wavelength topography and gravity anomalies, with mantle viscosity being one of the most critical parameters controlling the dynamic behavior of a planet. The joint analysis of gravity and topography data has become an important method for studying the internal viscosity structure of planets. On Earth, the analysis of seismic waves can provide detailed internal structures, which are used as driving forces for mantle convection in geodynamic simulations. These simulations have successfully explained geoid anomalies and provide important constraints on mantle viscosity. The topography and gravity anomalies from mantle convection models can be combined with other data, such as plate motion, post-glacial rebound, and mineral physics experiments, to jointly constrain the mantle's viscosity structure. The method of using topography and gravity anomalies for dynamic modeling to invert mantle viscosity structure can also be applied to Venus. The main difference from Earth is that Venus lacks seismic data on internal density structures. Instead, numerical simulations of thermal evolution can be used to generate the present-day mantle structure, thereby constraining Venus's internal viscosity structure. The primary difference in mantle viscosity structure between Earth and Venus is the absence of a pronounced viscosity jump between the upper and lower mantle on Venus and the likely existence of an asthenosphere on Earth. The absence of an asthenosphere in Venus's mantle is the reason why Venus lacks plate tectonics. With the improvement in the precision of observational data and the advancement of research methods, planetary topography and gravity have become increasingly significant in the study of mantle viscosity.
