抄録
Research on landform evolution has undergone fundamental transformations in theoretical frameworks and methodologies in recent years. The integration of multi-source data, such as high-resolution remote sensing, LiDAR, and geophysical surveys, with machine learning algorithms has significantly enhanced the analysis of geomorphic processes, driving the rapid development of data-driven studies. Investigations into rapid landscape changes induced by extreme events (e.g., catastrophic floods, large-scale landslides) have challenged the traditional paradigm of “uniformitarianism,” highlighting the critical role of abrupt processes in shaping surface patterns and promoting the establishment of a “catastrophism-uniformitarianism” coupling theory. Multi-process coupling studies have revealed the nonlinear interaction mechanisms among hydrological, gravitational, glacial, aeolian, and biogeomorphic forces under varying conditions. The application of quantitative techniques such as cosmogenic nuclide dating and optically stimulated luminescence dating has shifted landform evolution analysis from qualitative inference to high-precision quantification, greatly improving the resolution of long-term evolutionary histories. This paper summarizes frontier advances and challenges, explores future research pathways integrating data and physical models, and provides theoretical and methodological support for understanding surface system evolution and its responses to climate change.
