5:15 PM - 7:15 PM
[HTT15-P01] Automatic detection and classification of marine and fluvial terraces using statistical and stochastic clustering methods
★Invited Papers
Keywords:Marine terraces, Fluvial terraces, Relative sea-level changes, Digital elevation model, Geomorphometry, Automatic terrace classification
Marine and fluvial terraces play a significant role in various geoscience fields as records of past relative water-level changes caused by climate and tectonic activity. The identification of lateral continuity of synchronous terraces is one of the most important observations. However, the evaluation of terrace continuity often encounters difficulties due to erosion and weathering, and often relies on subjective judgment. While geochronological dating is the most effective tool for identifying synchronous terraces, finding suitable material for dating is not possible in some locations. These limitations have motivated a topography-based approach that enables objective and reliable terrace classifications.
To address this challenge, we developed an automated terrace identification method that finds and visualizes the lateral continuity of cliff features using digital elevation models (DEMs). This method employs Gaussian Mixture Models (GMM) and Akaike’s Information Criterion (AIC) to assess the detection probabilities of paleo-shoreline angles, enabling automatic mapping and visualization on 3D terrain models.
Our classification model demonstrates high versatility, applicable for a wide range of formation ages (Holocene to Pleistocene) and geomorphic settings (both marine and fluvial terraces). In this presentation, we showcase its application in diverse regions, including Holocene marine terraces in the Boso Peninsula, central Japan; Late Pleistocene marine terraces in the Huon Peninsula, Papua New Guinea; and Pleistocene fluvial terraces along the Waipawa River, New Zealand. The classification process is implemented in Python and published as open-source script on GitHub, promoting its broad application to terrace landforms worldwide.
To address this challenge, we developed an automated terrace identification method that finds and visualizes the lateral continuity of cliff features using digital elevation models (DEMs). This method employs Gaussian Mixture Models (GMM) and Akaike’s Information Criterion (AIC) to assess the detection probabilities of paleo-shoreline angles, enabling automatic mapping and visualization on 3D terrain models.
Our classification model demonstrates high versatility, applicable for a wide range of formation ages (Holocene to Pleistocene) and geomorphic settings (both marine and fluvial terraces). In this presentation, we showcase its application in diverse regions, including Holocene marine terraces in the Boso Peninsula, central Japan; Late Pleistocene marine terraces in the Huon Peninsula, Papua New Guinea; and Pleistocene fluvial terraces along the Waipawa River, New Zealand. The classification process is implemented in Python and published as open-source script on GitHub, promoting its broad application to terrace landforms worldwide.