[A-TT29] Machine Learning Techniques in Weather, Climate, Ocean, Hydrology and Disease Predictions

[ATT29-P03] Automatic optical sediment facies classification – A novel machine learning application to sedimentology

*Aaron Kuan Lun Chen¹, An-Sheng Lee¹, Wen-Ta Young¹, Sofia Ya Hsuan Liou¹ (1.National Taiwan University)

Keywords:machine learning, sediment facies

Geosciences are overall based on the delineation of sediment facies but are torn between the quality and quantity. The conventional method of facies discrimination relies on observer-dependent sedimentological descriptions, which are not always reanalyzable. Machine learning as a trend of the technology shows a great possibility to build up a novel automatic sediment facies classification for the further geoscience researches with a consistent quality. This approach was developed with the high-resolution photos comprise of sediment cores covering sundry sediment facies which are typical for the island Norderney (East Frisian Wadden Sea, Germany). We have the sediment facies recognized by conventional facies anlasis: moraine, eolian/fluvial, soil, peat, lagoon, sand flat, channel fill and beach-foreshore. The main scope is using the high-resolution photos with machine learning to imitate the observational behavior of sedimentologists, an optical machine-learning classification model. The Ultimate goal is to establish a training model to fit cores in other areas and exploit the optical machine learning limit. We also see the model as the key for regional geological mapping. Lastly, it is expected to be a cornerstone for searching the unprecedented depth.