Automatic extraction of potential impact structures from geospatial data : examples from Finnmark, Northern Norway

Abstract

Impact cratering is a fundamental process in the Solar System, and on solid planetary bodies like Mars and the Moon, impact cratering may be the most prominent landforming process. On the Earth several processes compete in shaping the surface. Consequently, the impact structures on Earth are often poorly preserved, difficult to spot and found in limited numbers (per 2010, 176). The impact crater formation process results in a circular shape of fresh craters, except for impacts at low angles. This circularity is found in e.g. morphology, the distribution of impact rocks and in geophysical anomalies. The analytical choice is then to use the circular shape as a feature descriptor in search approaches. This thesis describes techniques applied to automatic extract circular features from appropriate geospatial datasets, i.e. to locate potential impact structures. The data cover parts of Finnmark county, Northern Norway, and include digital elevation models, geophysical potential field data and multispectral images. Remote sensing or image analysis methodologies can only detect potential impact structures, the most promising structures for further field studies. Evidence must later come from sampled rocks. An impact structure search should not be based on a single technique or a single dataset because of the diverse impact crater catalog, but rather a combination of several techniques applied on various data. Unlike previous terrestrial search approaches of purely visual analysis of data or the use of automatic techniques relevant to only a limited set of data, the presented methodology offers a framework to search large regions and several types of data to extract promising structures prior to the visual inspection.