Aarhus University Seal / Aarhus Universitets segl

Master's thesis defence - Peter Kongstad

Predicting chemical compositions in Greenland with remote sensing data application and machine learning.

2019.06.07 | Krishna Maria Olsen

Date Mon 24 Jun
Time 13:00 15:00
Location Geoscience, Auditorium, 1671 - 137
  • Supervisor/examiner: 
    Christoffer Karoff, Geoscience, AU
  • External examiner:
    Thorkild Maack Rasmussen, Luleå University of Technology, Sweden 

Abstract:

This thesis investigates the potential of applying machine learning techniques in combination with widely available and large scale remote sensing data, in order to predict chemical composition of regional geology.

The project revolve around a geochemical data set provided by the mineral exploration company 21st North. This data set is the product of years of geological prospecting conducted in a part of south eastern Greenland, known as Ammassalik.

The continuous surveying and prospecting have produced data sets with chemical compositions and affiliated GPS coordinates which in combination with satellites, aeromagnetic measurements and structural geology, have been analyzed by machine learning algorithms.

Department of Geoscience, Masters degree exam