This code was initially created for my master's thesis: Gadylyaev, D. (2021): Applying Computed Tomography (CT) scanning for segmentation of permafrost constituents in drill cores , Master thesis, University of Potsdam, Institute of Geosciences. https://epic.awi.de/id/eprint/55281/
and after which it turned into a scientific paper: Nitzbon, J., Gadylyaev, D., Schlüter, S., Köhne, J. M., Grosse, G., and Boike, J.: Brief communication: Unravelling the composition and microstructure of a permafrost core using X-ray computed tomography, The Cryosphere, 16, 3507–3515, https://doi.org/10.5194/tc-16-3507-2022, 2022.
This pyhton script uses the below listed input files to perform and evaluate a regression analysis of the CT data against the laboratory data. The regression result is the composition of the CT-derived sediment phases (A,B) in terms of pore ice, organic, and mineral. The script furthermore computes evaluation metrics of the lab-CT comparison, and computes volumetric contents of pore ice, total ice, organic, and mineral at the high resolution of the original CT data.
Contains the volumetric contents of total ice, organic, and mineral measured in the laboratory at AWI Potsdam at a coarse resolution. It further contains the volumetric contents of gas, excess ice, and two sediment phases (A,B) derived from a CT scan at UFZ Halle, downsampled to the resolution of the laboratory samples.
Contains the volumetric contents of gas, excess ice, and two sediment phases (A,B) derived from a CT scan at UFZ Halle at the original resolution of 50µm.
This file can be reproduced by the files listed above and contains, in addition to the data contained in volumetric_contents_sampleRes_lab+CT.csv, the volumetric contents of pore ice, total ice, mineral, and organic as predicted by the regression model at the same (coarse) resolution as the laboratory samples.
This file can be reproduced by the files listed above and contains, in addition to the data contained in volumetric_contents_highRes_CT.csv, the volumetric contents of pore ice, total ice, mineral, and organic as predicted by the regression model at the same (high) resolution as the original CT data.