A new approach to quantitatively describe permafrost core using multi-energy CT scanning: composition fraction and morphological analysis

Monday, 15 December 2014
Chun Chang1,2, Catherine McKnight1 and Timothy J Kneafsey1, (1)Lawrence Berkeley National Laboratory, Berkeley, CA, United States, (2)China University of Geosciences Beijing, Beijing, China
Composition discernment, fraction calculation and morphological analysis of a shallow core retrieved from Barrow, AK as part of the Next Generation Ecosystem Experiments in the Arctic (NGEE-Arctic) were conducted to give a quantitative description of the core. Imaging of the core was performed using a medical X-ray computed tomography (CT) scanner, which gives a 3D image with a resolution of 0.195×0.195×0.625mm3. The core consists mainly of mineral, ice, organic matter and air and composition discernment and fraction calculation focus on the first three materials. Four scans with different energies were performed because materials with different density show different responses on scans with varying energies. A calibration curve showing the relationship between density and CT value was built by scanning standard materials having a wide range of density. CT value of the three compositions under four energies was determined by the calibration curve and the core scan. Composition fraction was calculated on the assumption that the core CT value is linearly proportional to the composition fraction and by solving linear least-squares problems with bounds. Comparison of the estimated and measured core CT value shows that the correlation coefficient is more than 0.99, indicating the accuracy of the calculation. Two regions with relatively high fraction of organic matter (10%) were distinguished, which are located at the top of the core and ice filled fractures at the bottom of the active layer. Morphological analysis was applied to the mineral and ice because of low fraction of organic matter. Three segmentations corresponding to ice-rich (with a density of 0.86 to 1.24 g/cm3), transition from ice to mineral (1.24 to 1.47 g/cm3) and mineral-rich (1.47 to 2.65 g/cm3) were applied to the core, and two area (area and area standard deviation) and three morphological (circulatory, roundness and rectangularity) parameters were analysed. By conducting Principle Component Analysis and Cluster Analysis, two principle components were distinguished from the five parameters, giving a quantitative description of the core. The study provides an innovative approach to quantitatively describe the arctic core, which may provide help in understanding geomorphological changes as a result of climate variation in the arctic.