In this thesis, we investigated the repeatability and consistency of soil colour determination using two methods: visual assessment with Munsell colour charts and a digital method using the Munsell Capsure spectrocolorimeter. We included 28 soil samples in the study and assessed the colour of three samples twice to check consistency. The visual assessments were performed by five evaluators, while the digital measurements were repeated five times for each sample. All assessments and measurements were performed in a laboratory under the same lighting conditions to ensure comparability between the two methods. We observed differences between the evaluators as the average repeatability of the visual assessments was low (κF = 0.240), indicating the subjective nature of the visual method. The digital method showed a significantly higher repeatability (κF = 0.765), confirming its reliability. The visual assessments matched the digitally determined colour codes in only 20.7% of cases. The average degree of agreement between the evaluators' assessments and the values measured with the Capsure spectrocolorimeter was very low (κF = 0.173). The reference colour, which was calculated as a modal value from the Capsure measurements, agreed better with the assessments of some evaluators than with others.
|
