This thesis examines how quantitative colour measurements and spectral recordings can be utilised to add traceability to the process of stratigraphic analysis carried out for archae- ological and geoscientific research. The non-destructive, quantitative and often extensive character of this data holds much potential for stratigraphic documentation and interpreta- tion. Quantitative methods are examined, which complement traditional perception-based documentation and interpretation of stratigraphies without the need of complex laboratory work. Likewise, it is shown how the results of such laboratory work can be further utilised by combining them with spectral imaging techniques. Digital photography, as carried out as a standard procedure during fieldwork, constitutes a basic form of spectral imaging. Throughout this thesis, digital photographs systematically recorded with a common RGB camera are therefore treated as physical measurements. By analysing RGB, multispectral and hyperspectral image data of soil and sediment profiles, it is shown, that digital image data offers a reliable basis for the transparent and reproducible delineation of stratigraphic layers. The produced stratigraphic sequences based on spectral data are of similar quality as sequences based on traditional sedimentological data obtained from laboratory work. Spectral data, however, can be acquired during fieldwork and offers a fast and non-destructive alternative to exhaustive laboratory analyses. Furthermore, it is shown that the extensive character of spectral image data can be utilised to produce maps of chemical and physical parameters along soil and sediment profiles. Through the application of statistical models which involve the spectral image data and sed- iment properties obtained from samples analysed in the laboratory, reliable maps are created for various parameters like soil organic matter or iron oxide contents. The presented results show that RGB images as well as hyperspectral images can be utilised for this purpose. Exploring possible applications of spectral imaging, the results presented in this thesis promote a more transparent and reproducible documentation for often destructive archaeo- logical and geoscientific fieldwork.
