GSA Annual Meeting in Seattle, Washington, USA - 2017

Paper No. 52-11
Presentation Time: 4:15 PM


VOORHEIS, Brennan M. and FARTHING, Dori J., Department of Geological Sciences, SUNY-Geneseo, 1 College Circle, Geneseo, NY 14454,

Color is often used to describe geological samples such as rocks and soils. In many cases, color is an indicator of chemical variation among samples of the same general type or region. Tools such as the Munsell system are commonly used to help identify a color. The Munsell system is robust and well-accepted, but has limitations when the color of interest is not present in the standard geologic or soil color books or other Munsell catalogs (e.g., plant tissues or bead colors). It is possible to customize alternative Munsell charts, but for samples such as historical iron slag that occurs in a wide range of colors, such an option is cost prohibitive.

Inspired by attempts to characterize the color of archaeological slag from the Adirondack region of NY, we investigated the potential of using reflectance spectrometry (RS). RS analyzes how a sample reflects and absorbs visible light. Each analysis produces a spectrum that is studied to identify the location of the strongest reflectance peak, the intensity of this peak, and the relative difference in intensity of this peak compared to the remainder of the spectrum. All of these values correspond to hue, lightness/darkness, and saturation.

The wavelength of the highest peak in a spectrum identifies the color's hue. The lightness or darkness of the color correlates with the actual intensity value for the highest peak because lighter samples will be more reflective along the entire spectrum. A color’s saturation corresponds to the calculated standard deviation for the spectrum’s intensity values between wavelengths of 400 and 700 nm. A highly saturated sample would have a large standard deviation, as there would be a bigger difference between the tallest peak’s intensity and the intensity of other wavelengths along the spectrum. Both Munsell and commercial paint chips were tested to investigate the sensitivity and reproducibility of RS. The best results came from samples with high-wavelength hues that had minimal to no interference from ultraviolet noise at the low end of the spectrum. Our study also found that RS was successful in defining a color even if it was “contaminated” (such as a blue sample with reddish-brown oxidation spots). The results of our work have helped generate a spectral catalog for common Munsell colors as well as additional colors found in archaeological slag samples.