P085-08
Creating Mineral “Frankenspectra” Using UV-VNIR-MIR Reflectance Data from Three Different Laboratories

Wednesday, 16 December 2020: 10:50
Virtual
Melissa D Lane1, Amanda R Hendrix2, Roger Nelson Clark2, Edward Cloutis3, Melinda Darby Dyar4, Jörn Helbert5, Alessandro Maturilli5, Neil Pearson6 and Toolbox for Research and Exploration (TREX) Team, (1)Fibernetics LLC, Lititz, PA, United States, (2)Planetary Science Institute Tucson, Tucson, AZ, United States, (3)University of Winnipeg, Department of Geography, Winnipeg, MB, Canada, (4)Mount Holyoke College, South Hadley, MA, United States, (5)German Aerospace Center DLR, Berlin, Germany, (6)Planetary Science Institute Tucson, Tucson, United States
Abstract:
Reflectance spectroscopy is a useful method for identifying an unknown mineral, if its spectrum can be compared to a spectral library that includes quality spectra of the same mineral. However, a mineral sample measured in different labs will often result in varied spectra due to the way the sample is placed into the sample cup, the calibration of the data, the spectrometer hardware and light sources used, or other issues. Our team measured a suite of 28 mineral samples in 3 different labs to expose differences in the resulting spectra in order to modify the data acquisition strategy both to improve and converge on the best sample spectra.

The labs involved were the Planetary Spectroscopy Lab at the German Aerospace Center, the Centre for Terrestrial and Planetary Exploration at the University of Winnipeg, and the Planetary Geosciences Lab at the Planetary Science Institute. After the best data were acquired from each of the labs and plotted together, it was clear that no single lab/instrument offered the best overall spectrum; rather, each lab/instrument best presented a portion of the overall wavelength region studied. Therefore, a spectral library would benefit from one best-quality, most representative spectrum of a sample, constructed using the best wavelength pieces of the spectra acquired at various labs.

We are calling these blended/spliced mineral spectra “Frankenspectra”, i.e., spectra built in a laboratory and stitched together with human modification, akin to Dr. Frankenstein’s monster (but in a good way).

We will present the details of our samples, the laboratories, our ambient-temperature data from the 3 labs, as well as our derived Frankenspectra of the minerals. These data will be archived at the Planetary Data System Geosciences Node (GEO), including all the original individual spectra and the best-representative Frankenspectrum of each mineral.

In the near future we will be acquiring similar measurements and creating similar spectra with meteorite samples that currently are being prepared for us at NASA Johnson Space Center’s curatorial facility.