Caltech/NHM Mineral VTIR Spectral Library

Grant, Kathleen; Brodrick, Philip G.; Celestian, Aaron J.; Rossman, George R.; West, A. Joshua

doi:10.5281/zenodo.15305345

Published June 18, 2025 | Version v1

Dataset Open

Caltech/NHM Mineral VTIR Spectral Library

1. University of Southern California
2. Jet Propulsion Laboratory
3. Natural History Museum of Los Angeles County
4. California Institute of Technology

Abstract

The Caltech/NHM Mineral VTIR Spectral Library was developed to address gaps in VTIR (visible to shortwave infrared and thermal infrared) spectral libraries used in remote sensing applications. Specifically, this library was generated to improve the retrieval of global, grain size separable, quartz and feldspar distributions using a Monte Carlo spectral mixing analysis (SMA) strategy. To address the gap in VTIR spectral libraries, we assembled the Caltech/NHM spectral library – a new dataset of 27 combined VTIR spectra of quartz and feldspar minerals with geochemical information about each specimen. Mineral specimens were sourced from and measured on x-ray diffraction (XRD) and x-ray fluorescence (XRF) at the Natural History Museum of Los Angeles County (NHM), and spectroscopic analyses were conducted at the California Institute of Technology (Caltech). This data package contains the raw spectra for all mineral samples, raw XRD and XRF data with ancillary files containing metadata, and photos of the specimens.

Methods

After specimen selection, samples were first cut using a rock saw, then disaggregated with a mortar and pestle, and finally sieved into five grain size fractions: unconsolidated (solid) minerals, fine to coarse sand (125–250 µm), coarse silt to fine sand (50–125 µm), very fine silt to coarse silt (2–50 µm), and clay (<2 µm). Each specimen was analyzed by XRD and each grain size fraction was analyzed by XRF to assess geochemical compositions.

Spectral measurements were acquired using three instruments: an ASD FieldSpec 4 for the VSWIR range and both a Nicolet 860 Magna series FTIR and Thermo-Nicolet iS50 FTIR for the TIR range. VSWIR spectra are reported in reflectance units, while TIR spectra are reported in emissivity units. The dataset includes key mineral types that were underrepresented in prior libraries—such as albite, oligoclase, microcline, bytownite, amazonite, labradorite, and quartz—measured across multiple grain sizes. Unconsolidated and clay sized specimens are not measured in the VSWIR wavelength region. The Caltech/NHM Mineral VTIR Spectral Library is intended to support a range of remote sensing studies, particularly those focused on VTIR spectral applications.

Notes

File Naming Conventions:

VSWIR raw data file names start with the NHM catalog number, followed by the largest grain size fraction (in microns). For example, the number 500 represents the fine to coarse sand grain size fraction. This is followed by four zeros (an ASD spectrometer convention) and a final digit indicating the measurement order, starting at 0 for the first measurement.
TIR raw data file names also begin with the NHM catalog number and grain size (e.g., 500 microns), followed by a letter denoting the measurement sequence, starting with “a”.
VTIR and ancillary XRF text files are named using the NHM catalog number, grain size (in microns), and lastly the mineral name.
XRD files (both .xlm and .xye) begin with the date of measurement (year_month_day), followed by the instrument position, and end with the NHM catalog number.
XRF raw files are simply labeled with the NHM catalog number.

Users can search the NHM database using the catalog number to access detailed information about each specimen.

Mineral Source Locations:

Bakersville, Mitchell, North Carolina, USA

George Howell mine, Spruce Pine, Mitchell, North Carolina, USA

Herkimer, New York, USA; Crystal Peak, Teller, Colorado, USA

Lake Superior, Minnesota, USA

Minas Gerais, Brazil

Nain, Labrador, Canada

Pikes Peak Bath, El Paso, Colorado, USA

Other

Funding: This study was supported by funding provided by both the JPL and NASA's Earth Science Division.

Dataset Acknowledgements: We would like to acknowledge the contributions of the entire EMIT engineering and science teams and the ISS team for enabling the EMIT mission. A portion of this research was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration (80NM0018D0004).

Files