BIROn - Birkbeck Institutional Research Online

    Spectral identification and quantification of salts in the Atacama Desert

    Harris, Jennifer and Cousins, C.R. and Claire, M.W. (2016) Spectral identification and quantification of salts in the Atacama Desert. SPIE Proceedings 10005 , ISSN 0277-786X.

    [img]
    Preview
    Text
    17871.pdf - Author's Accepted Manuscript

    Download (1MB) | Preview

    Abstract

    Salt minerals are an important natural resource. The ability to quickly and remotely identify and quantify salt deposits and salt contaminated soils and sands is therefore a priority goal for the various industries and agencies that utilise salts. The advent of global hyperspectral imagery from instruments such as Hyperion on NASA’s Earth-Observing 1 satellite has opened up a new source of data that can potentially be used for just this task. This study aims to assess the ability of Visible and Near Infrared (VNIR) spectroscopy to identify and quantify salt minerals through the use of spectral mixture analysis. The surface and near-surface soils of the Atacama Desert in Chile contain a variety of well-studied salts, which together with low cloud coverage, and high aridity, makes this region an ideal testbed for this technique. Two forms of spectral data ranging 0.35 – 2.5 μm were collected: laboratory spectra acquired using an ASD FieldSpec Pro instrument on samples from four locations in the Atacama desert known to have surface concentrations of sulfates, nitrates, chlorides and perchlorates; and images from the EO-1 satellite’s Hyperion instrument taken over the same four locations. Mineral identifications and abundances were confirmed using quantitative XRD of the physical samples. Spectral endmembers were extracted from within the laboratory and Hyperion spectral datasets and together with additional spectral library endmembers fed into a linear mixture model. The resulting identification and abundances from both dataset types were verified against the sample XRD values. Issues of spectral scale, SNR and how different mineral spectra interact are considered, and the utility of VNIR spectroscopy and Hyperion in particular for mapping specific salt concentrations in desert environments is established. Overall, SMA was successful at estimating abundances of sulfate minerals, particularly calcium sulfate, from both hyperspectral image and laboratory sample spectra, while abundance estimation of other salt phase spectral end-members was achieved with a higher degree of error.

    Metadata

    Item Type: Article
    Additional Information: Proc. SPIE 10005, Earth Resources and Environmental Remote Sensing/GIS Applications VII. Copyright 2016 Society of Photo Optical Instrumentation Engineers. One print or electronic copy may be made for personal use only. Systematic reproduction and distribution, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper are prohibited.
    School: Birkbeck Schools and Departments > School of Science > Earth and Planetary Sciences > UCL/Birkbeck Centre for Planetary Sciences
    Birkbeck Schools and Departments > School of Science > Earth and Planetary Sciences
    Research Centre: Planetary Sciences, Centre for (CPS)
    Depositing User: Dr Jennifer Harris
    Date Deposited: 13 Jan 2017 09:01
    Last Modified: 29 Jul 2019 21:55
    URI: http://eprints.bbk.ac.uk/id/eprint/17871

    Statistics

    Downloads
    Activity Overview
    198Downloads
    96Hits

    Additional statistics are available via IRStats2.

    Archive Staff Only (login required)

    Edit/View Item Edit/View Item