TY  - JOUR
A1  - Milewski, Robert
A1  - Chabrillat, Sabine
A1  - Brell, Maximillian
A1  - Schleicher, Anja Maria
A1  - Guanter, Luis
T1  - Assessment of the 1.75μm absorption feature for gypsum estimation using laboratory, air- and spaceborne hyperspectral sensors
JF  - International Journal of Applied Earth Observation and Geoinformation
N2  - High spectral resolution (hyperspectral) remote sensing has already demonstrated its capabilities for soil constituent mapping based on absorption feature parameters. This paper tests different parametrizations of the 1.75 μm gypsum feature for the determination of gypsum abundances, from the laboratory to remote sensing applications of recent as well as upcoming hyperspectral sensors. In particular, this study focuses on remote sensing imagery over the large body of the Omongwa pan located in the Namibian Kalahari. Four common absorption feature parameters are compared: band ratio through the introduction of the Normalized Differenced Gypsum Index (NDGI), the shape-based parameters Slope, and Half-Area, and the Continuum Removed Absorption Depth (CRAD). On laboratory soil samples from the pan, CRAD and NDGI approaches perform best to determine gypsum content tested in cross validated regression models with XRD mineralogical data (R² = 0.84 for NDGI and R² = 0.86 for CRAD). Subsequently the laboratory prediction functions are transferred to remote sensing imagery of spaceborne Hyperion, airborne HySpex and simulated spaceborne EnMAP sensor. Variable results were obtained depending on sensor characteristics, data quality, preprocessing and spectral parameters. Overall, the CRAD parameter in this wavelength region proved not to be robust for remote sensing applications, and the simple band ratio based parameter, the NDGI, proved robust and is recommended for future use for the determination of gypsum content in bare soils based on remote sensing hyperspectral imagery.
KW  - Hyperspectral
KW  - Gypsum quantification
KW  - Salt pan
KW  - Absorption feature parameters
KW  - Namibia
Y1  - 2017
U6  - https://doi.org/10.1016/j.jag.2018.12.012
SN  - 0303-2434
VL  - 77
SP  - 69
EP  - 83
PB  - Elsevier Science
CY  - Amsterdam [u.a.]
ER  -