MAPPING MINERALS WITH WORLDVIEW-3 DATA, EXAMPLES FROM NEVADA AND CALIFORNIA

Mineral mapping has benefited greatly in the last several years due to the increase in both spatial and spectral resolution of sensors on commercial satellites. WorldView-3, with 8 channels in the visible/near-infrared (VNIR), and 8 channels in the short wave infrared (SWIR), is the latest sensor with increased spectral capabilities. More channels means more information, so typically more materials can be mapped with data from this sensor than from some other multispectral sensors. For this study two WorldView-3 data sets were analyzed. One data set was collected over Cuprite, Nevada in 2014. The other data set was collected in May 2015 over the Clark Mountain Range near the California/Nevada border. The geology of Cuprite is fairly well known and the site has provided a good test bed for a variety of sensors, including hyperspectral. Exposures at Cuprite reveal an advanced argillic hydrothermal system with various alteration minerals including kaolinite, dickite, alunite, buddingtonite, and silica. These minerals were mapped using simple hyperspectral data processing routines. Results compare favorably with results from the analysis of both ASTER and AVIRIS data. WorldView-3 SWIR bands are very similar to those of ASTER, the main difference is that WorldView-3 does not have a channel near 2400 nm. The lack of a channel near 2400 can hinder the differentiation of such minerals as carbonates, chlorite, and epidote. However, WorldView-3's VNIR bands can help resolve the iron-related absorptions of chlorite to help to differentiate it from calcite and dolomite. This capability is demonstrated by results from the analysis of the Clark Mountain Range data set.