Visible/Shortwave and Thermal Infrared Measurements at Cuprite, Nevada: Complementary or Corroborating?

Abstract

Remote mapping of surface mineral composition often involves Visible to short wavelength infrared (VSWIR) and Thermal Infrared (TIR) imaging spectroscopy, as well as multiband instruments. This makes geologic mapping a compelling test case for quantitative information studies of remote sensing instrument suites. The degree to which instruments either corroborate or complement each other depends on the specific investigation objectives. Conventionally, VSWIR and TIR spectral windows are thought to be complementary; the VSWIR is highly sensitive to iron oxides, phyllosilicates, and iron oxides, while the TIR reveals Neso- and Tectosilicate minerals. In this work, we utilize state-of-the-art statistical techniques that model and exploit the underlying patterns between both types of measurements during mineralogical investigations. We use these models to quantitatively compute the shared information content between VSWIR and TIR measurements. Furthermore, we use these methods to leverage scene statistics to enable bidirectional predictions; that is, predict VSWIR observations from TIR measurements, and vice versa. We present a case study that consists in mineralogical investigations at Cuprite, NV using measurements from two different spectrometers: the Airborne Visible/Infrared Imaging Spectrometer Next-Generation (AVIRIS-NG) and the Hyperspectral Thermal Emission Spectrometer (HyTES). Similar models can help budget the remote sensing "flow of information" in other earth science disciplines, and enable more rigorous experimental design during remote planetary exploration.