Compositional Mapping of the Transantarctic Mountains Using Orbital Reflectance Data

Abstract:

We report on our progress of remotely mapping compositional variations throughout the Transantarctic Mountains (TAM) using orbital spectroscopic data. These techniques were originally proven effective in Antarctica using moderate spatial resolution (30 m/pixel) Advanced Land Imager (ALI) data, and showed great successes in identifying even minor variations in composition throughout the McMurdo Dry Valleys (MDV) [Salvatore et al., 2013]. However, due to the orbital inclination of the Earth Observing-1 spacecraft, ALI is unable to image the central and southern TAM, making comparable studies at comparable resolutions impossible on a continental scale. Fortunately, the WorldView-2 satellite (DigitalGlobe, Inc.) boasts high-resolution (2 m/pixel) multispectral capabilities, with 8 spectral bands located between 427 nm and 908 nm, and is able to image the entirety of the TAM through off-nadir pointing capabilities. This provides the ability to continue our remote spectral mapping campaign throughout the TAM to identify compositional variations in support of past and future field operations.

We present an updated map of relative spectral variability (RSV) in the vicinity of Shackleton Glacier. This mapping product consists of 91 individual WorldView-2 images, each corrected to top-of-atmosphere radiance and parameterized to highlight known compositional properties. The mapped area covers approximately 17,850 square kilometers of ice-covered and exposed terrain. Compositional variations are easily mapped, and small-scale variations in iron-bearing mineralogy are particularly well resolved. We also describe our updated atmospheric correction algorithm for the WorldView-2 dataset, which utilizes in-scene techniques to derive surface reflectance and does not necessitate the use of radiative transfer modeling. Our technique is validated using laboratory reflectance measurements. In conjunction with the Polar Rock Repository at the Ohio State University, we have measured hundreds of individual samples in an effort to verify and “ground-truth” this atmospheric removal algorithm. Using these methodologies and revised techniques, our objective is to make a fully calibrated and atmospherically corrected spectral map of the central TAM available to the scientific community.